diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,62940 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.999499081649691,
+  "eval_steps": 500,
+  "global_step": 8982,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0003339455668725998,
+      "grad_norm": 18.08110072524588,
+      "learning_rate": 1.1123470522803115e-08,
+      "loss": 1.1965,
+      "step": 1
+    },
+    {
+      "epoch": 0.0006678911337451996,
+      "grad_norm": 19.552894943582405,
+      "learning_rate": 2.224694104560623e-08,
+      "loss": 1.2467,
+      "step": 2
+    },
+    {
+      "epoch": 0.0010018367006177993,
+      "grad_norm": 18.45573642932544,
+      "learning_rate": 3.337041156840935e-08,
+      "loss": 1.1939,
+      "step": 3
+    },
+    {
+      "epoch": 0.0013357822674903992,
+      "grad_norm": 19.51921878235958,
+      "learning_rate": 4.449388209121246e-08,
+      "loss": 1.2352,
+      "step": 4
+    },
+    {
+      "epoch": 0.0016697278343629988,
+      "grad_norm": 19.438335705136836,
+      "learning_rate": 5.561735261401558e-08,
+      "loss": 1.2511,
+      "step": 5
+    },
+    {
+      "epoch": 0.0020036734012355987,
+      "grad_norm": 19.874015492468935,
+      "learning_rate": 6.67408231368187e-08,
+      "loss": 1.2588,
+      "step": 6
+    },
+    {
+      "epoch": 0.0023376189681081983,
+      "grad_norm": 18.81666516443105,
+      "learning_rate": 7.78642936596218e-08,
+      "loss": 1.1735,
+      "step": 7
+    },
+    {
+      "epoch": 0.0026715645349807983,
+      "grad_norm": 18.796402872724705,
+      "learning_rate": 8.898776418242492e-08,
+      "loss": 1.1907,
+      "step": 8
+    },
+    {
+      "epoch": 0.003005510101853398,
+      "grad_norm": 18.406989033693506,
+      "learning_rate": 1.0011123470522804e-07,
+      "loss": 1.1884,
+      "step": 9
+    },
+    {
+      "epoch": 0.0033394556687259976,
+      "grad_norm": 17.666473890858295,
+      "learning_rate": 1.1123470522803116e-07,
+      "loss": 1.156,
+      "step": 10
+    },
+    {
+      "epoch": 0.0036734012355985972,
+      "grad_norm": 18.312626054013474,
+      "learning_rate": 1.2235817575083427e-07,
+      "loss": 1.1896,
+      "step": 11
+    },
+    {
+      "epoch": 0.004007346802471197,
+      "grad_norm": 19.21746711412166,
+      "learning_rate": 1.334816462736374e-07,
+      "loss": 1.2486,
+      "step": 12
+    },
+    {
+      "epoch": 0.004341292369343797,
+      "grad_norm": 18.645287861138588,
+      "learning_rate": 1.446051167964405e-07,
+      "loss": 1.2237,
+      "step": 13
+    },
+    {
+      "epoch": 0.0046752379362163966,
+      "grad_norm": 18.389360599838056,
+      "learning_rate": 1.557285873192436e-07,
+      "loss": 1.174,
+      "step": 14
+    },
+    {
+      "epoch": 0.005009183503088996,
+      "grad_norm": 18.094045001801177,
+      "learning_rate": 1.6685205784204674e-07,
+      "loss": 1.2052,
+      "step": 15
+    },
+    {
+      "epoch": 0.005343129069961597,
+      "grad_norm": 18.4374027292432,
+      "learning_rate": 1.7797552836484985e-07,
+      "loss": 1.2309,
+      "step": 16
+    },
+    {
+      "epoch": 0.005677074636834196,
+      "grad_norm": 17.289354366652493,
+      "learning_rate": 1.8909899888765295e-07,
+      "loss": 1.1751,
+      "step": 17
+    },
+    {
+      "epoch": 0.006011020203706796,
+      "grad_norm": 18.24833811594456,
+      "learning_rate": 2.0022246941045608e-07,
+      "loss": 1.2278,
+      "step": 18
+    },
+    {
+      "epoch": 0.006344965770579396,
+      "grad_norm": 17.053322563470108,
+      "learning_rate": 2.113459399332592e-07,
+      "loss": 1.1721,
+      "step": 19
+    },
+    {
+      "epoch": 0.006678911337451995,
+      "grad_norm": 17.129792113985143,
+      "learning_rate": 2.2246941045606232e-07,
+      "loss": 1.1354,
+      "step": 20
+    },
+    {
+      "epoch": 0.007012856904324595,
+      "grad_norm": 13.47458828617283,
+      "learning_rate": 2.3359288097886543e-07,
+      "loss": 1.0349,
+      "step": 21
+    },
+    {
+      "epoch": 0.0073468024711971945,
+      "grad_norm": 14.161487265468253,
+      "learning_rate": 2.4471635150166853e-07,
+      "loss": 1.12,
+      "step": 22
+    },
+    {
+      "epoch": 0.007680748038069795,
+      "grad_norm": 13.161488444055747,
+      "learning_rate": 2.5583982202447166e-07,
+      "loss": 1.0514,
+      "step": 23
+    },
+    {
+      "epoch": 0.008014693604942395,
+      "grad_norm": 14.570394994187327,
+      "learning_rate": 2.669632925472748e-07,
+      "loss": 1.123,
+      "step": 24
+    },
+    {
+      "epoch": 0.008348639171814994,
+      "grad_norm": 14.666246390111493,
+      "learning_rate": 2.780867630700779e-07,
+      "loss": 1.1715,
+      "step": 25
+    },
+    {
+      "epoch": 0.008682584738687594,
+      "grad_norm": 14.577788314814974,
+      "learning_rate": 2.89210233592881e-07,
+      "loss": 1.1681,
+      "step": 26
+    },
+    {
+      "epoch": 0.009016530305560193,
+      "grad_norm": 13.214866163706802,
+      "learning_rate": 3.003337041156841e-07,
+      "loss": 1.0505,
+      "step": 27
+    },
+    {
+      "epoch": 0.009350475872432793,
+      "grad_norm": 10.769628248339709,
+      "learning_rate": 3.114571746384872e-07,
+      "loss": 0.9765,
+      "step": 28
+    },
+    {
+      "epoch": 0.009684421439305393,
+      "grad_norm": 10.098307468296536,
+      "learning_rate": 3.2258064516129035e-07,
+      "loss": 0.9672,
+      "step": 29
+    },
+    {
+      "epoch": 0.010018367006177992,
+      "grad_norm": 10.418771316687504,
+      "learning_rate": 3.337041156840935e-07,
+      "loss": 0.9561,
+      "step": 30
+    },
+    {
+      "epoch": 0.010352312573050592,
+      "grad_norm": 8.978717627730614,
+      "learning_rate": 3.4482758620689656e-07,
+      "loss": 0.8463,
+      "step": 31
+    },
+    {
+      "epoch": 0.010686258139923193,
+      "grad_norm": 10.286653113951482,
+      "learning_rate": 3.559510567296997e-07,
+      "loss": 0.9665,
+      "step": 32
+    },
+    {
+      "epoch": 0.011020203706795793,
+      "grad_norm": 10.462353476788321,
+      "learning_rate": 3.670745272525028e-07,
+      "loss": 0.9434,
+      "step": 33
+    },
+    {
+      "epoch": 0.011354149273668393,
+      "grad_norm": 9.714490170332663,
+      "learning_rate": 3.781979977753059e-07,
+      "loss": 0.9324,
+      "step": 34
+    },
+    {
+      "epoch": 0.011688094840540992,
+      "grad_norm": 10.083825623996413,
+      "learning_rate": 3.8932146829810904e-07,
+      "loss": 0.9361,
+      "step": 35
+    },
+    {
+      "epoch": 0.012022040407413592,
+      "grad_norm": 9.694323120504304,
+      "learning_rate": 4.0044493882091217e-07,
+      "loss": 0.9045,
+      "step": 36
+    },
+    {
+      "epoch": 0.012355985974286192,
+      "grad_norm": 9.10401434831957,
+      "learning_rate": 4.115684093437153e-07,
+      "loss": 0.8305,
+      "step": 37
+    },
+    {
+      "epoch": 0.012689931541158791,
+      "grad_norm": 8.912869023969314,
+      "learning_rate": 4.226918798665184e-07,
+      "loss": 0.7016,
+      "step": 38
+    },
+    {
+      "epoch": 0.01302387710803139,
+      "grad_norm": 9.949234492145816,
+      "learning_rate": 4.338153503893215e-07,
+      "loss": 0.6717,
+      "step": 39
+    },
+    {
+      "epoch": 0.01335782267490399,
+      "grad_norm": 9.107965901272914,
+      "learning_rate": 4.4493882091212464e-07,
+      "loss": 0.6031,
+      "step": 40
+    },
+    {
+      "epoch": 0.01369176824177659,
+      "grad_norm": 8.999748095069597,
+      "learning_rate": 4.560622914349278e-07,
+      "loss": 0.6396,
+      "step": 41
+    },
+    {
+      "epoch": 0.01402571380864919,
+      "grad_norm": 8.795599856241033,
+      "learning_rate": 4.6718576195773085e-07,
+      "loss": 0.5788,
+      "step": 42
+    },
+    {
+      "epoch": 0.01435965937552179,
+      "grad_norm": 9.433146095223425,
+      "learning_rate": 4.783092324805339e-07,
+      "loss": 0.6102,
+      "step": 43
+    },
+    {
+      "epoch": 0.014693604942394389,
+      "grad_norm": 7.960008599611662,
+      "learning_rate": 4.894327030033371e-07,
+      "loss": 0.5373,
+      "step": 44
+    },
+    {
+      "epoch": 0.01502755050926699,
+      "grad_norm": 7.366530046034347,
+      "learning_rate": 5.005561735261402e-07,
+      "loss": 0.5052,
+      "step": 45
+    },
+    {
+      "epoch": 0.01536149607613959,
+      "grad_norm": 6.363036907340728,
+      "learning_rate": 5.116796440489433e-07,
+      "loss": 0.4649,
+      "step": 46
+    },
+    {
+      "epoch": 0.01569544164301219,
+      "grad_norm": 5.556475848189961,
+      "learning_rate": 5.228031145717465e-07,
+      "loss": 0.4856,
+      "step": 47
+    },
+    {
+      "epoch": 0.01602938720988479,
+      "grad_norm": 3.83682323248533,
+      "learning_rate": 5.339265850945496e-07,
+      "loss": 0.4527,
+      "step": 48
+    },
+    {
+      "epoch": 0.01636333277675739,
+      "grad_norm": 3.4969259770523546,
+      "learning_rate": 5.450500556173527e-07,
+      "loss": 0.4612,
+      "step": 49
+    },
+    {
+      "epoch": 0.01669727834362999,
+      "grad_norm": 2.9909282013080722,
+      "learning_rate": 5.561735261401558e-07,
+      "loss": 0.4505,
+      "step": 50
+    },
+    {
+      "epoch": 0.017031223910502588,
+      "grad_norm": 2.7014984810055025,
+      "learning_rate": 5.672969966629589e-07,
+      "loss": 0.4161,
+      "step": 51
+    },
+    {
+      "epoch": 0.017365169477375188,
+      "grad_norm": 2.6387827313567604,
+      "learning_rate": 5.78420467185762e-07,
+      "loss": 0.4301,
+      "step": 52
+    },
+    {
+      "epoch": 0.017699115044247787,
+      "grad_norm": 2.4645004633554577,
+      "learning_rate": 5.89543937708565e-07,
+      "loss": 0.4103,
+      "step": 53
+    },
+    {
+      "epoch": 0.018033060611120387,
+      "grad_norm": 2.148956503468746,
+      "learning_rate": 6.006674082313682e-07,
+      "loss": 0.4049,
+      "step": 54
+    },
+    {
+      "epoch": 0.018367006177992987,
+      "grad_norm": 2.0812867730472076,
+      "learning_rate": 6.117908787541713e-07,
+      "loss": 0.4195,
+      "step": 55
+    },
+    {
+      "epoch": 0.018700951744865586,
+      "grad_norm": 1.97350230594877,
+      "learning_rate": 6.229143492769744e-07,
+      "loss": 0.3871,
+      "step": 56
+    },
+    {
+      "epoch": 0.019034897311738186,
+      "grad_norm": 2.0261853192304873,
+      "learning_rate": 6.340378197997777e-07,
+      "loss": 0.3907,
+      "step": 57
+    },
+    {
+      "epoch": 0.019368842878610786,
+      "grad_norm": 1.7884536259702397,
+      "learning_rate": 6.451612903225807e-07,
+      "loss": 0.3725,
+      "step": 58
+    },
+    {
+      "epoch": 0.019702788445483385,
+      "grad_norm": 1.7578524964487043,
+      "learning_rate": 6.562847608453838e-07,
+      "loss": 0.364,
+      "step": 59
+    },
+    {
+      "epoch": 0.020036734012355985,
+      "grad_norm": 1.575642860109224,
+      "learning_rate": 6.67408231368187e-07,
+      "loss": 0.3917,
+      "step": 60
+    },
+    {
+      "epoch": 0.020370679579228584,
+      "grad_norm": 1.4422891613432185,
+      "learning_rate": 6.785317018909901e-07,
+      "loss": 0.3494,
+      "step": 61
+    },
+    {
+      "epoch": 0.020704625146101184,
+      "grad_norm": 1.522233247296773,
+      "learning_rate": 6.896551724137931e-07,
+      "loss": 0.3539,
+      "step": 62
+    },
+    {
+      "epoch": 0.021038570712973784,
+      "grad_norm": 1.6394568552150803,
+      "learning_rate": 7.007786429365964e-07,
+      "loss": 0.3686,
+      "step": 63
+    },
+    {
+      "epoch": 0.021372516279846387,
+      "grad_norm": 1.4992220967488337,
+      "learning_rate": 7.119021134593994e-07,
+      "loss": 0.3675,
+      "step": 64
+    },
+    {
+      "epoch": 0.021706461846718986,
+      "grad_norm": 1.7873771972580739,
+      "learning_rate": 7.230255839822026e-07,
+      "loss": 0.3792,
+      "step": 65
+    },
+    {
+      "epoch": 0.022040407413591586,
+      "grad_norm": 1.5301039424791039,
+      "learning_rate": 7.341490545050057e-07,
+      "loss": 0.3561,
+      "step": 66
+    },
+    {
+      "epoch": 0.022374352980464186,
+      "grad_norm": 1.462805387783217,
+      "learning_rate": 7.452725250278087e-07,
+      "loss": 0.363,
+      "step": 67
+    },
+    {
+      "epoch": 0.022708298547336785,
+      "grad_norm": 1.4879055388402056,
+      "learning_rate": 7.563959955506118e-07,
+      "loss": 0.3724,
+      "step": 68
+    },
+    {
+      "epoch": 0.023042244114209385,
+      "grad_norm": 1.4681875012051924,
+      "learning_rate": 7.675194660734149e-07,
+      "loss": 0.3651,
+      "step": 69
+    },
+    {
+      "epoch": 0.023376189681081985,
+      "grad_norm": 1.547854625804747,
+      "learning_rate": 7.786429365962181e-07,
+      "loss": 0.3816,
+      "step": 70
+    },
+    {
+      "epoch": 0.023710135247954584,
+      "grad_norm": 1.7263953674720272,
+      "learning_rate": 7.897664071190211e-07,
+      "loss": 0.3699,
+      "step": 71
+    },
+    {
+      "epoch": 0.024044080814827184,
+      "grad_norm": 1.2751997064298404,
+      "learning_rate": 8.008898776418243e-07,
+      "loss": 0.3418,
+      "step": 72
+    },
+    {
+      "epoch": 0.024378026381699783,
+      "grad_norm": 1.4359100529226376,
+      "learning_rate": 8.120133481646274e-07,
+      "loss": 0.3693,
+      "step": 73
+    },
+    {
+      "epoch": 0.024711971948572383,
+      "grad_norm": 1.2749760525328362,
+      "learning_rate": 8.231368186874306e-07,
+      "loss": 0.3323,
+      "step": 74
+    },
+    {
+      "epoch": 0.025045917515444983,
+      "grad_norm": 1.2855735849326455,
+      "learning_rate": 8.342602892102336e-07,
+      "loss": 0.3407,
+      "step": 75
+    },
+    {
+      "epoch": 0.025379863082317582,
+      "grad_norm": 1.2508196970444088,
+      "learning_rate": 8.453837597330368e-07,
+      "loss": 0.3518,
+      "step": 76
+    },
+    {
+      "epoch": 0.025713808649190182,
+      "grad_norm": 1.406994531333698,
+      "learning_rate": 8.565072302558399e-07,
+      "loss": 0.3381,
+      "step": 77
+    },
+    {
+      "epoch": 0.02604775421606278,
+      "grad_norm": 1.4630422438671775,
+      "learning_rate": 8.67630700778643e-07,
+      "loss": 0.3213,
+      "step": 78
+    },
+    {
+      "epoch": 0.02638169978293538,
+      "grad_norm": 1.4111611274728295,
+      "learning_rate": 8.78754171301446e-07,
+      "loss": 0.348,
+      "step": 79
+    },
+    {
+      "epoch": 0.02671564534980798,
+      "grad_norm": 1.2796063016816839,
+      "learning_rate": 8.898776418242493e-07,
+      "loss": 0.3248,
+      "step": 80
+    },
+    {
+      "epoch": 0.02704959091668058,
+      "grad_norm": 1.5090068302982718,
+      "learning_rate": 9.010011123470523e-07,
+      "loss": 0.3559,
+      "step": 81
+    },
+    {
+      "epoch": 0.02738353648355318,
+      "grad_norm": 1.2343790950745581,
+      "learning_rate": 9.121245828698556e-07,
+      "loss": 0.3082,
+      "step": 82
+    },
+    {
+      "epoch": 0.02771748205042578,
+      "grad_norm": 1.1891735311805842,
+      "learning_rate": 9.232480533926586e-07,
+      "loss": 0.3116,
+      "step": 83
+    },
+    {
+      "epoch": 0.02805142761729838,
+      "grad_norm": 1.3062993215992929,
+      "learning_rate": 9.343715239154617e-07,
+      "loss": 0.3292,
+      "step": 84
+    },
+    {
+      "epoch": 0.02838537318417098,
+      "grad_norm": 1.2819392779749308,
+      "learning_rate": 9.454949944382647e-07,
+      "loss": 0.3395,
+      "step": 85
+    },
+    {
+      "epoch": 0.02871931875104358,
+      "grad_norm": 1.4416842302356845,
+      "learning_rate": 9.566184649610679e-07,
+      "loss": 0.336,
+      "step": 86
+    },
+    {
+      "epoch": 0.02905326431791618,
+      "grad_norm": 1.3987710774933,
+      "learning_rate": 9.67741935483871e-07,
+      "loss": 0.3063,
+      "step": 87
+    },
+    {
+      "epoch": 0.029387209884788778,
+      "grad_norm": 1.4418014348260515,
+      "learning_rate": 9.788654060066741e-07,
+      "loss": 0.3607,
+      "step": 88
+    },
+    {
+      "epoch": 0.029721155451661378,
+      "grad_norm": 1.585329350183969,
+      "learning_rate": 9.899888765294773e-07,
+      "loss": 0.3271,
+      "step": 89
+    },
+    {
+      "epoch": 0.03005510101853398,
+      "grad_norm": 1.6528669573974764,
+      "learning_rate": 1.0011123470522804e-06,
+      "loss": 0.3312,
+      "step": 90
+    },
+    {
+      "epoch": 0.03038904658540658,
+      "grad_norm": 1.5673416929359354,
+      "learning_rate": 1.0122358175750835e-06,
+      "loss": 0.3362,
+      "step": 91
+    },
+    {
+      "epoch": 0.03072299215227918,
+      "grad_norm": 1.3538512444375355,
+      "learning_rate": 1.0233592880978867e-06,
+      "loss": 0.3103,
+      "step": 92
+    },
+    {
+      "epoch": 0.03105693771915178,
+      "grad_norm": 1.266414511944769,
+      "learning_rate": 1.0344827586206898e-06,
+      "loss": 0.3476,
+      "step": 93
+    },
+    {
+      "epoch": 0.03139088328602438,
+      "grad_norm": 1.2381813473063292,
+      "learning_rate": 1.045606229143493e-06,
+      "loss": 0.322,
+      "step": 94
+    },
+    {
+      "epoch": 0.03172482885289698,
+      "grad_norm": 1.3517795579781238,
+      "learning_rate": 1.056729699666296e-06,
+      "loss": 0.3428,
+      "step": 95
+    },
+    {
+      "epoch": 0.03205877441976958,
+      "grad_norm": 1.2503553987008789,
+      "learning_rate": 1.0678531701890992e-06,
+      "loss": 0.3324,
+      "step": 96
+    },
+    {
+      "epoch": 0.03239271998664218,
+      "grad_norm": 1.2235838797074612,
+      "learning_rate": 1.0789766407119021e-06,
+      "loss": 0.3077,
+      "step": 97
+    },
+    {
+      "epoch": 0.03272666555351478,
+      "grad_norm": 1.2342368229237566,
+      "learning_rate": 1.0901001112347055e-06,
+      "loss": 0.3123,
+      "step": 98
+    },
+    {
+      "epoch": 0.03306061112038738,
+      "grad_norm": 1.3993217559841238,
+      "learning_rate": 1.1012235817575084e-06,
+      "loss": 0.3157,
+      "step": 99
+    },
+    {
+      "epoch": 0.03339455668725998,
+      "grad_norm": 1.187047636224087,
+      "learning_rate": 1.1123470522803115e-06,
+      "loss": 0.3115,
+      "step": 100
+    },
+    {
+      "epoch": 0.03372850225413258,
+      "grad_norm": 1.1888949396982949,
+      "learning_rate": 1.1234705228031146e-06,
+      "loss": 0.314,
+      "step": 101
+    },
+    {
+      "epoch": 0.034062447821005176,
+      "grad_norm": 1.3282629053014763,
+      "learning_rate": 1.1345939933259178e-06,
+      "loss": 0.3134,
+      "step": 102
+    },
+    {
+      "epoch": 0.034396393387877776,
+      "grad_norm": 1.409494192960795,
+      "learning_rate": 1.145717463848721e-06,
+      "loss": 0.343,
+      "step": 103
+    },
+    {
+      "epoch": 0.034730338954750375,
+      "grad_norm": 1.3955053377524733,
+      "learning_rate": 1.156840934371524e-06,
+      "loss": 0.3294,
+      "step": 104
+    },
+    {
+      "epoch": 0.035064284521622975,
+      "grad_norm": 1.1921009474473023,
+      "learning_rate": 1.1679644048943272e-06,
+      "loss": 0.2927,
+      "step": 105
+    },
+    {
+      "epoch": 0.035398230088495575,
+      "grad_norm": 1.342026976228603,
+      "learning_rate": 1.17908787541713e-06,
+      "loss": 0.3005,
+      "step": 106
+    },
+    {
+      "epoch": 0.035732175655368174,
+      "grad_norm": 1.4229030988806093,
+      "learning_rate": 1.1902113459399334e-06,
+      "loss": 0.333,
+      "step": 107
+    },
+    {
+      "epoch": 0.036066121222240774,
+      "grad_norm": 1.1373972993813148,
+      "learning_rate": 1.2013348164627363e-06,
+      "loss": 0.3059,
+      "step": 108
+    },
+    {
+      "epoch": 0.036400066789113374,
+      "grad_norm": 1.3470188636614417,
+      "learning_rate": 1.2124582869855397e-06,
+      "loss": 0.3168,
+      "step": 109
+    },
+    {
+      "epoch": 0.03673401235598597,
+      "grad_norm": 1.2640906475959344,
+      "learning_rate": 1.2235817575083426e-06,
+      "loss": 0.3064,
+      "step": 110
+    },
+    {
+      "epoch": 0.03706795792285857,
+      "grad_norm": 1.244712541251493,
+      "learning_rate": 1.2347052280311457e-06,
+      "loss": 0.3063,
+      "step": 111
+    },
+    {
+      "epoch": 0.03740190348973117,
+      "grad_norm": 1.0580996310724993,
+      "learning_rate": 1.2458286985539489e-06,
+      "loss": 0.2849,
+      "step": 112
+    },
+    {
+      "epoch": 0.03773584905660377,
+      "grad_norm": 1.3180651427074217,
+      "learning_rate": 1.256952169076752e-06,
+      "loss": 0.3112,
+      "step": 113
+    },
+    {
+      "epoch": 0.03806979462347637,
+      "grad_norm": 1.285650915666917,
+      "learning_rate": 1.2680756395995554e-06,
+      "loss": 0.3156,
+      "step": 114
+    },
+    {
+      "epoch": 0.03840374019034897,
+      "grad_norm": 1.2921513400307592,
+      "learning_rate": 1.2791991101223583e-06,
+      "loss": 0.3102,
+      "step": 115
+    },
+    {
+      "epoch": 0.03873768575722157,
+      "grad_norm": 1.1878797730987274,
+      "learning_rate": 1.2903225806451614e-06,
+      "loss": 0.2986,
+      "step": 116
+    },
+    {
+      "epoch": 0.03907163132409417,
+      "grad_norm": 1.2560584793651444,
+      "learning_rate": 1.3014460511679643e-06,
+      "loss": 0.2989,
+      "step": 117
+    },
+    {
+      "epoch": 0.03940557689096677,
+      "grad_norm": 1.2643345435783993,
+      "learning_rate": 1.3125695216907677e-06,
+      "loss": 0.3233,
+      "step": 118
+    },
+    {
+      "epoch": 0.03973952245783937,
+      "grad_norm": 1.3041047613547743,
+      "learning_rate": 1.3236929922135708e-06,
+      "loss": 0.336,
+      "step": 119
+    },
+    {
+      "epoch": 0.04007346802471197,
+      "grad_norm": 1.2031369067707778,
+      "learning_rate": 1.334816462736374e-06,
+      "loss": 0.304,
+      "step": 120
+    },
+    {
+      "epoch": 0.04040741359158457,
+      "grad_norm": 1.1399223289092129,
+      "learning_rate": 1.3459399332591769e-06,
+      "loss": 0.2892,
+      "step": 121
+    },
+    {
+      "epoch": 0.04074135915845717,
+      "grad_norm": 1.303855873648094,
+      "learning_rate": 1.3570634037819802e-06,
+      "loss": 0.3095,
+      "step": 122
+    },
+    {
+      "epoch": 0.04107530472532977,
+      "grad_norm": 1.0649829493695344,
+      "learning_rate": 1.3681868743047833e-06,
+      "loss": 0.285,
+      "step": 123
+    },
+    {
+      "epoch": 0.04140925029220237,
+      "grad_norm": 1.415926257488088,
+      "learning_rate": 1.3793103448275862e-06,
+      "loss": 0.3394,
+      "step": 124
+    },
+    {
+      "epoch": 0.04174319585907497,
+      "grad_norm": 1.183128643867534,
+      "learning_rate": 1.3904338153503894e-06,
+      "loss": 0.3079,
+      "step": 125
+    },
+    {
+      "epoch": 0.04207714142594757,
+      "grad_norm": 1.3177462295230522,
+      "learning_rate": 1.4015572858731927e-06,
+      "loss": 0.3369,
+      "step": 126
+    },
+    {
+      "epoch": 0.04241108699282017,
+      "grad_norm": 1.3473486897101608,
+      "learning_rate": 1.4126807563959956e-06,
+      "loss": 0.3045,
+      "step": 127
+    },
+    {
+      "epoch": 0.042745032559692774,
+      "grad_norm": 1.2390438395244772,
+      "learning_rate": 1.4238042269187988e-06,
+      "loss": 0.2975,
+      "step": 128
+    },
+    {
+      "epoch": 0.04307897812656537,
+      "grad_norm": 1.2740946670644597,
+      "learning_rate": 1.434927697441602e-06,
+      "loss": 0.3059,
+      "step": 129
+    },
+    {
+      "epoch": 0.04341292369343797,
+      "grad_norm": 1.3012091566446777,
+      "learning_rate": 1.4460511679644053e-06,
+      "loss": 0.3079,
+      "step": 130
+    },
+    {
+      "epoch": 0.04374686926031057,
+      "grad_norm": 1.1947097037496526,
+      "learning_rate": 1.4571746384872082e-06,
+      "loss": 0.3052,
+      "step": 131
+    },
+    {
+      "epoch": 0.04408081482718317,
+      "grad_norm": 1.1432786230589478,
+      "learning_rate": 1.4682981090100113e-06,
+      "loss": 0.3139,
+      "step": 132
+    },
+    {
+      "epoch": 0.04441476039405577,
+      "grad_norm": 1.3142624785291002,
+      "learning_rate": 1.4794215795328142e-06,
+      "loss": 0.3004,
+      "step": 133
+    },
+    {
+      "epoch": 0.04474870596092837,
+      "grad_norm": 1.1091244974414516,
+      "learning_rate": 1.4905450500556174e-06,
+      "loss": 0.2872,
+      "step": 134
+    },
+    {
+      "epoch": 0.04508265152780097,
+      "grad_norm": 1.3022437706824233,
+      "learning_rate": 1.5016685205784207e-06,
+      "loss": 0.3114,
+      "step": 135
+    },
+    {
+      "epoch": 0.04541659709467357,
+      "grad_norm": 1.323275845400312,
+      "learning_rate": 1.5127919911012236e-06,
+      "loss": 0.3265,
+      "step": 136
+    },
+    {
+      "epoch": 0.04575054266154617,
+      "grad_norm": 1.24079589522644,
+      "learning_rate": 1.5239154616240268e-06,
+      "loss": 0.2973,
+      "step": 137
+    },
+    {
+      "epoch": 0.04608448822841877,
+      "grad_norm": 1.0994743200549912,
+      "learning_rate": 1.5350389321468299e-06,
+      "loss": 0.2917,
+      "step": 138
+    },
+    {
+      "epoch": 0.04641843379529137,
+      "grad_norm": 1.1757509816140939,
+      "learning_rate": 1.5461624026696332e-06,
+      "loss": 0.2896,
+      "step": 139
+    },
+    {
+      "epoch": 0.04675237936216397,
+      "grad_norm": 1.097534938340695,
+      "learning_rate": 1.5572858731924361e-06,
+      "loss": 0.281,
+      "step": 140
+    },
+    {
+      "epoch": 0.04708632492903657,
+      "grad_norm": 1.1159015324002537,
+      "learning_rate": 1.5684093437152393e-06,
+      "loss": 0.2902,
+      "step": 141
+    },
+    {
+      "epoch": 0.04742027049590917,
+      "grad_norm": 1.2320769642378695,
+      "learning_rate": 1.5795328142380422e-06,
+      "loss": 0.3015,
+      "step": 142
+    },
+    {
+      "epoch": 0.04775421606278177,
+      "grad_norm": 1.2268199663025805,
+      "learning_rate": 1.5906562847608455e-06,
+      "loss": 0.2986,
+      "step": 143
+    },
+    {
+      "epoch": 0.04808816162965437,
+      "grad_norm": 1.106082896796778,
+      "learning_rate": 1.6017797552836487e-06,
+      "loss": 0.2824,
+      "step": 144
+    },
+    {
+      "epoch": 0.04842210719652697,
+      "grad_norm": 1.0688508167162885,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 0.2887,
+      "step": 145
+    },
+    {
+      "epoch": 0.04875605276339957,
+      "grad_norm": 1.2355684355753798,
+      "learning_rate": 1.6240266963292547e-06,
+      "loss": 0.3029,
+      "step": 146
+    },
+    {
+      "epoch": 0.049089998330272167,
+      "grad_norm": 1.2608745789622653,
+      "learning_rate": 1.635150166852058e-06,
+      "loss": 0.3053,
+      "step": 147
+    },
+    {
+      "epoch": 0.049423943897144766,
+      "grad_norm": 1.2087145826619097,
+      "learning_rate": 1.6462736373748612e-06,
+      "loss": 0.3074,
+      "step": 148
+    },
+    {
+      "epoch": 0.049757889464017366,
+      "grad_norm": 1.4191211752662398,
+      "learning_rate": 1.6573971078976641e-06,
+      "loss": 0.3018,
+      "step": 149
+    },
+    {
+      "epoch": 0.050091835030889965,
+      "grad_norm": 1.6005040312929983,
+      "learning_rate": 1.6685205784204673e-06,
+      "loss": 0.2992,
+      "step": 150
+    },
+    {
+      "epoch": 0.050425780597762565,
+      "grad_norm": 1.3177431441471539,
+      "learning_rate": 1.6796440489432706e-06,
+      "loss": 0.3013,
+      "step": 151
+    },
+    {
+      "epoch": 0.050759726164635165,
+      "grad_norm": 1.0667233844048953,
+      "learning_rate": 1.6907675194660735e-06,
+      "loss": 0.2853,
+      "step": 152
+    },
+    {
+      "epoch": 0.051093671731507764,
+      "grad_norm": 1.2197105719883872,
+      "learning_rate": 1.7018909899888767e-06,
+      "loss": 0.2957,
+      "step": 153
+    },
+    {
+      "epoch": 0.051427617298380364,
+      "grad_norm": 1.2749427439891368,
+      "learning_rate": 1.7130144605116798e-06,
+      "loss": 0.2826,
+      "step": 154
+    },
+    {
+      "epoch": 0.051761562865252964,
+      "grad_norm": 1.3919022833764747,
+      "learning_rate": 1.724137931034483e-06,
+      "loss": 0.2828,
+      "step": 155
+    },
+    {
+      "epoch": 0.05209550843212556,
+      "grad_norm": 1.1706667684484162,
+      "learning_rate": 1.735261401557286e-06,
+      "loss": 0.2834,
+      "step": 156
+    },
+    {
+      "epoch": 0.05242945399899816,
+      "grad_norm": 1.1385584136115947,
+      "learning_rate": 1.7463848720800892e-06,
+      "loss": 0.278,
+      "step": 157
+    },
+    {
+      "epoch": 0.05276339956587076,
+      "grad_norm": 1.2784468592463725,
+      "learning_rate": 1.757508342602892e-06,
+      "loss": 0.2878,
+      "step": 158
+    },
+    {
+      "epoch": 0.05309734513274336,
+      "grad_norm": 1.0828791709613537,
+      "learning_rate": 1.7686318131256954e-06,
+      "loss": 0.2843,
+      "step": 159
+    },
+    {
+      "epoch": 0.05343129069961596,
+      "grad_norm": 1.0858986938430426,
+      "learning_rate": 1.7797552836484986e-06,
+      "loss": 0.2803,
+      "step": 160
+    },
+    {
+      "epoch": 0.05376523626648856,
+      "grad_norm": 1.4160711682479377,
+      "learning_rate": 1.7908787541713015e-06,
+      "loss": 0.2681,
+      "step": 161
+    },
+    {
+      "epoch": 0.05409918183336116,
+      "grad_norm": 1.1622241308379426,
+      "learning_rate": 1.8020022246941046e-06,
+      "loss": 0.2942,
+      "step": 162
+    },
+    {
+      "epoch": 0.05443312740023376,
+      "grad_norm": 1.3244697401189178,
+      "learning_rate": 1.813125695216908e-06,
+      "loss": 0.3069,
+      "step": 163
+    },
+    {
+      "epoch": 0.05476707296710636,
+      "grad_norm": 1.1651688282873625,
+      "learning_rate": 1.824249165739711e-06,
+      "loss": 0.2942,
+      "step": 164
+    },
+    {
+      "epoch": 0.05510101853397896,
+      "grad_norm": 1.3401516990160998,
+      "learning_rate": 1.835372636262514e-06,
+      "loss": 0.2979,
+      "step": 165
+    },
+    {
+      "epoch": 0.05543496410085156,
+      "grad_norm": 1.1232106728216595,
+      "learning_rate": 1.8464961067853172e-06,
+      "loss": 0.2919,
+      "step": 166
+    },
+    {
+      "epoch": 0.05576890966772416,
+      "grad_norm": 0.9645342716426896,
+      "learning_rate": 1.85761957730812e-06,
+      "loss": 0.2833,
+      "step": 167
+    },
+    {
+      "epoch": 0.05610285523459676,
+      "grad_norm": 1.2735387795044881,
+      "learning_rate": 1.8687430478309234e-06,
+      "loss": 0.311,
+      "step": 168
+    },
+    {
+      "epoch": 0.05643680080146936,
+      "grad_norm": 1.1593324040382949,
+      "learning_rate": 1.8798665183537266e-06,
+      "loss": 0.3005,
+      "step": 169
+    },
+    {
+      "epoch": 0.05677074636834196,
+      "grad_norm": 1.0571733025122132,
+      "learning_rate": 1.8909899888765295e-06,
+      "loss": 0.2871,
+      "step": 170
+    },
+    {
+      "epoch": 0.05710469193521456,
+      "grad_norm": 1.1772637818135905,
+      "learning_rate": 1.9021134593993326e-06,
+      "loss": 0.2963,
+      "step": 171
+    },
+    {
+      "epoch": 0.05743863750208716,
+      "grad_norm": 1.2648890164784086,
+      "learning_rate": 1.9132369299221357e-06,
+      "loss": 0.284,
+      "step": 172
+    },
+    {
+      "epoch": 0.05777258306895976,
+      "grad_norm": 1.257741155911166,
+      "learning_rate": 1.924360400444939e-06,
+      "loss": 0.3123,
+      "step": 173
+    },
+    {
+      "epoch": 0.05810652863583236,
+      "grad_norm": 1.2076339575951947,
+      "learning_rate": 1.935483870967742e-06,
+      "loss": 0.2968,
+      "step": 174
+    },
+    {
+      "epoch": 0.058440474202704956,
+      "grad_norm": 1.3831077669462297,
+      "learning_rate": 1.946607341490545e-06,
+      "loss": 0.302,
+      "step": 175
+    },
+    {
+      "epoch": 0.058774419769577556,
+      "grad_norm": 1.2392310046049393,
+      "learning_rate": 1.9577308120133483e-06,
+      "loss": 0.3019,
+      "step": 176
+    },
+    {
+      "epoch": 0.059108365336450155,
+      "grad_norm": 1.0630699642754082,
+      "learning_rate": 1.9688542825361514e-06,
+      "loss": 0.2794,
+      "step": 177
+    },
+    {
+      "epoch": 0.059442310903322755,
+      "grad_norm": 1.2231288730901275,
+      "learning_rate": 1.9799777530589545e-06,
+      "loss": 0.3036,
+      "step": 178
+    },
+    {
+      "epoch": 0.059776256470195355,
+      "grad_norm": 1.054287375357212,
+      "learning_rate": 1.9911012235817577e-06,
+      "loss": 0.2776,
+      "step": 179
+    },
+    {
+      "epoch": 0.06011020203706796,
+      "grad_norm": 1.3813659231450461,
+      "learning_rate": 2.002224694104561e-06,
+      "loss": 0.312,
+      "step": 180
+    },
+    {
+      "epoch": 0.06044414760394056,
+      "grad_norm": 1.223098684904635,
+      "learning_rate": 2.013348164627364e-06,
+      "loss": 0.2986,
+      "step": 181
+    },
+    {
+      "epoch": 0.06077809317081316,
+      "grad_norm": 1.306919633091064,
+      "learning_rate": 2.024471635150167e-06,
+      "loss": 0.3263,
+      "step": 182
+    },
+    {
+      "epoch": 0.06111203873768576,
+      "grad_norm": 1.2007296717327645,
+      "learning_rate": 2.03559510567297e-06,
+      "loss": 0.2856,
+      "step": 183
+    },
+    {
+      "epoch": 0.06144598430455836,
+      "grad_norm": 1.2020773671581122,
+      "learning_rate": 2.0467185761957733e-06,
+      "loss": 0.2777,
+      "step": 184
+    },
+    {
+      "epoch": 0.06177992987143096,
+      "grad_norm": 1.196861175909905,
+      "learning_rate": 2.0578420467185764e-06,
+      "loss": 0.2833,
+      "step": 185
+    },
+    {
+      "epoch": 0.06211387543830356,
+      "grad_norm": 1.2558516865319087,
+      "learning_rate": 2.0689655172413796e-06,
+      "loss": 0.2763,
+      "step": 186
+    },
+    {
+      "epoch": 0.06244782100517616,
+      "grad_norm": 1.0444075488420261,
+      "learning_rate": 2.0800889877641823e-06,
+      "loss": 0.2759,
+      "step": 187
+    },
+    {
+      "epoch": 0.06278176657204876,
+      "grad_norm": 1.4420044463146366,
+      "learning_rate": 2.091212458286986e-06,
+      "loss": 0.3064,
+      "step": 188
+    },
+    {
+      "epoch": 0.06311571213892135,
+      "grad_norm": 1.2416958553865685,
+      "learning_rate": 2.102335928809789e-06,
+      "loss": 0.3063,
+      "step": 189
+    },
+    {
+      "epoch": 0.06344965770579396,
+      "grad_norm": 1.0748461696024694,
+      "learning_rate": 2.113459399332592e-06,
+      "loss": 0.2752,
+      "step": 190
+    },
+    {
+      "epoch": 0.06378360327266655,
+      "grad_norm": 1.1742498266635828,
+      "learning_rate": 2.124582869855395e-06,
+      "loss": 0.2744,
+      "step": 191
+    },
+    {
+      "epoch": 0.06411754883953916,
+      "grad_norm": 1.0861534931135144,
+      "learning_rate": 2.1357063403781984e-06,
+      "loss": 0.2776,
+      "step": 192
+    },
+    {
+      "epoch": 0.06445149440641175,
+      "grad_norm": 1.3347076544814778,
+      "learning_rate": 2.1468298109010015e-06,
+      "loss": 0.2735,
+      "step": 193
+    },
+    {
+      "epoch": 0.06478543997328436,
+      "grad_norm": 1.2383353005287574,
+      "learning_rate": 2.1579532814238042e-06,
+      "loss": 0.274,
+      "step": 194
+    },
+    {
+      "epoch": 0.06511938554015695,
+      "grad_norm": 1.054486003296379,
+      "learning_rate": 2.1690767519466073e-06,
+      "loss": 0.284,
+      "step": 195
+    },
+    {
+      "epoch": 0.06545333110702956,
+      "grad_norm": 1.2422895734951493,
+      "learning_rate": 2.180200222469411e-06,
+      "loss": 0.271,
+      "step": 196
+    },
+    {
+      "epoch": 0.06578727667390215,
+      "grad_norm": 1.2952773756126603,
+      "learning_rate": 2.1913236929922136e-06,
+      "loss": 0.2972,
+      "step": 197
+    },
+    {
+      "epoch": 0.06612122224077475,
+      "grad_norm": 1.2211536848947333,
+      "learning_rate": 2.2024471635150167e-06,
+      "loss": 0.2841,
+      "step": 198
+    },
+    {
+      "epoch": 0.06645516780764735,
+      "grad_norm": 1.197513650785667,
+      "learning_rate": 2.21357063403782e-06,
+      "loss": 0.3226,
+      "step": 199
+    },
+    {
+      "epoch": 0.06678911337451995,
+      "grad_norm": 1.2718959471586204,
+      "learning_rate": 2.224694104560623e-06,
+      "loss": 0.3052,
+      "step": 200
+    },
+    {
+      "epoch": 0.06712305894139255,
+      "grad_norm": 1.11207328849927,
+      "learning_rate": 2.235817575083426e-06,
+      "loss": 0.2752,
+      "step": 201
+    },
+    {
+      "epoch": 0.06745700450826515,
+      "grad_norm": 1.0206175010699121,
+      "learning_rate": 2.2469410456062293e-06,
+      "loss": 0.2919,
+      "step": 202
+    },
+    {
+      "epoch": 0.06779095007513775,
+      "grad_norm": 1.0839833033777182,
+      "learning_rate": 2.2580645161290324e-06,
+      "loss": 0.2942,
+      "step": 203
+    },
+    {
+      "epoch": 0.06812489564201035,
+      "grad_norm": 1.078908508453811,
+      "learning_rate": 2.2691879866518355e-06,
+      "loss": 0.2758,
+      "step": 204
+    },
+    {
+      "epoch": 0.06845884120888296,
+      "grad_norm": 1.4625898609451118,
+      "learning_rate": 2.2803114571746387e-06,
+      "loss": 0.2909,
+      "step": 205
+    },
+    {
+      "epoch": 0.06879278677575555,
+      "grad_norm": 1.1657831625637678,
+      "learning_rate": 2.291434927697442e-06,
+      "loss": 0.2901,
+      "step": 206
+    },
+    {
+      "epoch": 0.06912673234262816,
+      "grad_norm": 1.1163150602739882,
+      "learning_rate": 2.302558398220245e-06,
+      "loss": 0.2836,
+      "step": 207
+    },
+    {
+      "epoch": 0.06946067790950075,
+      "grad_norm": 1.0394061522813909,
+      "learning_rate": 2.313681868743048e-06,
+      "loss": 0.2928,
+      "step": 208
+    },
+    {
+      "epoch": 0.06979462347637336,
+      "grad_norm": 1.0257281218495777,
+      "learning_rate": 2.324805339265851e-06,
+      "loss": 0.282,
+      "step": 209
+    },
+    {
+      "epoch": 0.07012856904324595,
+      "grad_norm": 1.060343943193521,
+      "learning_rate": 2.3359288097886543e-06,
+      "loss": 0.2494,
+      "step": 210
+    },
+    {
+      "epoch": 0.07046251461011856,
+      "grad_norm": 1.1754632761919408,
+      "learning_rate": 2.3470522803114575e-06,
+      "loss": 0.2875,
+      "step": 211
+    },
+    {
+      "epoch": 0.07079646017699115,
+      "grad_norm": 1.3889723569823251,
+      "learning_rate": 2.35817575083426e-06,
+      "loss": 0.2776,
+      "step": 212
+    },
+    {
+      "epoch": 0.07113040574386376,
+      "grad_norm": 1.1026304990108282,
+      "learning_rate": 2.3692992213570637e-06,
+      "loss": 0.2813,
+      "step": 213
+    },
+    {
+      "epoch": 0.07146435131073635,
+      "grad_norm": 1.3540421203167579,
+      "learning_rate": 2.380422691879867e-06,
+      "loss": 0.3084,
+      "step": 214
+    },
+    {
+      "epoch": 0.07179829687760896,
+      "grad_norm": 0.9644640688215597,
+      "learning_rate": 2.39154616240267e-06,
+      "loss": 0.2526,
+      "step": 215
+    },
+    {
+      "epoch": 0.07213224244448155,
+      "grad_norm": 1.077328685424387,
+      "learning_rate": 2.4026696329254727e-06,
+      "loss": 0.2925,
+      "step": 216
+    },
+    {
+      "epoch": 0.07246618801135415,
+      "grad_norm": 1.1941534402934753,
+      "learning_rate": 2.4137931034482762e-06,
+      "loss": 0.2912,
+      "step": 217
+    },
+    {
+      "epoch": 0.07280013357822675,
+      "grad_norm": 1.2216996006544307,
+      "learning_rate": 2.4249165739710794e-06,
+      "loss": 0.2926,
+      "step": 218
+    },
+    {
+      "epoch": 0.07313407914509935,
+      "grad_norm": 1.0288345472230127,
+      "learning_rate": 2.436040044493882e-06,
+      "loss": 0.2699,
+      "step": 219
+    },
+    {
+      "epoch": 0.07346802471197195,
+      "grad_norm": 1.108118580799365,
+      "learning_rate": 2.4471635150166852e-06,
+      "loss": 0.2705,
+      "step": 220
+    },
+    {
+      "epoch": 0.07380197027884455,
+      "grad_norm": 1.0837312318999974,
+      "learning_rate": 2.4582869855394888e-06,
+      "loss": 0.2926,
+      "step": 221
+    },
+    {
+      "epoch": 0.07413591584571715,
+      "grad_norm": 1.078034579186348,
+      "learning_rate": 2.4694104560622915e-06,
+      "loss": 0.2625,
+      "step": 222
+    },
+    {
+      "epoch": 0.07446986141258975,
+      "grad_norm": 1.1387659740587366,
+      "learning_rate": 2.4805339265850946e-06,
+      "loss": 0.2878,
+      "step": 223
+    },
+    {
+      "epoch": 0.07480380697946235,
+      "grad_norm": 1.1289059064466833,
+      "learning_rate": 2.4916573971078977e-06,
+      "loss": 0.2725,
+      "step": 224
+    },
+    {
+      "epoch": 0.07513775254633495,
+      "grad_norm": 1.1154801276679143,
+      "learning_rate": 2.502780867630701e-06,
+      "loss": 0.2796,
+      "step": 225
+    },
+    {
+      "epoch": 0.07547169811320754,
+      "grad_norm": 1.1637231705683622,
+      "learning_rate": 2.513904338153504e-06,
+      "loss": 0.2988,
+      "step": 226
+    },
+    {
+      "epoch": 0.07580564368008015,
+      "grad_norm": 1.0890226181580156,
+      "learning_rate": 2.5250278086763076e-06,
+      "loss": 0.2884,
+      "step": 227
+    },
+    {
+      "epoch": 0.07613958924695274,
+      "grad_norm": 1.133532164208556,
+      "learning_rate": 2.5361512791991107e-06,
+      "loss": 0.2896,
+      "step": 228
+    },
+    {
+      "epoch": 0.07647353481382535,
+      "grad_norm": 1.0864431293764834,
+      "learning_rate": 2.5472747497219134e-06,
+      "loss": 0.2778,
+      "step": 229
+    },
+    {
+      "epoch": 0.07680748038069794,
+      "grad_norm": 1.146220316681448,
+      "learning_rate": 2.5583982202447165e-06,
+      "loss": 0.3008,
+      "step": 230
+    },
+    {
+      "epoch": 0.07714142594757055,
+      "grad_norm": 0.9719070473034946,
+      "learning_rate": 2.5695216907675197e-06,
+      "loss": 0.2865,
+      "step": 231
+    },
+    {
+      "epoch": 0.07747537151444314,
+      "grad_norm": 0.9482297604156307,
+      "learning_rate": 2.580645161290323e-06,
+      "loss": 0.2541,
+      "step": 232
+    },
+    {
+      "epoch": 0.07780931708131575,
+      "grad_norm": 1.154538692385191,
+      "learning_rate": 2.591768631813126e-06,
+      "loss": 0.2743,
+      "step": 233
+    },
+    {
+      "epoch": 0.07814326264818834,
+      "grad_norm": 1.228755098669564,
+      "learning_rate": 2.6028921023359286e-06,
+      "loss": 0.2903,
+      "step": 234
+    },
+    {
+      "epoch": 0.07847720821506095,
+      "grad_norm": 1.225373106150885,
+      "learning_rate": 2.6140155728587318e-06,
+      "loss": 0.2766,
+      "step": 235
+    },
+    {
+      "epoch": 0.07881115378193354,
+      "grad_norm": 1.1650718474694142,
+      "learning_rate": 2.6251390433815353e-06,
+      "loss": 0.2647,
+      "step": 236
+    },
+    {
+      "epoch": 0.07914509934880615,
+      "grad_norm": 0.9954249618414314,
+      "learning_rate": 2.6362625139043385e-06,
+      "loss": 0.263,
+      "step": 237
+    },
+    {
+      "epoch": 0.07947904491567874,
+      "grad_norm": 1.137248317039158,
+      "learning_rate": 2.6473859844271416e-06,
+      "loss": 0.2792,
+      "step": 238
+    },
+    {
+      "epoch": 0.07981299048255135,
+      "grad_norm": 1.0766424441389253,
+      "learning_rate": 2.6585094549499447e-06,
+      "loss": 0.295,
+      "step": 239
+    },
+    {
+      "epoch": 0.08014693604942394,
+      "grad_norm": 1.133299162391096,
+      "learning_rate": 2.669632925472748e-06,
+      "loss": 0.2874,
+      "step": 240
+    },
+    {
+      "epoch": 0.08048088161629655,
+      "grad_norm": 1.6190237814839594,
+      "learning_rate": 2.6807563959955506e-06,
+      "loss": 0.2775,
+      "step": 241
+    },
+    {
+      "epoch": 0.08081482718316914,
+      "grad_norm": 1.0781192090641312,
+      "learning_rate": 2.6918798665183537e-06,
+      "loss": 0.2782,
+      "step": 242
+    },
+    {
+      "epoch": 0.08114877275004174,
+      "grad_norm": 1.1272432543088633,
+      "learning_rate": 2.703003337041157e-06,
+      "loss": 0.2843,
+      "step": 243
+    },
+    {
+      "epoch": 0.08148271831691434,
+      "grad_norm": 1.2925306771786247,
+      "learning_rate": 2.7141268075639604e-06,
+      "loss": 0.2803,
+      "step": 244
+    },
+    {
+      "epoch": 0.08181666388378694,
+      "grad_norm": 1.016699529674434,
+      "learning_rate": 2.7252502780867635e-06,
+      "loss": 0.2605,
+      "step": 245
+    },
+    {
+      "epoch": 0.08215060945065954,
+      "grad_norm": 1.0899875102025414,
+      "learning_rate": 2.7363737486095667e-06,
+      "loss": 0.2841,
+      "step": 246
+    },
+    {
+      "epoch": 0.08248455501753214,
+      "grad_norm": 1.289631959826523,
+      "learning_rate": 2.7474972191323694e-06,
+      "loss": 0.2904,
+      "step": 247
+    },
+    {
+      "epoch": 0.08281850058440474,
+      "grad_norm": 1.2578086396698145,
+      "learning_rate": 2.7586206896551725e-06,
+      "loss": 0.2691,
+      "step": 248
+    },
+    {
+      "epoch": 0.08315244615127734,
+      "grad_norm": 0.9712514626265429,
+      "learning_rate": 2.7697441601779756e-06,
+      "loss": 0.2771,
+      "step": 249
+    },
+    {
+      "epoch": 0.08348639171814994,
+      "grad_norm": 1.2110417920398469,
+      "learning_rate": 2.7808676307007788e-06,
+      "loss": 0.2746,
+      "step": 250
+    },
+    {
+      "epoch": 0.08382033728502254,
+      "grad_norm": 1.1917994392155153,
+      "learning_rate": 2.791991101223582e-06,
+      "loss": 0.281,
+      "step": 251
+    },
+    {
+      "epoch": 0.08415428285189513,
+      "grad_norm": 1.0773885616519046,
+      "learning_rate": 2.8031145717463854e-06,
+      "loss": 0.2766,
+      "step": 252
+    },
+    {
+      "epoch": 0.08448822841876774,
+      "grad_norm": 1.1477058723357543,
+      "learning_rate": 2.8142380422691886e-06,
+      "loss": 0.2868,
+      "step": 253
+    },
+    {
+      "epoch": 0.08482217398564033,
+      "grad_norm": 1.1308647572547197,
+      "learning_rate": 2.8253615127919913e-06,
+      "loss": 0.2809,
+      "step": 254
+    },
+    {
+      "epoch": 0.08515611955251294,
+      "grad_norm": 1.1886012198694917,
+      "learning_rate": 2.8364849833147944e-06,
+      "loss": 0.2838,
+      "step": 255
+    },
+    {
+      "epoch": 0.08549006511938555,
+      "grad_norm": 1.0917302174368306,
+      "learning_rate": 2.8476084538375975e-06,
+      "loss": 0.2966,
+      "step": 256
+    },
+    {
+      "epoch": 0.08582401068625814,
+      "grad_norm": 1.0628706382181317,
+      "learning_rate": 2.8587319243604007e-06,
+      "loss": 0.2746,
+      "step": 257
+    },
+    {
+      "epoch": 0.08615795625313075,
+      "grad_norm": 0.9011209090285437,
+      "learning_rate": 2.869855394883204e-06,
+      "loss": 0.2663,
+      "step": 258
+    },
+    {
+      "epoch": 0.08649190182000334,
+      "grad_norm": 0.9699244069128968,
+      "learning_rate": 2.8809788654060065e-06,
+      "loss": 0.2732,
+      "step": 259
+    },
+    {
+      "epoch": 0.08682584738687595,
+      "grad_norm": 1.0368860734439413,
+      "learning_rate": 2.8921023359288105e-06,
+      "loss": 0.2703,
+      "step": 260
+    },
+    {
+      "epoch": 0.08715979295374854,
+      "grad_norm": 1.1688091750555711,
+      "learning_rate": 2.903225806451613e-06,
+      "loss": 0.2888,
+      "step": 261
+    },
+    {
+      "epoch": 0.08749373852062114,
+      "grad_norm": 1.004959730063203,
+      "learning_rate": 2.9143492769744163e-06,
+      "loss": 0.2576,
+      "step": 262
+    },
+    {
+      "epoch": 0.08782768408749374,
+      "grad_norm": 0.9991808356904461,
+      "learning_rate": 2.9254727474972195e-06,
+      "loss": 0.2625,
+      "step": 263
+    },
+    {
+      "epoch": 0.08816162965436634,
+      "grad_norm": 0.9825510412554146,
+      "learning_rate": 2.9365962180200226e-06,
+      "loss": 0.2632,
+      "step": 264
+    },
+    {
+      "epoch": 0.08849557522123894,
+      "grad_norm": 0.9821153217304937,
+      "learning_rate": 2.9477196885428257e-06,
+      "loss": 0.2713,
+      "step": 265
+    },
+    {
+      "epoch": 0.08882952078811154,
+      "grad_norm": 1.1614726347763442,
+      "learning_rate": 2.9588431590656284e-06,
+      "loss": 0.282,
+      "step": 266
+    },
+    {
+      "epoch": 0.08916346635498414,
+      "grad_norm": 1.7691717435465628,
+      "learning_rate": 2.9699666295884316e-06,
+      "loss": 0.2799,
+      "step": 267
+    },
+    {
+      "epoch": 0.08949741192185674,
+      "grad_norm": 0.9974536728011792,
+      "learning_rate": 2.9810901001112347e-06,
+      "loss": 0.2754,
+      "step": 268
+    },
+    {
+      "epoch": 0.08983135748872934,
+      "grad_norm": 1.1409552236974378,
+      "learning_rate": 2.9922135706340383e-06,
+      "loss": 0.2974,
+      "step": 269
+    },
+    {
+      "epoch": 0.09016530305560194,
+      "grad_norm": 0.938700704239528,
+      "learning_rate": 3.0033370411568414e-06,
+      "loss": 0.2595,
+      "step": 270
+    },
+    {
+      "epoch": 0.09049924862247453,
+      "grad_norm": 1.1094304035830367,
+      "learning_rate": 3.0144605116796445e-06,
+      "loss": 0.2783,
+      "step": 271
+    },
+    {
+      "epoch": 0.09083319418934714,
+      "grad_norm": 1.0187603452237721,
+      "learning_rate": 3.0255839822024472e-06,
+      "loss": 0.272,
+      "step": 272
+    },
+    {
+      "epoch": 0.09116713975621973,
+      "grad_norm": 0.9507482461915819,
+      "learning_rate": 3.0367074527252504e-06,
+      "loss": 0.2786,
+      "step": 273
+    },
+    {
+      "epoch": 0.09150108532309234,
+      "grad_norm": 0.9623017053402451,
+      "learning_rate": 3.0478309232480535e-06,
+      "loss": 0.2563,
+      "step": 274
+    },
+    {
+      "epoch": 0.09183503088996493,
+      "grad_norm": 0.9705519249362503,
+      "learning_rate": 3.0589543937708566e-06,
+      "loss": 0.2634,
+      "step": 275
+    },
+    {
+      "epoch": 0.09216897645683754,
+      "grad_norm": 1.1162146922113336,
+      "learning_rate": 3.0700778642936598e-06,
+      "loss": 0.2814,
+      "step": 276
+    },
+    {
+      "epoch": 0.09250292202371013,
+      "grad_norm": 1.0488911817978013,
+      "learning_rate": 3.0812013348164633e-06,
+      "loss": 0.2736,
+      "step": 277
+    },
+    {
+      "epoch": 0.09283686759058274,
+      "grad_norm": 1.143098412646465,
+      "learning_rate": 3.0923248053392665e-06,
+      "loss": 0.2961,
+      "step": 278
+    },
+    {
+      "epoch": 0.09317081315745533,
+      "grad_norm": 1.0543653338829788,
+      "learning_rate": 3.103448275862069e-06,
+      "loss": 0.2921,
+      "step": 279
+    },
+    {
+      "epoch": 0.09350475872432794,
+      "grad_norm": 0.9983554927496489,
+      "learning_rate": 3.1145717463848723e-06,
+      "loss": 0.2694,
+      "step": 280
+    },
+    {
+      "epoch": 0.09383870429120053,
+      "grad_norm": 1.0555480006962337,
+      "learning_rate": 3.1256952169076754e-06,
+      "loss": 0.2583,
+      "step": 281
+    },
+    {
+      "epoch": 0.09417264985807314,
+      "grad_norm": 0.986411867922012,
+      "learning_rate": 3.1368186874304786e-06,
+      "loss": 0.2814,
+      "step": 282
+    },
+    {
+      "epoch": 0.09450659542494573,
+      "grad_norm": 1.0115618448923618,
+      "learning_rate": 3.1479421579532817e-06,
+      "loss": 0.2602,
+      "step": 283
+    },
+    {
+      "epoch": 0.09484054099181834,
+      "grad_norm": 1.074066553143631,
+      "learning_rate": 3.1590656284760844e-06,
+      "loss": 0.2577,
+      "step": 284
+    },
+    {
+      "epoch": 0.09517448655869093,
+      "grad_norm": 0.978641957387922,
+      "learning_rate": 3.170189098998888e-06,
+      "loss": 0.278,
+      "step": 285
+    },
+    {
+      "epoch": 0.09550843212556354,
+      "grad_norm": 0.9937876447676249,
+      "learning_rate": 3.181312569521691e-06,
+      "loss": 0.2532,
+      "step": 286
+    },
+    {
+      "epoch": 0.09584237769243613,
+      "grad_norm": 1.0298119478855567,
+      "learning_rate": 3.1924360400444942e-06,
+      "loss": 0.2669,
+      "step": 287
+    },
+    {
+      "epoch": 0.09617632325930874,
+      "grad_norm": 1.0380635876797846,
+      "learning_rate": 3.2035595105672973e-06,
+      "loss": 0.2849,
+      "step": 288
+    },
+    {
+      "epoch": 0.09651026882618133,
+      "grad_norm": 0.8705028041214296,
+      "learning_rate": 3.2146829810901005e-06,
+      "loss": 0.2631,
+      "step": 289
+    },
+    {
+      "epoch": 0.09684421439305393,
+      "grad_norm": 0.9554719138367959,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 0.2699,
+      "step": 290
+    },
+    {
+      "epoch": 0.09717815995992653,
+      "grad_norm": 1.1070871658729817,
+      "learning_rate": 3.2369299221357063e-06,
+      "loss": 0.2699,
+      "step": 291
+    },
+    {
+      "epoch": 0.09751210552679913,
+      "grad_norm": 1.1569397616168657,
+      "learning_rate": 3.2480533926585095e-06,
+      "loss": 0.2901,
+      "step": 292
+    },
+    {
+      "epoch": 0.09784605109367173,
+      "grad_norm": 0.8906942173665268,
+      "learning_rate": 3.259176863181313e-06,
+      "loss": 0.2491,
+      "step": 293
+    },
+    {
+      "epoch": 0.09817999666054433,
+      "grad_norm": 0.9616142547096681,
+      "learning_rate": 3.270300333704116e-06,
+      "loss": 0.2479,
+      "step": 294
+    },
+    {
+      "epoch": 0.09851394222741693,
+      "grad_norm": 0.9424174400332765,
+      "learning_rate": 3.2814238042269193e-06,
+      "loss": 0.2555,
+      "step": 295
+    },
+    {
+      "epoch": 0.09884788779428953,
+      "grad_norm": 0.969602671488551,
+      "learning_rate": 3.2925472747497224e-06,
+      "loss": 0.2522,
+      "step": 296
+    },
+    {
+      "epoch": 0.09918183336116213,
+      "grad_norm": 1.0778054535009527,
+      "learning_rate": 3.303670745272525e-06,
+      "loss": 0.2916,
+      "step": 297
+    },
+    {
+      "epoch": 0.09951577892803473,
+      "grad_norm": 0.950681407817474,
+      "learning_rate": 3.3147942157953282e-06,
+      "loss": 0.2624,
+      "step": 298
+    },
+    {
+      "epoch": 0.09984972449490732,
+      "grad_norm": 1.0136126498415492,
+      "learning_rate": 3.3259176863181314e-06,
+      "loss": 0.2933,
+      "step": 299
+    },
+    {
+      "epoch": 0.10018367006177993,
+      "grad_norm": 1.1157879759767526,
+      "learning_rate": 3.3370411568409345e-06,
+      "loss": 0.2752,
+      "step": 300
+    },
+    {
+      "epoch": 0.10051761562865252,
+      "grad_norm": 0.9659786911496452,
+      "learning_rate": 3.3481646273637376e-06,
+      "loss": 0.2637,
+      "step": 301
+    },
+    {
+      "epoch": 0.10085156119552513,
+      "grad_norm": 1.0334037162956897,
+      "learning_rate": 3.359288097886541e-06,
+      "loss": 0.2826,
+      "step": 302
+    },
+    {
+      "epoch": 0.10118550676239772,
+      "grad_norm": 0.8974929775545794,
+      "learning_rate": 3.3704115684093443e-06,
+      "loss": 0.2592,
+      "step": 303
+    },
+    {
+      "epoch": 0.10151945232927033,
+      "grad_norm": 0.9908654321167929,
+      "learning_rate": 3.381535038932147e-06,
+      "loss": 0.2514,
+      "step": 304
+    },
+    {
+      "epoch": 0.10185339789614292,
+      "grad_norm": 1.013861682980457,
+      "learning_rate": 3.39265850945495e-06,
+      "loss": 0.2659,
+      "step": 305
+    },
+    {
+      "epoch": 0.10218734346301553,
+      "grad_norm": 1.0310009709599692,
+      "learning_rate": 3.4037819799777533e-06,
+      "loss": 0.2936,
+      "step": 306
+    },
+    {
+      "epoch": 0.10252128902988812,
+      "grad_norm": 0.8989162049291942,
+      "learning_rate": 3.4149054505005564e-06,
+      "loss": 0.2524,
+      "step": 307
+    },
+    {
+      "epoch": 0.10285523459676073,
+      "grad_norm": 1.0325604088157159,
+      "learning_rate": 3.4260289210233596e-06,
+      "loss": 0.2725,
+      "step": 308
+    },
+    {
+      "epoch": 0.10318918016363333,
+      "grad_norm": 0.9980252685246547,
+      "learning_rate": 3.4371523915461623e-06,
+      "loss": 0.2663,
+      "step": 309
+    },
+    {
+      "epoch": 0.10352312573050593,
+      "grad_norm": 0.9993480947882514,
+      "learning_rate": 3.448275862068966e-06,
+      "loss": 0.2619,
+      "step": 310
+    },
+    {
+      "epoch": 0.10385707129737853,
+      "grad_norm": 1.2466860459144706,
+      "learning_rate": 3.459399332591769e-06,
+      "loss": 0.2864,
+      "step": 311
+    },
+    {
+      "epoch": 0.10419101686425113,
+      "grad_norm": 1.2924240507601645,
+      "learning_rate": 3.470522803114572e-06,
+      "loss": 0.2759,
+      "step": 312
+    },
+    {
+      "epoch": 0.10452496243112373,
+      "grad_norm": 1.117231355808104,
+      "learning_rate": 3.4816462736373752e-06,
+      "loss": 0.2493,
+      "step": 313
+    },
+    {
+      "epoch": 0.10485890799799633,
+      "grad_norm": 1.0013319983489994,
+      "learning_rate": 3.4927697441601784e-06,
+      "loss": 0.2694,
+      "step": 314
+    },
+    {
+      "epoch": 0.10519285356486893,
+      "grad_norm": 0.8939567136330437,
+      "learning_rate": 3.503893214682981e-06,
+      "loss": 0.2555,
+      "step": 315
+    },
+    {
+      "epoch": 0.10552679913174152,
+      "grad_norm": 1.239242556867951,
+      "learning_rate": 3.515016685205784e-06,
+      "loss": 0.2728,
+      "step": 316
+    },
+    {
+      "epoch": 0.10586074469861413,
+      "grad_norm": 1.0565860100298228,
+      "learning_rate": 3.5261401557285873e-06,
+      "loss": 0.2663,
+      "step": 317
+    },
+    {
+      "epoch": 0.10619469026548672,
+      "grad_norm": 0.9597333088021871,
+      "learning_rate": 3.537263626251391e-06,
+      "loss": 0.2729,
+      "step": 318
+    },
+    {
+      "epoch": 0.10652863583235933,
+      "grad_norm": 0.8695321802277968,
+      "learning_rate": 3.548387096774194e-06,
+      "loss": 0.2586,
+      "step": 319
+    },
+    {
+      "epoch": 0.10686258139923192,
+      "grad_norm": 1.035738979892239,
+      "learning_rate": 3.559510567296997e-06,
+      "loss": 0.2671,
+      "step": 320
+    },
+    {
+      "epoch": 0.10719652696610453,
+      "grad_norm": 0.9575178575212548,
+      "learning_rate": 3.5706340378198003e-06,
+      "loss": 0.2828,
+      "step": 321
+    },
+    {
+      "epoch": 0.10753047253297712,
+      "grad_norm": 1.1300397562818,
+      "learning_rate": 3.581757508342603e-06,
+      "loss": 0.2949,
+      "step": 322
+    },
+    {
+      "epoch": 0.10786441809984973,
+      "grad_norm": 1.1537677216616982,
+      "learning_rate": 3.592880978865406e-06,
+      "loss": 0.2656,
+      "step": 323
+    },
+    {
+      "epoch": 0.10819836366672232,
+      "grad_norm": 0.9627144838033714,
+      "learning_rate": 3.6040044493882093e-06,
+      "loss": 0.2785,
+      "step": 324
+    },
+    {
+      "epoch": 0.10853230923359493,
+      "grad_norm": 0.9449257156505781,
+      "learning_rate": 3.6151279199110124e-06,
+      "loss": 0.2676,
+      "step": 325
+    },
+    {
+      "epoch": 0.10886625480046752,
+      "grad_norm": 1.0206213863568905,
+      "learning_rate": 3.626251390433816e-06,
+      "loss": 0.2643,
+      "step": 326
+    },
+    {
+      "epoch": 0.10920020036734013,
+      "grad_norm": 0.9815487853303332,
+      "learning_rate": 3.637374860956619e-06,
+      "loss": 0.2789,
+      "step": 327
+    },
+    {
+      "epoch": 0.10953414593421272,
+      "grad_norm": 1.010306702915483,
+      "learning_rate": 3.648498331479422e-06,
+      "loss": 0.2683,
+      "step": 328
+    },
+    {
+      "epoch": 0.10986809150108533,
+      "grad_norm": 1.0690064824290435,
+      "learning_rate": 3.659621802002225e-06,
+      "loss": 0.2819,
+      "step": 329
+    },
+    {
+      "epoch": 0.11020203706795792,
+      "grad_norm": 1.0546653253133171,
+      "learning_rate": 3.670745272525028e-06,
+      "loss": 0.2869,
+      "step": 330
+    },
+    {
+      "epoch": 0.11053598263483053,
+      "grad_norm": 1.0596495202097613,
+      "learning_rate": 3.681868743047831e-06,
+      "loss": 0.2692,
+      "step": 331
+    },
+    {
+      "epoch": 0.11086992820170312,
+      "grad_norm": 0.9257332414938142,
+      "learning_rate": 3.6929922135706343e-06,
+      "loss": 0.2468,
+      "step": 332
+    },
+    {
+      "epoch": 0.11120387376857573,
+      "grad_norm": 0.9835419475940986,
+      "learning_rate": 3.7041156840934374e-06,
+      "loss": 0.2793,
+      "step": 333
+    },
+    {
+      "epoch": 0.11153781933544832,
+      "grad_norm": 0.9771967275391085,
+      "learning_rate": 3.71523915461624e-06,
+      "loss": 0.2845,
+      "step": 334
+    },
+    {
+      "epoch": 0.11187176490232092,
+      "grad_norm": 0.9602977118557694,
+      "learning_rate": 3.7263626251390437e-06,
+      "loss": 0.2742,
+      "step": 335
+    },
+    {
+      "epoch": 0.11220571046919352,
+      "grad_norm": 1.0029205944461674,
+      "learning_rate": 3.737486095661847e-06,
+      "loss": 0.2687,
+      "step": 336
+    },
+    {
+      "epoch": 0.11253965603606612,
+      "grad_norm": 1.0201328545542836,
+      "learning_rate": 3.74860956618465e-06,
+      "loss": 0.2708,
+      "step": 337
+    },
+    {
+      "epoch": 0.11287360160293872,
+      "grad_norm": 0.9103855657952339,
+      "learning_rate": 3.759733036707453e-06,
+      "loss": 0.2792,
+      "step": 338
+    },
+    {
+      "epoch": 0.11320754716981132,
+      "grad_norm": 0.9601017467552823,
+      "learning_rate": 3.7708565072302562e-06,
+      "loss": 0.2678,
+      "step": 339
+    },
+    {
+      "epoch": 0.11354149273668392,
+      "grad_norm": 1.0089053411102482,
+      "learning_rate": 3.781979977753059e-06,
+      "loss": 0.2743,
+      "step": 340
+    },
+    {
+      "epoch": 0.11387543830355652,
+      "grad_norm": 0.9504969314518689,
+      "learning_rate": 3.793103448275862e-06,
+      "loss": 0.2725,
+      "step": 341
+    },
+    {
+      "epoch": 0.11420938387042912,
+      "grad_norm": 0.9049400864683504,
+      "learning_rate": 3.804226918798665e-06,
+      "loss": 0.2545,
+      "step": 342
+    },
+    {
+      "epoch": 0.11454332943730172,
+      "grad_norm": 0.9801080200173922,
+      "learning_rate": 3.815350389321469e-06,
+      "loss": 0.2602,
+      "step": 343
+    },
+    {
+      "epoch": 0.11487727500417431,
+      "grad_norm": 0.9494051183285684,
+      "learning_rate": 3.8264738598442715e-06,
+      "loss": 0.2718,
+      "step": 344
+    },
+    {
+      "epoch": 0.11521122057104692,
+      "grad_norm": 0.9624224416135067,
+      "learning_rate": 3.837597330367075e-06,
+      "loss": 0.273,
+      "step": 345
+    },
+    {
+      "epoch": 0.11554516613791951,
+      "grad_norm": 1.2213226228441383,
+      "learning_rate": 3.848720800889878e-06,
+      "loss": 0.2825,
+      "step": 346
+    },
+    {
+      "epoch": 0.11587911170479212,
+      "grad_norm": 0.9776624862349191,
+      "learning_rate": 3.859844271412681e-06,
+      "loss": 0.277,
+      "step": 347
+    },
+    {
+      "epoch": 0.11621305727166471,
+      "grad_norm": 0.9947846421904093,
+      "learning_rate": 3.870967741935484e-06,
+      "loss": 0.2629,
+      "step": 348
+    },
+    {
+      "epoch": 0.11654700283853732,
+      "grad_norm": 0.9240560147186092,
+      "learning_rate": 3.8820912124582876e-06,
+      "loss": 0.2546,
+      "step": 349
+    },
+    {
+      "epoch": 0.11688094840540991,
+      "grad_norm": 0.963768770591079,
+      "learning_rate": 3.89321468298109e-06,
+      "loss": 0.2765,
+      "step": 350
+    },
+    {
+      "epoch": 0.11721489397228252,
+      "grad_norm": 0.9589949517884051,
+      "learning_rate": 3.904338153503894e-06,
+      "loss": 0.2775,
+      "step": 351
+    },
+    {
+      "epoch": 0.11754883953915511,
+      "grad_norm": 0.9782433798778546,
+      "learning_rate": 3.9154616240266965e-06,
+      "loss": 0.2799,
+      "step": 352
+    },
+    {
+      "epoch": 0.11788278510602772,
+      "grad_norm": 0.9858329287017443,
+      "learning_rate": 3.9265850945495e-06,
+      "loss": 0.2711,
+      "step": 353
+    },
+    {
+      "epoch": 0.11821673067290031,
+      "grad_norm": 0.9039100597255357,
+      "learning_rate": 3.937708565072303e-06,
+      "loss": 0.2581,
+      "step": 354
+    },
+    {
+      "epoch": 0.11855067623977292,
+      "grad_norm": 1.1034122762806373,
+      "learning_rate": 3.948832035595106e-06,
+      "loss": 0.2783,
+      "step": 355
+    },
+    {
+      "epoch": 0.11888462180664551,
+      "grad_norm": 0.960339590267913,
+      "learning_rate": 3.959955506117909e-06,
+      "loss": 0.2892,
+      "step": 356
+    },
+    {
+      "epoch": 0.11921856737351812,
+      "grad_norm": 1.0002994472569884,
+      "learning_rate": 3.971078976640712e-06,
+      "loss": 0.2866,
+      "step": 357
+    },
+    {
+      "epoch": 0.11955251294039071,
+      "grad_norm": 1.0057324471749156,
+      "learning_rate": 3.982202447163515e-06,
+      "loss": 0.2577,
+      "step": 358
+    },
+    {
+      "epoch": 0.11988645850726332,
+      "grad_norm": 0.9933056307505863,
+      "learning_rate": 3.993325917686319e-06,
+      "loss": 0.2671,
+      "step": 359
+    },
+    {
+      "epoch": 0.12022040407413592,
+      "grad_norm": 1.0118000022039888,
+      "learning_rate": 4.004449388209122e-06,
+      "loss": 0.2782,
+      "step": 360
+    },
+    {
+      "epoch": 0.12055434964100852,
+      "grad_norm": 0.9104306956415678,
+      "learning_rate": 4.015572858731925e-06,
+      "loss": 0.2719,
+      "step": 361
+    },
+    {
+      "epoch": 0.12088829520788112,
+      "grad_norm": 1.0409646407223883,
+      "learning_rate": 4.026696329254728e-06,
+      "loss": 0.2824,
+      "step": 362
+    },
+    {
+      "epoch": 0.12122224077475371,
+      "grad_norm": 0.9353332418061575,
+      "learning_rate": 4.0378197997775306e-06,
+      "loss": 0.2779,
+      "step": 363
+    },
+    {
+      "epoch": 0.12155618634162632,
+      "grad_norm": 0.9117928694164866,
+      "learning_rate": 4.048943270300334e-06,
+      "loss": 0.2645,
+      "step": 364
+    },
+    {
+      "epoch": 0.12189013190849891,
+      "grad_norm": 0.9516320491269368,
+      "learning_rate": 4.060066740823137e-06,
+      "loss": 0.2763,
+      "step": 365
+    },
+    {
+      "epoch": 0.12222407747537152,
+      "grad_norm": 0.8980496133272654,
+      "learning_rate": 4.07119021134594e-06,
+      "loss": 0.2479,
+      "step": 366
+    },
+    {
+      "epoch": 0.12255802304224411,
+      "grad_norm": 0.8722081882192233,
+      "learning_rate": 4.082313681868743e-06,
+      "loss": 0.254,
+      "step": 367
+    },
+    {
+      "epoch": 0.12289196860911672,
+      "grad_norm": 0.9489762799911274,
+      "learning_rate": 4.093437152391547e-06,
+      "loss": 0.2794,
+      "step": 368
+    },
+    {
+      "epoch": 0.12322591417598931,
+      "grad_norm": 0.8260848475511359,
+      "learning_rate": 4.104560622914349e-06,
+      "loss": 0.2581,
+      "step": 369
+    },
+    {
+      "epoch": 0.12355985974286192,
+      "grad_norm": 1.012870852293905,
+      "learning_rate": 4.115684093437153e-06,
+      "loss": 0.2795,
+      "step": 370
+    },
+    {
+      "epoch": 0.12389380530973451,
+      "grad_norm": 0.9078068672567016,
+      "learning_rate": 4.126807563959956e-06,
+      "loss": 0.2724,
+      "step": 371
+    },
+    {
+      "epoch": 0.12422775087660712,
+      "grad_norm": 0.9367356712295183,
+      "learning_rate": 4.137931034482759e-06,
+      "loss": 0.2772,
+      "step": 372
+    },
+    {
+      "epoch": 0.12456169644347971,
+      "grad_norm": 0.9769142088201448,
+      "learning_rate": 4.149054505005562e-06,
+      "loss": 0.273,
+      "step": 373
+    },
+    {
+      "epoch": 0.12489564201035232,
+      "grad_norm": 0.8714374593023121,
+      "learning_rate": 4.160177975528365e-06,
+      "loss": 0.2523,
+      "step": 374
+    },
+    {
+      "epoch": 0.1252295875772249,
+      "grad_norm": 0.9713759181008551,
+      "learning_rate": 4.171301446051168e-06,
+      "loss": 0.2627,
+      "step": 375
+    },
+    {
+      "epoch": 0.12556353314409752,
+      "grad_norm": 1.0024543615178043,
+      "learning_rate": 4.182424916573972e-06,
+      "loss": 0.2736,
+      "step": 376
+    },
+    {
+      "epoch": 0.12589747871097012,
+      "grad_norm": 0.8841635902459595,
+      "learning_rate": 4.193548387096774e-06,
+      "loss": 0.266,
+      "step": 377
+    },
+    {
+      "epoch": 0.1262314242778427,
+      "grad_norm": 0.9422878272730257,
+      "learning_rate": 4.204671857619578e-06,
+      "loss": 0.2763,
+      "step": 378
+    },
+    {
+      "epoch": 0.1265653698447153,
+      "grad_norm": 0.9737880805971011,
+      "learning_rate": 4.215795328142381e-06,
+      "loss": 0.2604,
+      "step": 379
+    },
+    {
+      "epoch": 0.12689931541158792,
+      "grad_norm": 0.9311503938608767,
+      "learning_rate": 4.226918798665184e-06,
+      "loss": 0.2687,
+      "step": 380
+    },
+    {
+      "epoch": 0.12723326097846052,
+      "grad_norm": 0.9099240809857687,
+      "learning_rate": 4.238042269187987e-06,
+      "loss": 0.2615,
+      "step": 381
+    },
+    {
+      "epoch": 0.1275672065453331,
+      "grad_norm": 0.8644535162182144,
+      "learning_rate": 4.24916573971079e-06,
+      "loss": 0.2629,
+      "step": 382
+    },
+    {
+      "epoch": 0.1279011521122057,
+      "grad_norm": 1.08354660489156,
+      "learning_rate": 4.260289210233593e-06,
+      "loss": 0.2757,
+      "step": 383
+    },
+    {
+      "epoch": 0.1282350976790783,
+      "grad_norm": 0.9059885118793621,
+      "learning_rate": 4.271412680756397e-06,
+      "loss": 0.2741,
+      "step": 384
+    },
+    {
+      "epoch": 0.12856904324595092,
+      "grad_norm": 0.8214511989438817,
+      "learning_rate": 4.2825361512791995e-06,
+      "loss": 0.2588,
+      "step": 385
+    },
+    {
+      "epoch": 0.1289029888128235,
+      "grad_norm": 0.8920712731302586,
+      "learning_rate": 4.293659621802003e-06,
+      "loss": 0.2628,
+      "step": 386
+    },
+    {
+      "epoch": 0.1292369343796961,
+      "grad_norm": 1.02140475986345,
+      "learning_rate": 4.304783092324806e-06,
+      "loss": 0.2769,
+      "step": 387
+    },
+    {
+      "epoch": 0.1295708799465687,
+      "grad_norm": 1.1041718598040544,
+      "learning_rate": 4.3159065628476084e-06,
+      "loss": 0.2829,
+      "step": 388
+    },
+    {
+      "epoch": 0.12990482551344132,
+      "grad_norm": 0.9103286130796249,
+      "learning_rate": 4.327030033370412e-06,
+      "loss": 0.2616,
+      "step": 389
+    },
+    {
+      "epoch": 0.1302387710803139,
+      "grad_norm": 0.8952232727330045,
+      "learning_rate": 4.338153503893215e-06,
+      "loss": 0.2484,
+      "step": 390
+    },
+    {
+      "epoch": 0.1305727166471865,
+      "grad_norm": 0.7877032707324497,
+      "learning_rate": 4.349276974416018e-06,
+      "loss": 0.2615,
+      "step": 391
+    },
+    {
+      "epoch": 0.1309066622140591,
+      "grad_norm": 0.881238803534704,
+      "learning_rate": 4.360400444938822e-06,
+      "loss": 0.2606,
+      "step": 392
+    },
+    {
+      "epoch": 0.13124060778093172,
+      "grad_norm": 0.8156813680994354,
+      "learning_rate": 4.3715239154616245e-06,
+      "loss": 0.2483,
+      "step": 393
+    },
+    {
+      "epoch": 0.1315745533478043,
+      "grad_norm": 0.8460744085644406,
+      "learning_rate": 4.382647385984427e-06,
+      "loss": 0.266,
+      "step": 394
+    },
+    {
+      "epoch": 0.1319084989146769,
+      "grad_norm": 0.8909110880065707,
+      "learning_rate": 4.393770856507231e-06,
+      "loss": 0.2507,
+      "step": 395
+    },
+    {
+      "epoch": 0.1322424444815495,
+      "grad_norm": 1.0610350746398265,
+      "learning_rate": 4.4048943270300335e-06,
+      "loss": 0.2962,
+      "step": 396
+    },
+    {
+      "epoch": 0.13257639004842212,
+      "grad_norm": 1.2159344229127298,
+      "learning_rate": 4.416017797552837e-06,
+      "loss": 0.2628,
+      "step": 397
+    },
+    {
+      "epoch": 0.1329103356152947,
+      "grad_norm": 0.9343938441218402,
+      "learning_rate": 4.42714126807564e-06,
+      "loss": 0.2809,
+      "step": 398
+    },
+    {
+      "epoch": 0.1332442811821673,
+      "grad_norm": 0.9460286809085141,
+      "learning_rate": 4.4382647385984425e-06,
+      "loss": 0.2672,
+      "step": 399
+    },
+    {
+      "epoch": 0.1335782267490399,
+      "grad_norm": 0.9028504129341358,
+      "learning_rate": 4.449388209121246e-06,
+      "loss": 0.253,
+      "step": 400
+    },
+    {
+      "epoch": 0.13391217231591251,
+      "grad_norm": 0.777636732453647,
+      "learning_rate": 4.4605116796440496e-06,
+      "loss": 0.252,
+      "step": 401
+    },
+    {
+      "epoch": 0.1342461178827851,
+      "grad_norm": 0.8787403586719423,
+      "learning_rate": 4.471635150166852e-06,
+      "loss": 0.2681,
+      "step": 402
+    },
+    {
+      "epoch": 0.1345800634496577,
+      "grad_norm": 0.8623473953489393,
+      "learning_rate": 4.482758620689656e-06,
+      "loss": 0.2615,
+      "step": 403
+    },
+    {
+      "epoch": 0.1349140090165303,
+      "grad_norm": 0.8590329595872587,
+      "learning_rate": 4.4938820912124585e-06,
+      "loss": 0.2693,
+      "step": 404
+    },
+    {
+      "epoch": 0.1352479545834029,
+      "grad_norm": 1.0942605325821617,
+      "learning_rate": 4.505005561735262e-06,
+      "loss": 0.276,
+      "step": 405
+    },
+    {
+      "epoch": 0.1355819001502755,
+      "grad_norm": 0.8384871494712324,
+      "learning_rate": 4.516129032258065e-06,
+      "loss": 0.2539,
+      "step": 406
+    },
+    {
+      "epoch": 0.1359158457171481,
+      "grad_norm": 0.9713974468286866,
+      "learning_rate": 4.5272525027808675e-06,
+      "loss": 0.248,
+      "step": 407
+    },
+    {
+      "epoch": 0.1362497912840207,
+      "grad_norm": 0.8912405350992786,
+      "learning_rate": 4.538375973303671e-06,
+      "loss": 0.2606,
+      "step": 408
+    },
+    {
+      "epoch": 0.1365837368508933,
+      "grad_norm": 0.8504112799682982,
+      "learning_rate": 4.549499443826475e-06,
+      "loss": 0.2478,
+      "step": 409
+    },
+    {
+      "epoch": 0.13691768241776592,
+      "grad_norm": 0.8845213824948364,
+      "learning_rate": 4.560622914349277e-06,
+      "loss": 0.2549,
+      "step": 410
+    },
+    {
+      "epoch": 0.1372516279846385,
+      "grad_norm": 0.9170303075568879,
+      "learning_rate": 4.571746384872081e-06,
+      "loss": 0.2824,
+      "step": 411
+    },
+    {
+      "epoch": 0.1375855735515111,
+      "grad_norm": 0.9418395539373073,
+      "learning_rate": 4.582869855394884e-06,
+      "loss": 0.2512,
+      "step": 412
+    },
+    {
+      "epoch": 0.1379195191183837,
+      "grad_norm": 0.961468515239987,
+      "learning_rate": 4.593993325917686e-06,
+      "loss": 0.2798,
+      "step": 413
+    },
+    {
+      "epoch": 0.13825346468525632,
+      "grad_norm": 0.884383561019613,
+      "learning_rate": 4.60511679644049e-06,
+      "loss": 0.2693,
+      "step": 414
+    },
+    {
+      "epoch": 0.1385874102521289,
+      "grad_norm": 0.8938825429085708,
+      "learning_rate": 4.6162402669632926e-06,
+      "loss": 0.2686,
+      "step": 415
+    },
+    {
+      "epoch": 0.1389213558190015,
+      "grad_norm": 0.8354946158818547,
+      "learning_rate": 4.627363737486096e-06,
+      "loss": 0.245,
+      "step": 416
+    },
+    {
+      "epoch": 0.1392553013858741,
+      "grad_norm": 0.9373626614204349,
+      "learning_rate": 4.6384872080089e-06,
+      "loss": 0.2802,
+      "step": 417
+    },
+    {
+      "epoch": 0.13958924695274672,
+      "grad_norm": 0.7859209899855684,
+      "learning_rate": 4.649610678531702e-06,
+      "loss": 0.2445,
+      "step": 418
+    },
+    {
+      "epoch": 0.1399231925196193,
+      "grad_norm": 0.9807671068134438,
+      "learning_rate": 4.660734149054505e-06,
+      "loss": 0.2644,
+      "step": 419
+    },
+    {
+      "epoch": 0.1402571380864919,
+      "grad_norm": 0.851020616239925,
+      "learning_rate": 4.671857619577309e-06,
+      "loss": 0.2551,
+      "step": 420
+    },
+    {
+      "epoch": 0.1405910836533645,
+      "grad_norm": 0.9306635039098415,
+      "learning_rate": 4.682981090100111e-06,
+      "loss": 0.2559,
+      "step": 421
+    },
+    {
+      "epoch": 0.1409250292202371,
+      "grad_norm": 0.9028478012185194,
+      "learning_rate": 4.694104560622915e-06,
+      "loss": 0.2797,
+      "step": 422
+    },
+    {
+      "epoch": 0.1412589747871097,
+      "grad_norm": 0.8282408886330277,
+      "learning_rate": 4.705228031145718e-06,
+      "loss": 0.2634,
+      "step": 423
+    },
+    {
+      "epoch": 0.1415929203539823,
+      "grad_norm": 0.9583571910070512,
+      "learning_rate": 4.71635150166852e-06,
+      "loss": 0.2737,
+      "step": 424
+    },
+    {
+      "epoch": 0.1419268659208549,
+      "grad_norm": 0.8598845314007896,
+      "learning_rate": 4.727474972191325e-06,
+      "loss": 0.2734,
+      "step": 425
+    },
+    {
+      "epoch": 0.1422608114877275,
+      "grad_norm": 0.9823961957996282,
+      "learning_rate": 4.7385984427141274e-06,
+      "loss": 0.2756,
+      "step": 426
+    },
+    {
+      "epoch": 0.1425947570546001,
+      "grad_norm": 0.8402666268767217,
+      "learning_rate": 4.74972191323693e-06,
+      "loss": 0.2644,
+      "step": 427
+    },
+    {
+      "epoch": 0.1429287026214727,
+      "grad_norm": 0.9504374640973223,
+      "learning_rate": 4.760845383759734e-06,
+      "loss": 0.2802,
+      "step": 428
+    },
+    {
+      "epoch": 0.1432626481883453,
+      "grad_norm": 0.860942841549608,
+      "learning_rate": 4.771968854282536e-06,
+      "loss": 0.2642,
+      "step": 429
+    },
+    {
+      "epoch": 0.1435965937552179,
+      "grad_norm": 0.7976260693507284,
+      "learning_rate": 4.78309232480534e-06,
+      "loss": 0.2396,
+      "step": 430
+    },
+    {
+      "epoch": 0.1439305393220905,
+      "grad_norm": 0.9444299816616959,
+      "learning_rate": 4.794215795328143e-06,
+      "loss": 0.2647,
+      "step": 431
+    },
+    {
+      "epoch": 0.1442644848889631,
+      "grad_norm": 0.9810683931758797,
+      "learning_rate": 4.805339265850945e-06,
+      "loss": 0.2755,
+      "step": 432
+    },
+    {
+      "epoch": 0.1445984304558357,
+      "grad_norm": 0.8774017600524443,
+      "learning_rate": 4.816462736373749e-06,
+      "loss": 0.2516,
+      "step": 433
+    },
+    {
+      "epoch": 0.1449323760227083,
+      "grad_norm": 0.9482573960208934,
+      "learning_rate": 4.8275862068965525e-06,
+      "loss": 0.2862,
+      "step": 434
+    },
+    {
+      "epoch": 0.1452663215895809,
+      "grad_norm": 0.8175895495379112,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 0.2573,
+      "step": 435
+    },
+    {
+      "epoch": 0.1456002671564535,
+      "grad_norm": 0.9000739545892836,
+      "learning_rate": 4.849833147942159e-06,
+      "loss": 0.2596,
+      "step": 436
+    },
+    {
+      "epoch": 0.1459342127233261,
+      "grad_norm": 0.9000554099286614,
+      "learning_rate": 4.8609566184649615e-06,
+      "loss": 0.2525,
+      "step": 437
+    },
+    {
+      "epoch": 0.1462681582901987,
+      "grad_norm": 1.035965080184084,
+      "learning_rate": 4.872080088987764e-06,
+      "loss": 0.2746,
+      "step": 438
+    },
+    {
+      "epoch": 0.1466021038570713,
+      "grad_norm": 0.9495692424587069,
+      "learning_rate": 4.883203559510568e-06,
+      "loss": 0.2765,
+      "step": 439
+    },
+    {
+      "epoch": 0.1469360494239439,
+      "grad_norm": 0.8468127334478587,
+      "learning_rate": 4.8943270300333704e-06,
+      "loss": 0.2499,
+      "step": 440
+    },
+    {
+      "epoch": 0.1472699949908165,
+      "grad_norm": 0.9430177643905899,
+      "learning_rate": 4.905450500556174e-06,
+      "loss": 0.2596,
+      "step": 441
+    },
+    {
+      "epoch": 0.1476039405576891,
+      "grad_norm": 0.8564111776747324,
+      "learning_rate": 4.9165739710789776e-06,
+      "loss": 0.2742,
+      "step": 442
+    },
+    {
+      "epoch": 0.14793788612456168,
+      "grad_norm": 0.9034833859232495,
+      "learning_rate": 4.92769744160178e-06,
+      "loss": 0.2624,
+      "step": 443
+    },
+    {
+      "epoch": 0.1482718316914343,
+      "grad_norm": 1.0085778060835962,
+      "learning_rate": 4.938820912124583e-06,
+      "loss": 0.2788,
+      "step": 444
+    },
+    {
+      "epoch": 0.1486057772583069,
+      "grad_norm": 0.8486548720543253,
+      "learning_rate": 4.9499443826473865e-06,
+      "loss": 0.2499,
+      "step": 445
+    },
+    {
+      "epoch": 0.1489397228251795,
+      "grad_norm": 0.8647171380949372,
+      "learning_rate": 4.961067853170189e-06,
+      "loss": 0.2467,
+      "step": 446
+    },
+    {
+      "epoch": 0.14927366839205208,
+      "grad_norm": 0.9458194360965803,
+      "learning_rate": 4.972191323692993e-06,
+      "loss": 0.26,
+      "step": 447
+    },
+    {
+      "epoch": 0.1496076139589247,
+      "grad_norm": 0.98845534464712,
+      "learning_rate": 4.9833147942157955e-06,
+      "loss": 0.277,
+      "step": 448
+    },
+    {
+      "epoch": 0.1499415595257973,
+      "grad_norm": 1.0038333784339768,
+      "learning_rate": 4.994438264738598e-06,
+      "loss": 0.2607,
+      "step": 449
+    },
+    {
+      "epoch": 0.1502755050926699,
+      "grad_norm": 0.9882288469242262,
+      "learning_rate": 5.005561735261402e-06,
+      "loss": 0.2736,
+      "step": 450
+    },
+    {
+      "epoch": 0.15060945065954248,
+      "grad_norm": 0.9332132918013903,
+      "learning_rate": 5.016685205784205e-06,
+      "loss": 0.2746,
+      "step": 451
+    },
+    {
+      "epoch": 0.1509433962264151,
+      "grad_norm": 0.8514088935668446,
+      "learning_rate": 5.027808676307008e-06,
+      "loss": 0.2667,
+      "step": 452
+    },
+    {
+      "epoch": 0.1512773417932877,
+      "grad_norm": 0.8462811908982392,
+      "learning_rate": 5.038932146829812e-06,
+      "loss": 0.2675,
+      "step": 453
+    },
+    {
+      "epoch": 0.1516112873601603,
+      "grad_norm": 0.8691114025176568,
+      "learning_rate": 5.050055617352615e-06,
+      "loss": 0.289,
+      "step": 454
+    },
+    {
+      "epoch": 0.15194523292703288,
+      "grad_norm": 0.969925154555584,
+      "learning_rate": 5.061179087875418e-06,
+      "loss": 0.2658,
+      "step": 455
+    },
+    {
+      "epoch": 0.1522791784939055,
+      "grad_norm": 0.7612750857319712,
+      "learning_rate": 5.072302558398221e-06,
+      "loss": 0.246,
+      "step": 456
+    },
+    {
+      "epoch": 0.1526131240607781,
+      "grad_norm": 0.9012049865349984,
+      "learning_rate": 5.083426028921023e-06,
+      "loss": 0.2791,
+      "step": 457
+    },
+    {
+      "epoch": 0.1529470696276507,
+      "grad_norm": 0.8335869019920015,
+      "learning_rate": 5.094549499443827e-06,
+      "loss": 0.2542,
+      "step": 458
+    },
+    {
+      "epoch": 0.15328101519452328,
+      "grad_norm": 0.8982538630222419,
+      "learning_rate": 5.1056729699666295e-06,
+      "loss": 0.2693,
+      "step": 459
+    },
+    {
+      "epoch": 0.15361496076139589,
+      "grad_norm": 0.8430594873139436,
+      "learning_rate": 5.116796440489433e-06,
+      "loss": 0.2533,
+      "step": 460
+    },
+    {
+      "epoch": 0.1539489063282685,
+      "grad_norm": 0.8653242092400866,
+      "learning_rate": 5.127919911012236e-06,
+      "loss": 0.264,
+      "step": 461
+    },
+    {
+      "epoch": 0.1542828518951411,
+      "grad_norm": 0.8994436945219121,
+      "learning_rate": 5.139043381535039e-06,
+      "loss": 0.2615,
+      "step": 462
+    },
+    {
+      "epoch": 0.1546167974620137,
+      "grad_norm": 0.8101799725720696,
+      "learning_rate": 5.150166852057843e-06,
+      "loss": 0.2532,
+      "step": 463
+    },
+    {
+      "epoch": 0.15495074302888628,
+      "grad_norm": 0.8855034068759076,
+      "learning_rate": 5.161290322580646e-06,
+      "loss": 0.2732,
+      "step": 464
+    },
+    {
+      "epoch": 0.1552846885957589,
+      "grad_norm": 0.8043484789939199,
+      "learning_rate": 5.172413793103449e-06,
+      "loss": 0.2379,
+      "step": 465
+    },
+    {
+      "epoch": 0.1556186341626315,
+      "grad_norm": 0.8962988729164553,
+      "learning_rate": 5.183537263626252e-06,
+      "loss": 0.2799,
+      "step": 466
+    },
+    {
+      "epoch": 0.1559525797295041,
+      "grad_norm": 0.8232384777691226,
+      "learning_rate": 5.1946607341490554e-06,
+      "loss": 0.2586,
+      "step": 467
+    },
+    {
+      "epoch": 0.15628652529637668,
+      "grad_norm": 0.8472967462900588,
+      "learning_rate": 5.205784204671857e-06,
+      "loss": 0.2591,
+      "step": 468
+    },
+    {
+      "epoch": 0.1566204708632493,
+      "grad_norm": 0.7977382518203232,
+      "learning_rate": 5.216907675194661e-06,
+      "loss": 0.2517,
+      "step": 469
+    },
+    {
+      "epoch": 0.1569544164301219,
+      "grad_norm": 0.9780305545037385,
+      "learning_rate": 5.2280311457174636e-06,
+      "loss": 0.2772,
+      "step": 470
+    },
+    {
+      "epoch": 0.1572883619969945,
+      "grad_norm": 0.7622852005224084,
+      "learning_rate": 5.239154616240267e-06,
+      "loss": 0.2538,
+      "step": 471
+    },
+    {
+      "epoch": 0.15762230756386708,
+      "grad_norm": 0.8593964452116354,
+      "learning_rate": 5.250278086763071e-06,
+      "loss": 0.2568,
+      "step": 472
+    },
+    {
+      "epoch": 0.1579562531307397,
+      "grad_norm": 0.8100335857942087,
+      "learning_rate": 5.261401557285873e-06,
+      "loss": 0.2501,
+      "step": 473
+    },
+    {
+      "epoch": 0.1582901986976123,
+      "grad_norm": 0.8014373233797282,
+      "learning_rate": 5.272525027808677e-06,
+      "loss": 0.2535,
+      "step": 474
+    },
+    {
+      "epoch": 0.1586241442644849,
+      "grad_norm": 0.7975499651795742,
+      "learning_rate": 5.28364849833148e-06,
+      "loss": 0.2504,
+      "step": 475
+    },
+    {
+      "epoch": 0.15895808983135748,
+      "grad_norm": 0.8394801448362739,
+      "learning_rate": 5.294771968854283e-06,
+      "loss": 0.2707,
+      "step": 476
+    },
+    {
+      "epoch": 0.1592920353982301,
+      "grad_norm": 0.9828129623333244,
+      "learning_rate": 5.305895439377086e-06,
+      "loss": 0.2775,
+      "step": 477
+    },
+    {
+      "epoch": 0.1596259809651027,
+      "grad_norm": 0.732105226843958,
+      "learning_rate": 5.3170189098998895e-06,
+      "loss": 0.2289,
+      "step": 478
+    },
+    {
+      "epoch": 0.1599599265319753,
+      "grad_norm": 0.8965879378663164,
+      "learning_rate": 5.328142380422693e-06,
+      "loss": 0.2401,
+      "step": 479
+    },
+    {
+      "epoch": 0.16029387209884788,
+      "grad_norm": 0.9057075540665339,
+      "learning_rate": 5.339265850945496e-06,
+      "loss": 0.2681,
+      "step": 480
+    },
+    {
+      "epoch": 0.16062781766572048,
+      "grad_norm": 0.7990495639152133,
+      "learning_rate": 5.350389321468299e-06,
+      "loss": 0.2485,
+      "step": 481
+    },
+    {
+      "epoch": 0.1609617632325931,
+      "grad_norm": 1.015072564398485,
+      "learning_rate": 5.361512791991101e-06,
+      "loss": 0.2662,
+      "step": 482
+    },
+    {
+      "epoch": 0.1612957087994657,
+      "grad_norm": 0.8030254740021361,
+      "learning_rate": 5.372636262513905e-06,
+      "loss": 0.2585,
+      "step": 483
+    },
+    {
+      "epoch": 0.16162965436633828,
+      "grad_norm": 1.0859820471858048,
+      "learning_rate": 5.383759733036707e-06,
+      "loss": 0.2715,
+      "step": 484
+    },
+    {
+      "epoch": 0.16196359993321088,
+      "grad_norm": 0.8205452365480986,
+      "learning_rate": 5.394883203559511e-06,
+      "loss": 0.2596,
+      "step": 485
+    },
+    {
+      "epoch": 0.1622975455000835,
+      "grad_norm": 0.875783049907706,
+      "learning_rate": 5.406006674082314e-06,
+      "loss": 0.2605,
+      "step": 486
+    },
+    {
+      "epoch": 0.1626314910669561,
+      "grad_norm": 0.8802931215752893,
+      "learning_rate": 5.417130144605117e-06,
+      "loss": 0.2599,
+      "step": 487
+    },
+    {
+      "epoch": 0.16296543663382868,
+      "grad_norm": 0.8179528891483596,
+      "learning_rate": 5.428253615127921e-06,
+      "loss": 0.2527,
+      "step": 488
+    },
+    {
+      "epoch": 0.16329938220070128,
+      "grad_norm": 0.734992877721891,
+      "learning_rate": 5.4393770856507235e-06,
+      "loss": 0.241,
+      "step": 489
+    },
+    {
+      "epoch": 0.1636333277675739,
+      "grad_norm": 0.8158038849222073,
+      "learning_rate": 5.450500556173527e-06,
+      "loss": 0.2506,
+      "step": 490
+    },
+    {
+      "epoch": 0.1639672733344465,
+      "grad_norm": 0.835866048908736,
+      "learning_rate": 5.46162402669633e-06,
+      "loss": 0.2568,
+      "step": 491
+    },
+    {
+      "epoch": 0.16430121890131907,
+      "grad_norm": 0.895021641794215,
+      "learning_rate": 5.472747497219133e-06,
+      "loss": 0.2632,
+      "step": 492
+    },
+    {
+      "epoch": 0.16463516446819168,
+      "grad_norm": 0.8090105148370232,
+      "learning_rate": 5.483870967741935e-06,
+      "loss": 0.2473,
+      "step": 493
+    },
+    {
+      "epoch": 0.1649691100350643,
+      "grad_norm": 0.810000176014472,
+      "learning_rate": 5.494994438264739e-06,
+      "loss": 0.242,
+      "step": 494
+    },
+    {
+      "epoch": 0.1653030556019369,
+      "grad_norm": 0.8836235099488476,
+      "learning_rate": 5.506117908787543e-06,
+      "loss": 0.2455,
+      "step": 495
+    },
+    {
+      "epoch": 0.16563700116880947,
+      "grad_norm": 0.8254870164296962,
+      "learning_rate": 5.517241379310345e-06,
+      "loss": 0.2539,
+      "step": 496
+    },
+    {
+      "epoch": 0.16597094673568208,
+      "grad_norm": 0.9525772225779567,
+      "learning_rate": 5.5283648498331485e-06,
+      "loss": 0.261,
+      "step": 497
+    },
+    {
+      "epoch": 0.16630489230255469,
+      "grad_norm": 0.8377853082667718,
+      "learning_rate": 5.539488320355951e-06,
+      "loss": 0.253,
+      "step": 498
+    },
+    {
+      "epoch": 0.1666388378694273,
+      "grad_norm": 0.9616580429013377,
+      "learning_rate": 5.550611790878755e-06,
+      "loss": 0.2587,
+      "step": 499
+    },
+    {
+      "epoch": 0.16697278343629987,
+      "grad_norm": 0.7796060710868616,
+      "learning_rate": 5.5617352614015575e-06,
+      "loss": 0.2588,
+      "step": 500
+    },
+    {
+      "epoch": 0.16730672900317248,
+      "grad_norm": 0.9319120249495298,
+      "learning_rate": 5.572858731924361e-06,
+      "loss": 0.2451,
+      "step": 501
+    },
+    {
+      "epoch": 0.16764067457004508,
+      "grad_norm": 1.0596544782036998,
+      "learning_rate": 5.583982202447164e-06,
+      "loss": 0.2749,
+      "step": 502
+    },
+    {
+      "epoch": 0.1679746201369177,
+      "grad_norm": 0.8863515672067872,
+      "learning_rate": 5.595105672969967e-06,
+      "loss": 0.2511,
+      "step": 503
+    },
+    {
+      "epoch": 0.16830856570379027,
+      "grad_norm": 0.8225742978610687,
+      "learning_rate": 5.606229143492771e-06,
+      "loss": 0.2377,
+      "step": 504
+    },
+    {
+      "epoch": 0.16864251127066288,
+      "grad_norm": 0.7637996051027277,
+      "learning_rate": 5.617352614015574e-06,
+      "loss": 0.2335,
+      "step": 505
+    },
+    {
+      "epoch": 0.16897645683753548,
+      "grad_norm": 0.8249967207095058,
+      "learning_rate": 5.628476084538377e-06,
+      "loss": 0.2748,
+      "step": 506
+    },
+    {
+      "epoch": 0.1693104024044081,
+      "grad_norm": 0.8161862924539601,
+      "learning_rate": 5.639599555061179e-06,
+      "loss": 0.2649,
+      "step": 507
+    },
+    {
+      "epoch": 0.16964434797128067,
+      "grad_norm": 0.8749861083035826,
+      "learning_rate": 5.6507230255839826e-06,
+      "loss": 0.2641,
+      "step": 508
+    },
+    {
+      "epoch": 0.16997829353815327,
+      "grad_norm": 0.8934466811434186,
+      "learning_rate": 5.661846496106785e-06,
+      "loss": 0.2679,
+      "step": 509
+    },
+    {
+      "epoch": 0.17031223910502588,
+      "grad_norm": 0.8787600615518483,
+      "learning_rate": 5.672969966629589e-06,
+      "loss": 0.2735,
+      "step": 510
+    },
+    {
+      "epoch": 0.1706461846718985,
+      "grad_norm": 0.7797278565537913,
+      "learning_rate": 5.6840934371523915e-06,
+      "loss": 0.2652,
+      "step": 511
+    },
+    {
+      "epoch": 0.1709801302387711,
+      "grad_norm": 0.7628477421169882,
+      "learning_rate": 5.695216907675195e-06,
+      "loss": 0.2424,
+      "step": 512
+    },
+    {
+      "epoch": 0.17131407580564367,
+      "grad_norm": 0.840464455052028,
+      "learning_rate": 5.706340378197999e-06,
+      "loss": 0.2659,
+      "step": 513
+    },
+    {
+      "epoch": 0.17164802137251628,
+      "grad_norm": 0.985780992202247,
+      "learning_rate": 5.717463848720801e-06,
+      "loss": 0.2507,
+      "step": 514
+    },
+    {
+      "epoch": 0.1719819669393889,
+      "grad_norm": 1.4268260954113605,
+      "learning_rate": 5.728587319243605e-06,
+      "loss": 0.2718,
+      "step": 515
+    },
+    {
+      "epoch": 0.1723159125062615,
+      "grad_norm": 0.7437022173759871,
+      "learning_rate": 5.739710789766408e-06,
+      "loss": 0.2543,
+      "step": 516
+    },
+    {
+      "epoch": 0.17264985807313407,
+      "grad_norm": 0.7373193238546386,
+      "learning_rate": 5.750834260289211e-06,
+      "loss": 0.2451,
+      "step": 517
+    },
+    {
+      "epoch": 0.17298380364000668,
+      "grad_norm": 0.9049436389723264,
+      "learning_rate": 5.761957730812013e-06,
+      "loss": 0.2634,
+      "step": 518
+    },
+    {
+      "epoch": 0.17331774920687928,
+      "grad_norm": 0.8105545282139595,
+      "learning_rate": 5.773081201334817e-06,
+      "loss": 0.2588,
+      "step": 519
+    },
+    {
+      "epoch": 0.1736516947737519,
+      "grad_norm": 0.7844885492997241,
+      "learning_rate": 5.784204671857621e-06,
+      "loss": 0.2507,
+      "step": 520
+    },
+    {
+      "epoch": 0.17398564034062447,
+      "grad_norm": 1.5001443038134146,
+      "learning_rate": 5.795328142380423e-06,
+      "loss": 0.258,
+      "step": 521
+    },
+    {
+      "epoch": 0.17431958590749708,
+      "grad_norm": 0.9479471274537018,
+      "learning_rate": 5.806451612903226e-06,
+      "loss": 0.2547,
+      "step": 522
+    },
+    {
+      "epoch": 0.17465353147436968,
+      "grad_norm": 0.8165944158582878,
+      "learning_rate": 5.817575083426029e-06,
+      "loss": 0.2618,
+      "step": 523
+    },
+    {
+      "epoch": 0.1749874770412423,
+      "grad_norm": 0.793372998639345,
+      "learning_rate": 5.828698553948833e-06,
+      "loss": 0.2538,
+      "step": 524
+    },
+    {
+      "epoch": 0.17532142260811487,
+      "grad_norm": 0.7708291500668256,
+      "learning_rate": 5.839822024471635e-06,
+      "loss": 0.2561,
+      "step": 525
+    },
+    {
+      "epoch": 0.17565536817498748,
+      "grad_norm": 0.7448808234488508,
+      "learning_rate": 5.850945494994439e-06,
+      "loss": 0.2502,
+      "step": 526
+    },
+    {
+      "epoch": 0.17598931374186008,
+      "grad_norm": 0.8021357502415758,
+      "learning_rate": 5.862068965517242e-06,
+      "loss": 0.2783,
+      "step": 527
+    },
+    {
+      "epoch": 0.1763232593087327,
+      "grad_norm": 0.7495063590638561,
+      "learning_rate": 5.873192436040045e-06,
+      "loss": 0.2476,
+      "step": 528
+    },
+    {
+      "epoch": 0.17665720487560527,
+      "grad_norm": 0.7961292168845412,
+      "learning_rate": 5.884315906562849e-06,
+      "loss": 0.2726,
+      "step": 529
+    },
+    {
+      "epoch": 0.17699115044247787,
+      "grad_norm": 0.8074466377944476,
+      "learning_rate": 5.8954393770856515e-06,
+      "loss": 0.2582,
+      "step": 530
+    },
+    {
+      "epoch": 0.17732509600935048,
+      "grad_norm": 0.8345256528144968,
+      "learning_rate": 5.906562847608455e-06,
+      "loss": 0.245,
+      "step": 531
+    },
+    {
+      "epoch": 0.1776590415762231,
+      "grad_norm": 0.7869923386555632,
+      "learning_rate": 5.917686318131257e-06,
+      "loss": 0.2569,
+      "step": 532
+    },
+    {
+      "epoch": 0.17799298714309567,
+      "grad_norm": 0.8147246527010965,
+      "learning_rate": 5.9288097886540604e-06,
+      "loss": 0.2362,
+      "step": 533
+    },
+    {
+      "epoch": 0.17832693270996827,
+      "grad_norm": 0.8341283135652426,
+      "learning_rate": 5.939933259176863e-06,
+      "loss": 0.2764,
+      "step": 534
+    },
+    {
+      "epoch": 0.17866087827684088,
+      "grad_norm": 0.7045070493955722,
+      "learning_rate": 5.951056729699667e-06,
+      "loss": 0.2354,
+      "step": 535
+    },
+    {
+      "epoch": 0.17899482384371349,
+      "grad_norm": 0.7421340011554978,
+      "learning_rate": 5.962180200222469e-06,
+      "loss": 0.2386,
+      "step": 536
+    },
+    {
+      "epoch": 0.17932876941058606,
+      "grad_norm": 0.8226524022867789,
+      "learning_rate": 5.973303670745273e-06,
+      "loss": 0.2594,
+      "step": 537
+    },
+    {
+      "epoch": 0.17966271497745867,
+      "grad_norm": 0.9397477018343224,
+      "learning_rate": 5.9844271412680765e-06,
+      "loss": 0.2639,
+      "step": 538
+    },
+    {
+      "epoch": 0.17999666054433128,
+      "grad_norm": 0.9390525099666275,
+      "learning_rate": 5.995550611790879e-06,
+      "loss": 0.2477,
+      "step": 539
+    },
+    {
+      "epoch": 0.18033060611120388,
+      "grad_norm": 0.8300240065923898,
+      "learning_rate": 6.006674082313683e-06,
+      "loss": 0.2751,
+      "step": 540
+    },
+    {
+      "epoch": 0.18066455167807646,
+      "grad_norm": 0.813819268112457,
+      "learning_rate": 6.0177975528364855e-06,
+      "loss": 0.2617,
+      "step": 541
+    },
+    {
+      "epoch": 0.18099849724494907,
+      "grad_norm": 0.9523295641887335,
+      "learning_rate": 6.028921023359289e-06,
+      "loss": 0.2642,
+      "step": 542
+    },
+    {
+      "epoch": 0.18133244281182168,
+      "grad_norm": 0.8610172814649126,
+      "learning_rate": 6.040044493882091e-06,
+      "loss": 0.2738,
+      "step": 543
+    },
+    {
+      "epoch": 0.18166638837869428,
+      "grad_norm": 0.8247418133291398,
+      "learning_rate": 6.0511679644048945e-06,
+      "loss": 0.2794,
+      "step": 544
+    },
+    {
+      "epoch": 0.18200033394556686,
+      "grad_norm": 0.8563775168824261,
+      "learning_rate": 6.062291434927698e-06,
+      "loss": 0.255,
+      "step": 545
+    },
+    {
+      "epoch": 0.18233427951243947,
+      "grad_norm": 0.8968217199385246,
+      "learning_rate": 6.073414905450501e-06,
+      "loss": 0.2861,
+      "step": 546
+    },
+    {
+      "epoch": 0.18266822507931207,
+      "grad_norm": 0.8235810669629106,
+      "learning_rate": 6.084538375973304e-06,
+      "loss": 0.2663,
+      "step": 547
+    },
+    {
+      "epoch": 0.18300217064618468,
+      "grad_norm": 0.8348617646793106,
+      "learning_rate": 6.095661846496107e-06,
+      "loss": 0.2486,
+      "step": 548
+    },
+    {
+      "epoch": 0.18333611621305726,
+      "grad_norm": 0.8185061661036209,
+      "learning_rate": 6.1067853170189106e-06,
+      "loss": 0.2885,
+      "step": 549
+    },
+    {
+      "epoch": 0.18367006177992987,
+      "grad_norm": 0.8056111861107552,
+      "learning_rate": 6.117908787541713e-06,
+      "loss": 0.2651,
+      "step": 550
+    },
+    {
+      "epoch": 0.18400400734680247,
+      "grad_norm": 0.8746807107987543,
+      "learning_rate": 6.129032258064517e-06,
+      "loss": 0.2568,
+      "step": 551
+    },
+    {
+      "epoch": 0.18433795291367508,
+      "grad_norm": 0.8681694547363,
+      "learning_rate": 6.1401557285873195e-06,
+      "loss": 0.2638,
+      "step": 552
+    },
+    {
+      "epoch": 0.18467189848054766,
+      "grad_norm": 0.8221512767810651,
+      "learning_rate": 6.151279199110123e-06,
+      "loss": 0.2415,
+      "step": 553
+    },
+    {
+      "epoch": 0.18500584404742026,
+      "grad_norm": 0.698387186152687,
+      "learning_rate": 6.162402669632927e-06,
+      "loss": 0.245,
+      "step": 554
+    },
+    {
+      "epoch": 0.18533978961429287,
+      "grad_norm": 0.8433273577991696,
+      "learning_rate": 6.173526140155729e-06,
+      "loss": 0.2649,
+      "step": 555
+    },
+    {
+      "epoch": 0.18567373518116548,
+      "grad_norm": 0.8021748456512204,
+      "learning_rate": 6.184649610678533e-06,
+      "loss": 0.2469,
+      "step": 556
+    },
+    {
+      "epoch": 0.18600768074803806,
+      "grad_norm": 0.7364557546886381,
+      "learning_rate": 6.195773081201335e-06,
+      "loss": 0.255,
+      "step": 557
+    },
+    {
+      "epoch": 0.18634162631491066,
+      "grad_norm": 0.7762130750403377,
+      "learning_rate": 6.206896551724138e-06,
+      "loss": 0.2524,
+      "step": 558
+    },
+    {
+      "epoch": 0.18667557188178327,
+      "grad_norm": 0.8366629532210769,
+      "learning_rate": 6.218020022246941e-06,
+      "loss": 0.2746,
+      "step": 559
+    },
+    {
+      "epoch": 0.18700951744865588,
+      "grad_norm": 0.752997914423049,
+      "learning_rate": 6.229143492769745e-06,
+      "loss": 0.2523,
+      "step": 560
+    },
+    {
+      "epoch": 0.18734346301552846,
+      "grad_norm": 0.7260230957886581,
+      "learning_rate": 6.240266963292548e-06,
+      "loss": 0.2541,
+      "step": 561
+    },
+    {
+      "epoch": 0.18767740858240106,
+      "grad_norm": 0.8788917141496118,
+      "learning_rate": 6.251390433815351e-06,
+      "loss": 0.2715,
+      "step": 562
+    },
+    {
+      "epoch": 0.18801135414927367,
+      "grad_norm": 0.800907436768764,
+      "learning_rate": 6.262513904338154e-06,
+      "loss": 0.2645,
+      "step": 563
+    },
+    {
+      "epoch": 0.18834529971614627,
+      "grad_norm": 0.6844788488749838,
+      "learning_rate": 6.273637374860957e-06,
+      "loss": 0.2427,
+      "step": 564
+    },
+    {
+      "epoch": 0.18867924528301888,
+      "grad_norm": 0.8650404495590827,
+      "learning_rate": 6.284760845383761e-06,
+      "loss": 0.2517,
+      "step": 565
+    },
+    {
+      "epoch": 0.18901319084989146,
+      "grad_norm": 0.9007698022142087,
+      "learning_rate": 6.295884315906563e-06,
+      "loss": 0.2321,
+      "step": 566
+    },
+    {
+      "epoch": 0.18934713641676407,
+      "grad_norm": 0.7993098878472382,
+      "learning_rate": 6.307007786429367e-06,
+      "loss": 0.2431,
+      "step": 567
+    },
+    {
+      "epoch": 0.18968108198363667,
+      "grad_norm": 0.9721899662902379,
+      "learning_rate": 6.318131256952169e-06,
+      "loss": 0.2605,
+      "step": 568
+    },
+    {
+      "epoch": 0.19001502755050928,
+      "grad_norm": 0.8389462490316344,
+      "learning_rate": 6.329254727474972e-06,
+      "loss": 0.2601,
+      "step": 569
+    },
+    {
+      "epoch": 0.19034897311738186,
+      "grad_norm": 0.8414565668740682,
+      "learning_rate": 6.340378197997776e-06,
+      "loss": 0.2621,
+      "step": 570
+    },
+    {
+      "epoch": 0.19068291868425447,
+      "grad_norm": 0.8398750862803184,
+      "learning_rate": 6.351501668520579e-06,
+      "loss": 0.2695,
+      "step": 571
+    },
+    {
+      "epoch": 0.19101686425112707,
+      "grad_norm": 0.8866607655134973,
+      "learning_rate": 6.362625139043382e-06,
+      "loss": 0.2533,
+      "step": 572
+    },
+    {
+      "epoch": 0.19135080981799968,
+      "grad_norm": 0.9367172444562246,
+      "learning_rate": 6.373748609566185e-06,
+      "loss": 0.2586,
+      "step": 573
+    },
+    {
+      "epoch": 0.19168475538487226,
+      "grad_norm": 0.8225843599787163,
+      "learning_rate": 6.3848720800889884e-06,
+      "loss": 0.2565,
+      "step": 574
+    },
+    {
+      "epoch": 0.19201870095174486,
+      "grad_norm": 0.8444529094519592,
+      "learning_rate": 6.395995550611791e-06,
+      "loss": 0.2752,
+      "step": 575
+    },
+    {
+      "epoch": 0.19235264651861747,
+      "grad_norm": 0.8144999053793032,
+      "learning_rate": 6.407119021134595e-06,
+      "loss": 0.25,
+      "step": 576
+    },
+    {
+      "epoch": 0.19268659208549008,
+      "grad_norm": 0.9646835095492448,
+      "learning_rate": 6.418242491657397e-06,
+      "loss": 0.2797,
+      "step": 577
+    },
+    {
+      "epoch": 0.19302053765236266,
+      "grad_norm": 0.8030364087953927,
+      "learning_rate": 6.429365962180201e-06,
+      "loss": 0.2521,
+      "step": 578
+    },
+    {
+      "epoch": 0.19335448321923526,
+      "grad_norm": 0.7321772482455484,
+      "learning_rate": 6.4404894327030045e-06,
+      "loss": 0.2407,
+      "step": 579
+    },
+    {
+      "epoch": 0.19368842878610787,
+      "grad_norm": 0.8945349521153622,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 0.2606,
+      "step": 580
+    },
+    {
+      "epoch": 0.19402237435298048,
+      "grad_norm": 0.9583738923059661,
+      "learning_rate": 6.462736373748611e-06,
+      "loss": 0.2688,
+      "step": 581
+    },
+    {
+      "epoch": 0.19435631991985305,
+      "grad_norm": 0.7713487168779097,
+      "learning_rate": 6.473859844271413e-06,
+      "loss": 0.2557,
+      "step": 582
+    },
+    {
+      "epoch": 0.19469026548672566,
+      "grad_norm": 0.7917343133180527,
+      "learning_rate": 6.484983314794216e-06,
+      "loss": 0.2802,
+      "step": 583
+    },
+    {
+      "epoch": 0.19502421105359827,
+      "grad_norm": 0.8943120918526262,
+      "learning_rate": 6.496106785317019e-06,
+      "loss": 0.2501,
+      "step": 584
+    },
+    {
+      "epoch": 0.19535815662047087,
+      "grad_norm": 1.0548584285291802,
+      "learning_rate": 6.5072302558398225e-06,
+      "loss": 0.27,
+      "step": 585
+    },
+    {
+      "epoch": 0.19569210218734345,
+      "grad_norm": 0.8891494426082305,
+      "learning_rate": 6.518353726362626e-06,
+      "loss": 0.2762,
+      "step": 586
+    },
+    {
+      "epoch": 0.19602604775421606,
+      "grad_norm": 0.8248070845769653,
+      "learning_rate": 6.529477196885429e-06,
+      "loss": 0.2662,
+      "step": 587
+    },
+    {
+      "epoch": 0.19635999332108867,
+      "grad_norm": 0.7767078346910501,
+      "learning_rate": 6.540600667408232e-06,
+      "loss": 0.2714,
+      "step": 588
+    },
+    {
+      "epoch": 0.19669393888796127,
+      "grad_norm": 0.9437342483702248,
+      "learning_rate": 6.551724137931035e-06,
+      "loss": 0.2565,
+      "step": 589
+    },
+    {
+      "epoch": 0.19702788445483385,
+      "grad_norm": 0.7538036441233125,
+      "learning_rate": 6.5628476084538385e-06,
+      "loss": 0.2489,
+      "step": 590
+    },
+    {
+      "epoch": 0.19736183002170646,
+      "grad_norm": 1.3943431533296153,
+      "learning_rate": 6.573971078976641e-06,
+      "loss": 0.253,
+      "step": 591
+    },
+    {
+      "epoch": 0.19769577558857906,
+      "grad_norm": 0.8252629179756918,
+      "learning_rate": 6.585094549499445e-06,
+      "loss": 0.2514,
+      "step": 592
+    },
+    {
+      "epoch": 0.19802972115545167,
+      "grad_norm": 0.7701423212327584,
+      "learning_rate": 6.596218020022247e-06,
+      "loss": 0.2637,
+      "step": 593
+    },
+    {
+      "epoch": 0.19836366672232425,
+      "grad_norm": 0.9533433627482636,
+      "learning_rate": 6.60734149054505e-06,
+      "loss": 0.2611,
+      "step": 594
+    },
+    {
+      "epoch": 0.19869761228919686,
+      "grad_norm": 1.0300962240262892,
+      "learning_rate": 6.618464961067854e-06,
+      "loss": 0.2606,
+      "step": 595
+    },
+    {
+      "epoch": 0.19903155785606946,
+      "grad_norm": 0.7274771616034863,
+      "learning_rate": 6.6295884315906565e-06,
+      "loss": 0.247,
+      "step": 596
+    },
+    {
+      "epoch": 0.19936550342294207,
+      "grad_norm": 0.8521863744652877,
+      "learning_rate": 6.64071190211346e-06,
+      "loss": 0.2633,
+      "step": 597
+    },
+    {
+      "epoch": 0.19969944898981465,
+      "grad_norm": 0.857057035238085,
+      "learning_rate": 6.651835372636263e-06,
+      "loss": 0.2589,
+      "step": 598
+    },
+    {
+      "epoch": 0.20003339455668726,
+      "grad_norm": 0.7632237378212952,
+      "learning_rate": 6.662958843159066e-06,
+      "loss": 0.252,
+      "step": 599
+    },
+    {
+      "epoch": 0.20036734012355986,
+      "grad_norm": 0.7570323690891942,
+      "learning_rate": 6.674082313681869e-06,
+      "loss": 0.2493,
+      "step": 600
+    },
+    {
+      "epoch": 0.20070128569043247,
+      "grad_norm": 0.7808753856394541,
+      "learning_rate": 6.6852057842046726e-06,
+      "loss": 0.2525,
+      "step": 601
+    },
+    {
+      "epoch": 0.20103523125730505,
+      "grad_norm": 0.7516839563088439,
+      "learning_rate": 6.696329254727475e-06,
+      "loss": 0.2531,
+      "step": 602
+    },
+    {
+      "epoch": 0.20136917682417765,
+      "grad_norm": 0.9130303291108971,
+      "learning_rate": 6.707452725250279e-06,
+      "loss": 0.2602,
+      "step": 603
+    },
+    {
+      "epoch": 0.20170312239105026,
+      "grad_norm": 0.7675587493061325,
+      "learning_rate": 6.718576195773082e-06,
+      "loss": 0.2558,
+      "step": 604
+    },
+    {
+      "epoch": 0.20203706795792287,
+      "grad_norm": 0.8499898280803776,
+      "learning_rate": 6.729699666295884e-06,
+      "loss": 0.2521,
+      "step": 605
+    },
+    {
+      "epoch": 0.20237101352479545,
+      "grad_norm": 0.6792136287125394,
+      "learning_rate": 6.740823136818689e-06,
+      "loss": 0.241,
+      "step": 606
+    },
+    {
+      "epoch": 0.20270495909166805,
+      "grad_norm": 0.8162142022991196,
+      "learning_rate": 6.7519466073414905e-06,
+      "loss": 0.2527,
+      "step": 607
+    },
+    {
+      "epoch": 0.20303890465854066,
+      "grad_norm": 0.7296861143295201,
+      "learning_rate": 6.763070077864294e-06,
+      "loss": 0.2435,
+      "step": 608
+    },
+    {
+      "epoch": 0.20337285022541327,
+      "grad_norm": 0.7502154456382952,
+      "learning_rate": 6.774193548387097e-06,
+      "loss": 0.2614,
+      "step": 609
+    },
+    {
+      "epoch": 0.20370679579228584,
+      "grad_norm": 0.7661648899048032,
+      "learning_rate": 6.7853170189099e-06,
+      "loss": 0.2559,
+      "step": 610
+    },
+    {
+      "epoch": 0.20404074135915845,
+      "grad_norm": 0.7204037874440861,
+      "learning_rate": 6.796440489432704e-06,
+      "loss": 0.2414,
+      "step": 611
+    },
+    {
+      "epoch": 0.20437468692603106,
+      "grad_norm": 0.8284127688423369,
+      "learning_rate": 6.807563959955507e-06,
+      "loss": 0.2486,
+      "step": 612
+    },
+    {
+      "epoch": 0.20470863249290366,
+      "grad_norm": 0.7646957746452056,
+      "learning_rate": 6.81868743047831e-06,
+      "loss": 0.2482,
+      "step": 613
+    },
+    {
+      "epoch": 0.20504257805977624,
+      "grad_norm": 0.6666997972855363,
+      "learning_rate": 6.829810901001113e-06,
+      "loss": 0.2412,
+      "step": 614
+    },
+    {
+      "epoch": 0.20537652362664885,
+      "grad_norm": 0.8060448469342122,
+      "learning_rate": 6.840934371523916e-06,
+      "loss": 0.2592,
+      "step": 615
+    },
+    {
+      "epoch": 0.20571046919352146,
+      "grad_norm": 0.7430816705543792,
+      "learning_rate": 6.852057842046719e-06,
+      "loss": 0.2633,
+      "step": 616
+    },
+    {
+      "epoch": 0.20604441476039406,
+      "grad_norm": 0.705554520404781,
+      "learning_rate": 6.863181312569523e-06,
+      "loss": 0.2547,
+      "step": 617
+    },
+    {
+      "epoch": 0.20637836032726667,
+      "grad_norm": 0.7321042548247252,
+      "learning_rate": 6.8743047830923245e-06,
+      "loss": 0.2465,
+      "step": 618
+    },
+    {
+      "epoch": 0.20671230589413925,
+      "grad_norm": 0.7046105183839342,
+      "learning_rate": 6.885428253615128e-06,
+      "loss": 0.2422,
+      "step": 619
+    },
+    {
+      "epoch": 0.20704625146101185,
+      "grad_norm": 0.6631002359153175,
+      "learning_rate": 6.896551724137932e-06,
+      "loss": 0.2338,
+      "step": 620
+    },
+    {
+      "epoch": 0.20738019702788446,
+      "grad_norm": 0.820463308715899,
+      "learning_rate": 6.907675194660734e-06,
+      "loss": 0.2357,
+      "step": 621
+    },
+    {
+      "epoch": 0.20771414259475707,
+      "grad_norm": 0.8757054022253623,
+      "learning_rate": 6.918798665183538e-06,
+      "loss": 0.2576,
+      "step": 622
+    },
+    {
+      "epoch": 0.20804808816162965,
+      "grad_norm": 0.7617466716742917,
+      "learning_rate": 6.929922135706341e-06,
+      "loss": 0.2625,
+      "step": 623
+    },
+    {
+      "epoch": 0.20838203372850225,
+      "grad_norm": 0.7507182463449198,
+      "learning_rate": 6.941045606229144e-06,
+      "loss": 0.2484,
+      "step": 624
+    },
+    {
+      "epoch": 0.20871597929537486,
+      "grad_norm": 0.653627768081853,
+      "learning_rate": 6.952169076751947e-06,
+      "loss": 0.228,
+      "step": 625
+    },
+    {
+      "epoch": 0.20904992486224747,
+      "grad_norm": 0.7532509534249405,
+      "learning_rate": 6.9632925472747504e-06,
+      "loss": 0.2594,
+      "step": 626
+    },
+    {
+      "epoch": 0.20938387042912004,
+      "grad_norm": 0.6656450442528803,
+      "learning_rate": 6.974416017797554e-06,
+      "loss": 0.2434,
+      "step": 627
+    },
+    {
+      "epoch": 0.20971781599599265,
+      "grad_norm": 0.9213023326955349,
+      "learning_rate": 6.985539488320357e-06,
+      "loss": 0.2577,
+      "step": 628
+    },
+    {
+      "epoch": 0.21005176156286526,
+      "grad_norm": 0.8151666294986606,
+      "learning_rate": 6.99666295884316e-06,
+      "loss": 0.2582,
+      "step": 629
+    },
+    {
+      "epoch": 0.21038570712973786,
+      "grad_norm": 0.7631009288121973,
+      "learning_rate": 7.007786429365962e-06,
+      "loss": 0.2578,
+      "step": 630
+    },
+    {
+      "epoch": 0.21071965269661044,
+      "grad_norm": 0.8363790627747725,
+      "learning_rate": 7.0189098998887665e-06,
+      "loss": 0.2781,
+      "step": 631
+    },
+    {
+      "epoch": 0.21105359826348305,
+      "grad_norm": 0.839505394509319,
+      "learning_rate": 7.030033370411568e-06,
+      "loss": 0.2314,
+      "step": 632
+    },
+    {
+      "epoch": 0.21138754383035566,
+      "grad_norm": 1.0596615080003053,
+      "learning_rate": 7.041156840934372e-06,
+      "loss": 0.284,
+      "step": 633
+    },
+    {
+      "epoch": 0.21172148939722826,
+      "grad_norm": 0.7122109086577653,
+      "learning_rate": 7.052280311457175e-06,
+      "loss": 0.2304,
+      "step": 634
+    },
+    {
+      "epoch": 0.21205543496410084,
+      "grad_norm": 0.8054756743369217,
+      "learning_rate": 7.063403781979978e-06,
+      "loss": 0.2714,
+      "step": 635
+    },
+    {
+      "epoch": 0.21238938053097345,
+      "grad_norm": 0.7150051499547507,
+      "learning_rate": 7.074527252502782e-06,
+      "loss": 0.2456,
+      "step": 636
+    },
+    {
+      "epoch": 0.21272332609784605,
+      "grad_norm": 0.7974489943926535,
+      "learning_rate": 7.0856507230255845e-06,
+      "loss": 0.2571,
+      "step": 637
+    },
+    {
+      "epoch": 0.21305727166471866,
+      "grad_norm": 0.9680300225233516,
+      "learning_rate": 7.096774193548388e-06,
+      "loss": 0.2745,
+      "step": 638
+    },
+    {
+      "epoch": 0.21339121723159124,
+      "grad_norm": 0.7427730803660618,
+      "learning_rate": 7.107897664071191e-06,
+      "loss": 0.2518,
+      "step": 639
+    },
+    {
+      "epoch": 0.21372516279846385,
+      "grad_norm": 0.741805688846562,
+      "learning_rate": 7.119021134593994e-06,
+      "loss": 0.2444,
+      "step": 640
+    },
+    {
+      "epoch": 0.21405910836533645,
+      "grad_norm": 0.7821624254946435,
+      "learning_rate": 7.130144605116797e-06,
+      "loss": 0.2727,
+      "step": 641
+    },
+    {
+      "epoch": 0.21439305393220906,
+      "grad_norm": 0.6585665747997141,
+      "learning_rate": 7.1412680756396006e-06,
+      "loss": 0.2408,
+      "step": 642
+    },
+    {
+      "epoch": 0.21472699949908164,
+      "grad_norm": 0.6560687036372677,
+      "learning_rate": 7.152391546162402e-06,
+      "loss": 0.2396,
+      "step": 643
+    },
+    {
+      "epoch": 0.21506094506595425,
+      "grad_norm": 0.746510194063174,
+      "learning_rate": 7.163515016685206e-06,
+      "loss": 0.244,
+      "step": 644
+    },
+    {
+      "epoch": 0.21539489063282685,
+      "grad_norm": 0.7591594490569733,
+      "learning_rate": 7.1746384872080095e-06,
+      "loss": 0.2501,
+      "step": 645
+    },
+    {
+      "epoch": 0.21572883619969946,
+      "grad_norm": 0.9390381754063648,
+      "learning_rate": 7.185761957730812e-06,
+      "loss": 0.2597,
+      "step": 646
+    },
+    {
+      "epoch": 0.21606278176657204,
+      "grad_norm": 0.7677869272615818,
+      "learning_rate": 7.196885428253616e-06,
+      "loss": 0.2309,
+      "step": 647
+    },
+    {
+      "epoch": 0.21639672733344464,
+      "grad_norm": 0.7831044566641154,
+      "learning_rate": 7.2080088987764185e-06,
+      "loss": 0.2674,
+      "step": 648
+    },
+    {
+      "epoch": 0.21673067290031725,
+      "grad_norm": 0.7513498348722037,
+      "learning_rate": 7.219132369299222e-06,
+      "loss": 0.2608,
+      "step": 649
+    },
+    {
+      "epoch": 0.21706461846718986,
+      "grad_norm": 0.8318654606544825,
+      "learning_rate": 7.230255839822025e-06,
+      "loss": 0.277,
+      "step": 650
+    },
+    {
+      "epoch": 0.21739856403406244,
+      "grad_norm": 0.6802811622216082,
+      "learning_rate": 7.241379310344828e-06,
+      "loss": 0.2319,
+      "step": 651
+    },
+    {
+      "epoch": 0.21773250960093504,
+      "grad_norm": 1.106089563161147,
+      "learning_rate": 7.252502780867632e-06,
+      "loss": 0.2582,
+      "step": 652
+    },
+    {
+      "epoch": 0.21806645516780765,
+      "grad_norm": 0.7584213630649976,
+      "learning_rate": 7.263626251390435e-06,
+      "loss": 0.2495,
+      "step": 653
+    },
+    {
+      "epoch": 0.21840040073468026,
+      "grad_norm": 0.7854757276260729,
+      "learning_rate": 7.274749721913238e-06,
+      "loss": 0.2557,
+      "step": 654
+    },
+    {
+      "epoch": 0.21873434630155283,
+      "grad_norm": 0.7612163368733534,
+      "learning_rate": 7.28587319243604e-06,
+      "loss": 0.2451,
+      "step": 655
+    },
+    {
+      "epoch": 0.21906829186842544,
+      "grad_norm": 0.7748245742742014,
+      "learning_rate": 7.296996662958844e-06,
+      "loss": 0.2442,
+      "step": 656
+    },
+    {
+      "epoch": 0.21940223743529805,
+      "grad_norm": 0.6807686068804447,
+      "learning_rate": 7.308120133481646e-06,
+      "loss": 0.2408,
+      "step": 657
+    },
+    {
+      "epoch": 0.21973618300217065,
+      "grad_norm": 0.9239556234849106,
+      "learning_rate": 7.31924360400445e-06,
+      "loss": 0.2702,
+      "step": 658
+    },
+    {
+      "epoch": 0.22007012856904323,
+      "grad_norm": 0.7321576339149546,
+      "learning_rate": 7.3303670745272525e-06,
+      "loss": 0.2644,
+      "step": 659
+    },
+    {
+      "epoch": 0.22040407413591584,
+      "grad_norm": 0.7996011030861904,
+      "learning_rate": 7.341490545050056e-06,
+      "loss": 0.2489,
+      "step": 660
+    },
+    {
+      "epoch": 0.22073801970278845,
+      "grad_norm": 0.8764342155392295,
+      "learning_rate": 7.35261401557286e-06,
+      "loss": 0.2751,
+      "step": 661
+    },
+    {
+      "epoch": 0.22107196526966105,
+      "grad_norm": 0.8597656221387565,
+      "learning_rate": 7.363737486095662e-06,
+      "loss": 0.2786,
+      "step": 662
+    },
+    {
+      "epoch": 0.22140591083653363,
+      "grad_norm": 0.6958608408138943,
+      "learning_rate": 7.374860956618466e-06,
+      "loss": 0.2538,
+      "step": 663
+    },
+    {
+      "epoch": 0.22173985640340624,
+      "grad_norm": 0.7016554454611178,
+      "learning_rate": 7.385984427141269e-06,
+      "loss": 0.2467,
+      "step": 664
+    },
+    {
+      "epoch": 0.22207380197027884,
+      "grad_norm": 0.7640514477559327,
+      "learning_rate": 7.397107897664072e-06,
+      "loss": 0.2634,
+      "step": 665
+    },
+    {
+      "epoch": 0.22240774753715145,
+      "grad_norm": 0.7363554934776556,
+      "learning_rate": 7.408231368186875e-06,
+      "loss": 0.2298,
+      "step": 666
+    },
+    {
+      "epoch": 0.22274169310402406,
+      "grad_norm": 0.8144798385573445,
+      "learning_rate": 7.4193548387096784e-06,
+      "loss": 0.2443,
+      "step": 667
+    },
+    {
+      "epoch": 0.22307563867089664,
+      "grad_norm": 0.8918153080014487,
+      "learning_rate": 7.43047830923248e-06,
+      "loss": 0.2578,
+      "step": 668
+    },
+    {
+      "epoch": 0.22340958423776924,
+      "grad_norm": 0.7303457303558588,
+      "learning_rate": 7.441601779755284e-06,
+      "loss": 0.2674,
+      "step": 669
+    },
+    {
+      "epoch": 0.22374352980464185,
+      "grad_norm": 0.6891671510625942,
+      "learning_rate": 7.452725250278087e-06,
+      "loss": 0.2528,
+      "step": 670
+    },
+    {
+      "epoch": 0.22407747537151446,
+      "grad_norm": 0.7806126310729112,
+      "learning_rate": 7.46384872080089e-06,
+      "loss": 0.2571,
+      "step": 671
+    },
+    {
+      "epoch": 0.22441142093838704,
+      "grad_norm": 0.6843873330049303,
+      "learning_rate": 7.474972191323694e-06,
+      "loss": 0.2255,
+      "step": 672
+    },
+    {
+      "epoch": 0.22474536650525964,
+      "grad_norm": 0.8453798784973325,
+      "learning_rate": 7.486095661846496e-06,
+      "loss": 0.2625,
+      "step": 673
+    },
+    {
+      "epoch": 0.22507931207213225,
+      "grad_norm": 0.8434557324517794,
+      "learning_rate": 7.4972191323693e-06,
+      "loss": 0.2776,
+      "step": 674
+    },
+    {
+      "epoch": 0.22541325763900485,
+      "grad_norm": 0.7048585497108847,
+      "learning_rate": 7.508342602892103e-06,
+      "loss": 0.2585,
+      "step": 675
+    },
+    {
+      "epoch": 0.22574720320587743,
+      "grad_norm": 0.7659379836914064,
+      "learning_rate": 7.519466073414906e-06,
+      "loss": 0.2548,
+      "step": 676
+    },
+    {
+      "epoch": 0.22608114877275004,
+      "grad_norm": 0.6934227319885261,
+      "learning_rate": 7.53058954393771e-06,
+      "loss": 0.2415,
+      "step": 677
+    },
+    {
+      "epoch": 0.22641509433962265,
+      "grad_norm": 0.7811227987102337,
+      "learning_rate": 7.5417130144605125e-06,
+      "loss": 0.2517,
+      "step": 678
+    },
+    {
+      "epoch": 0.22674903990649525,
+      "grad_norm": 0.804992221647555,
+      "learning_rate": 7.552836484983316e-06,
+      "loss": 0.2741,
+      "step": 679
+    },
+    {
+      "epoch": 0.22708298547336783,
+      "grad_norm": 0.8687196799605786,
+      "learning_rate": 7.563959955506118e-06,
+      "loss": 0.2727,
+      "step": 680
+    },
+    {
+      "epoch": 0.22741693104024044,
+      "grad_norm": 0.7320254452663593,
+      "learning_rate": 7.575083426028922e-06,
+      "loss": 0.2593,
+      "step": 681
+    },
+    {
+      "epoch": 0.22775087660711305,
+      "grad_norm": 0.8031326712488748,
+      "learning_rate": 7.586206896551724e-06,
+      "loss": 0.2539,
+      "step": 682
+    },
+    {
+      "epoch": 0.22808482217398565,
+      "grad_norm": 0.7676874292506355,
+      "learning_rate": 7.597330367074528e-06,
+      "loss": 0.2327,
+      "step": 683
+    },
+    {
+      "epoch": 0.22841876774085823,
+      "grad_norm": 0.7010316070246935,
+      "learning_rate": 7.60845383759733e-06,
+      "loss": 0.2351,
+      "step": 684
+    },
+    {
+      "epoch": 0.22875271330773084,
+      "grad_norm": 0.8074417436579385,
+      "learning_rate": 7.619577308120134e-06,
+      "loss": 0.2521,
+      "step": 685
+    },
+    {
+      "epoch": 0.22908665887460344,
+      "grad_norm": 0.8071936582217873,
+      "learning_rate": 7.630700778642938e-06,
+      "loss": 0.244,
+      "step": 686
+    },
+    {
+      "epoch": 0.22942060444147605,
+      "grad_norm": 0.6359795454476252,
+      "learning_rate": 7.64182424916574e-06,
+      "loss": 0.2368,
+      "step": 687
+    },
+    {
+      "epoch": 0.22975455000834863,
+      "grad_norm": 0.7777196776901433,
+      "learning_rate": 7.652947719688543e-06,
+      "loss": 0.2414,
+      "step": 688
+    },
+    {
+      "epoch": 0.23008849557522124,
+      "grad_norm": 0.8592388607477667,
+      "learning_rate": 7.664071190211346e-06,
+      "loss": 0.2455,
+      "step": 689
+    },
+    {
+      "epoch": 0.23042244114209384,
+      "grad_norm": 0.7537444209652607,
+      "learning_rate": 7.67519466073415e-06,
+      "loss": 0.2568,
+      "step": 690
+    },
+    {
+      "epoch": 0.23075638670896645,
+      "grad_norm": 0.7514964866360881,
+      "learning_rate": 7.686318131256953e-06,
+      "loss": 0.2543,
+      "step": 691
+    },
+    {
+      "epoch": 0.23109033227583903,
+      "grad_norm": 0.7856637220526238,
+      "learning_rate": 7.697441601779755e-06,
+      "loss": 0.2396,
+      "step": 692
+    },
+    {
+      "epoch": 0.23142427784271163,
+      "grad_norm": 0.6766482208992999,
+      "learning_rate": 7.70856507230256e-06,
+      "loss": 0.2306,
+      "step": 693
+    },
+    {
+      "epoch": 0.23175822340958424,
+      "grad_norm": 0.7190582180915687,
+      "learning_rate": 7.719688542825363e-06,
+      "loss": 0.2332,
+      "step": 694
+    },
+    {
+      "epoch": 0.23209216897645685,
+      "grad_norm": 0.7365246732401293,
+      "learning_rate": 7.730812013348165e-06,
+      "loss": 0.2462,
+      "step": 695
+    },
+    {
+      "epoch": 0.23242611454332943,
+      "grad_norm": 0.7531487377575422,
+      "learning_rate": 7.741935483870968e-06,
+      "loss": 0.2516,
+      "step": 696
+    },
+    {
+      "epoch": 0.23276006011020203,
+      "grad_norm": 0.7856970157332456,
+      "learning_rate": 7.753058954393772e-06,
+      "loss": 0.2595,
+      "step": 697
+    },
+    {
+      "epoch": 0.23309400567707464,
+      "grad_norm": 0.7465491947839247,
+      "learning_rate": 7.764182424916575e-06,
+      "loss": 0.243,
+      "step": 698
+    },
+    {
+      "epoch": 0.23342795124394725,
+      "grad_norm": 0.78553071211408,
+      "learning_rate": 7.775305895439378e-06,
+      "loss": 0.2727,
+      "step": 699
+    },
+    {
+      "epoch": 0.23376189681081982,
+      "grad_norm": 0.6936051447843483,
+      "learning_rate": 7.78642936596218e-06,
+      "loss": 0.2406,
+      "step": 700
+    },
+    {
+      "epoch": 0.23409584237769243,
+      "grad_norm": 0.7284183041445422,
+      "learning_rate": 7.797552836484983e-06,
+      "loss": 0.2498,
+      "step": 701
+    },
+    {
+      "epoch": 0.23442978794456504,
+      "grad_norm": 0.7500119738086012,
+      "learning_rate": 7.808676307007788e-06,
+      "loss": 0.2353,
+      "step": 702
+    },
+    {
+      "epoch": 0.23476373351143764,
+      "grad_norm": 0.7405650072009493,
+      "learning_rate": 7.81979977753059e-06,
+      "loss": 0.2565,
+      "step": 703
+    },
+    {
+      "epoch": 0.23509767907831022,
+      "grad_norm": 0.7259042405639793,
+      "learning_rate": 7.830923248053393e-06,
+      "loss": 0.2448,
+      "step": 704
+    },
+    {
+      "epoch": 0.23543162464518283,
+      "grad_norm": 0.6356529795786438,
+      "learning_rate": 7.842046718576196e-06,
+      "loss": 0.2275,
+      "step": 705
+    },
+    {
+      "epoch": 0.23576557021205544,
+      "grad_norm": 0.7052503065077377,
+      "learning_rate": 7.853170189099e-06,
+      "loss": 0.2475,
+      "step": 706
+    },
+    {
+      "epoch": 0.23609951577892804,
+      "grad_norm": 0.7686445726114107,
+      "learning_rate": 7.864293659621803e-06,
+      "loss": 0.2644,
+      "step": 707
+    },
+    {
+      "epoch": 0.23643346134580062,
+      "grad_norm": 0.7484144633557975,
+      "learning_rate": 7.875417130144606e-06,
+      "loss": 0.2575,
+      "step": 708
+    },
+    {
+      "epoch": 0.23676740691267323,
+      "grad_norm": 0.7357228659660728,
+      "learning_rate": 7.886540600667408e-06,
+      "loss": 0.2372,
+      "step": 709
+    },
+    {
+      "epoch": 0.23710135247954583,
+      "grad_norm": 0.6982740541543035,
+      "learning_rate": 7.897664071190213e-06,
+      "loss": 0.2508,
+      "step": 710
+    },
+    {
+      "epoch": 0.23743529804641844,
+      "grad_norm": 0.7398245740088609,
+      "learning_rate": 7.908787541713015e-06,
+      "loss": 0.2381,
+      "step": 711
+    },
+    {
+      "epoch": 0.23776924361329102,
+      "grad_norm": 0.7895941667598881,
+      "learning_rate": 7.919911012235818e-06,
+      "loss": 0.2283,
+      "step": 712
+    },
+    {
+      "epoch": 0.23810318918016363,
+      "grad_norm": 0.779972681262703,
+      "learning_rate": 7.93103448275862e-06,
+      "loss": 0.2577,
+      "step": 713
+    },
+    {
+      "epoch": 0.23843713474703623,
+      "grad_norm": 0.7064211035558918,
+      "learning_rate": 7.942157953281424e-06,
+      "loss": 0.2391,
+      "step": 714
+    },
+    {
+      "epoch": 0.23877108031390884,
+      "grad_norm": 0.8977366266428429,
+      "learning_rate": 7.953281423804228e-06,
+      "loss": 0.2735,
+      "step": 715
+    },
+    {
+      "epoch": 0.23910502588078142,
+      "grad_norm": 0.8146695015510214,
+      "learning_rate": 7.96440489432703e-06,
+      "loss": 0.2553,
+      "step": 716
+    },
+    {
+      "epoch": 0.23943897144765403,
+      "grad_norm": 0.8075630691246333,
+      "learning_rate": 7.975528364849833e-06,
+      "loss": 0.24,
+      "step": 717
+    },
+    {
+      "epoch": 0.23977291701452663,
+      "grad_norm": 0.7417746252500675,
+      "learning_rate": 7.986651835372638e-06,
+      "loss": 0.2455,
+      "step": 718
+    },
+    {
+      "epoch": 0.24010686258139924,
+      "grad_norm": 0.7960193301121595,
+      "learning_rate": 7.99777530589544e-06,
+      "loss": 0.2537,
+      "step": 719
+    },
+    {
+      "epoch": 0.24044080814827185,
+      "grad_norm": 0.6581455395066982,
+      "learning_rate": 8.008898776418243e-06,
+      "loss": 0.2508,
+      "step": 720
+    },
+    {
+      "epoch": 0.24077475371514442,
+      "grad_norm": 0.762028236187238,
+      "learning_rate": 8.020022246941046e-06,
+      "loss": 0.2726,
+      "step": 721
+    },
+    {
+      "epoch": 0.24110869928201703,
+      "grad_norm": 0.7200877982327974,
+      "learning_rate": 8.03114571746385e-06,
+      "loss": 0.2551,
+      "step": 722
+    },
+    {
+      "epoch": 0.24144264484888964,
+      "grad_norm": 0.6356328284318833,
+      "learning_rate": 8.042269187986651e-06,
+      "loss": 0.2249,
+      "step": 723
+    },
+    {
+      "epoch": 0.24177659041576224,
+      "grad_norm": 0.7300735104858539,
+      "learning_rate": 8.053392658509456e-06,
+      "loss": 0.2535,
+      "step": 724
+    },
+    {
+      "epoch": 0.24211053598263482,
+      "grad_norm": 0.7236816257988127,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 0.2521,
+      "step": 725
+    },
+    {
+      "epoch": 0.24244448154950743,
+      "grad_norm": 0.6587602905727439,
+      "learning_rate": 8.075639599555061e-06,
+      "loss": 0.2425,
+      "step": 726
+    },
+    {
+      "epoch": 0.24277842711638004,
+      "grad_norm": 0.7780907894191817,
+      "learning_rate": 8.086763070077866e-06,
+      "loss": 0.264,
+      "step": 727
+    },
+    {
+      "epoch": 0.24311237268325264,
+      "grad_norm": 0.686670566205572,
+      "learning_rate": 8.097886540600668e-06,
+      "loss": 0.2509,
+      "step": 728
+    },
+    {
+      "epoch": 0.24344631825012522,
+      "grad_norm": 0.7534125635472662,
+      "learning_rate": 8.109010011123471e-06,
+      "loss": 0.2402,
+      "step": 729
+    },
+    {
+      "epoch": 0.24378026381699783,
+      "grad_norm": 0.607443586736483,
+      "learning_rate": 8.120133481646274e-06,
+      "loss": 0.2397,
+      "step": 730
+    },
+    {
+      "epoch": 0.24411420938387043,
+      "grad_norm": 1.3378541145711298,
+      "learning_rate": 8.131256952169078e-06,
+      "loss": 0.255,
+      "step": 731
+    },
+    {
+      "epoch": 0.24444815495074304,
+      "grad_norm": 0.7693531003770262,
+      "learning_rate": 8.14238042269188e-06,
+      "loss": 0.2469,
+      "step": 732
+    },
+    {
+      "epoch": 0.24478210051761562,
+      "grad_norm": 0.7630786614916105,
+      "learning_rate": 8.153503893214683e-06,
+      "loss": 0.2529,
+      "step": 733
+    },
+    {
+      "epoch": 0.24511604608448823,
+      "grad_norm": 0.6830755477799588,
+      "learning_rate": 8.164627363737486e-06,
+      "loss": 0.2593,
+      "step": 734
+    },
+    {
+      "epoch": 0.24544999165136083,
+      "grad_norm": 0.769470982645176,
+      "learning_rate": 8.17575083426029e-06,
+      "loss": 0.2635,
+      "step": 735
+    },
+    {
+      "epoch": 0.24578393721823344,
+      "grad_norm": 0.6411315581669028,
+      "learning_rate": 8.186874304783093e-06,
+      "loss": 0.2333,
+      "step": 736
+    },
+    {
+      "epoch": 0.24611788278510602,
+      "grad_norm": 0.7021751041939086,
+      "learning_rate": 8.197997775305896e-06,
+      "loss": 0.2614,
+      "step": 737
+    },
+    {
+      "epoch": 0.24645182835197862,
+      "grad_norm": 0.6973361840787954,
+      "learning_rate": 8.209121245828699e-06,
+      "loss": 0.2517,
+      "step": 738
+    },
+    {
+      "epoch": 0.24678577391885123,
+      "grad_norm": 0.7235562634782384,
+      "learning_rate": 8.220244716351501e-06,
+      "loss": 0.247,
+      "step": 739
+    },
+    {
+      "epoch": 0.24711971948572384,
+      "grad_norm": 0.7907039415300885,
+      "learning_rate": 8.231368186874306e-06,
+      "loss": 0.2747,
+      "step": 740
+    },
+    {
+      "epoch": 0.24745366505259642,
+      "grad_norm": 0.9920594690593687,
+      "learning_rate": 8.242491657397109e-06,
+      "loss": 0.2504,
+      "step": 741
+    },
+    {
+      "epoch": 0.24778761061946902,
+      "grad_norm": 0.6620117662895729,
+      "learning_rate": 8.253615127919911e-06,
+      "loss": 0.2539,
+      "step": 742
+    },
+    {
+      "epoch": 0.24812155618634163,
+      "grad_norm": 0.8115893003059279,
+      "learning_rate": 8.264738598442716e-06,
+      "loss": 0.2445,
+      "step": 743
+    },
+    {
+      "epoch": 0.24845550175321424,
+      "grad_norm": 0.7888030282533821,
+      "learning_rate": 8.275862068965518e-06,
+      "loss": 0.2447,
+      "step": 744
+    },
+    {
+      "epoch": 0.24878944732008682,
+      "grad_norm": 0.7396486295280933,
+      "learning_rate": 8.286985539488321e-06,
+      "loss": 0.2502,
+      "step": 745
+    },
+    {
+      "epoch": 0.24912339288695942,
+      "grad_norm": 0.7245100088509931,
+      "learning_rate": 8.298109010011124e-06,
+      "loss": 0.2484,
+      "step": 746
+    },
+    {
+      "epoch": 0.24945733845383203,
+      "grad_norm": 0.7768451542815072,
+      "learning_rate": 8.309232480533928e-06,
+      "loss": 0.2537,
+      "step": 747
+    },
+    {
+      "epoch": 0.24979128402070463,
+      "grad_norm": 0.7260972447956803,
+      "learning_rate": 8.32035595105673e-06,
+      "loss": 0.2664,
+      "step": 748
+    },
+    {
+      "epoch": 0.2501252295875772,
+      "grad_norm": 0.7727276441288169,
+      "learning_rate": 8.331479421579534e-06,
+      "loss": 0.2457,
+      "step": 749
+    },
+    {
+      "epoch": 0.2504591751544498,
+      "grad_norm": 0.7116614209879674,
+      "learning_rate": 8.342602892102336e-06,
+      "loss": 0.228,
+      "step": 750
+    },
+    {
+      "epoch": 0.2507931207213224,
+      "grad_norm": 0.9127998577319423,
+      "learning_rate": 8.353726362625139e-06,
+      "loss": 0.2362,
+      "step": 751
+    },
+    {
+      "epoch": 0.25112706628819503,
+      "grad_norm": 0.8377483636026994,
+      "learning_rate": 8.364849833147943e-06,
+      "loss": 0.2464,
+      "step": 752
+    },
+    {
+      "epoch": 0.25146101185506764,
+      "grad_norm": 0.6816783448014366,
+      "learning_rate": 8.375973303670746e-06,
+      "loss": 0.2369,
+      "step": 753
+    },
+    {
+      "epoch": 0.25179495742194025,
+      "grad_norm": 0.8758433268313643,
+      "learning_rate": 8.387096774193549e-06,
+      "loss": 0.2795,
+      "step": 754
+    },
+    {
+      "epoch": 0.2521289029888128,
+      "grad_norm": 0.9056051609541013,
+      "learning_rate": 8.398220244716352e-06,
+      "loss": 0.263,
+      "step": 755
+    },
+    {
+      "epoch": 0.2524628485556854,
+      "grad_norm": 0.8917755495799926,
+      "learning_rate": 8.409343715239156e-06,
+      "loss": 0.2599,
+      "step": 756
+    },
+    {
+      "epoch": 0.252796794122558,
+      "grad_norm": 0.6905229275325627,
+      "learning_rate": 8.420467185761959e-06,
+      "loss": 0.2328,
+      "step": 757
+    },
+    {
+      "epoch": 0.2531307396894306,
+      "grad_norm": 0.68920478639127,
+      "learning_rate": 8.431590656284761e-06,
+      "loss": 0.2477,
+      "step": 758
+    },
+    {
+      "epoch": 0.2534646852563032,
+      "grad_norm": 0.745543256443487,
+      "learning_rate": 8.442714126807566e-06,
+      "loss": 0.2405,
+      "step": 759
+    },
+    {
+      "epoch": 0.25379863082317583,
+      "grad_norm": 0.7884756984382175,
+      "learning_rate": 8.453837597330368e-06,
+      "loss": 0.2362,
+      "step": 760
+    },
+    {
+      "epoch": 0.25413257639004844,
+      "grad_norm": 0.7468278824055963,
+      "learning_rate": 8.464961067853171e-06,
+      "loss": 0.262,
+      "step": 761
+    },
+    {
+      "epoch": 0.25446652195692104,
+      "grad_norm": 0.8653261962193426,
+      "learning_rate": 8.476084538375974e-06,
+      "loss": 0.2586,
+      "step": 762
+    },
+    {
+      "epoch": 0.2548004675237936,
+      "grad_norm": 0.6837818557349334,
+      "learning_rate": 8.487208008898777e-06,
+      "loss": 0.2366,
+      "step": 763
+    },
+    {
+      "epoch": 0.2551344130906662,
+      "grad_norm": 0.7536532409242654,
+      "learning_rate": 8.49833147942158e-06,
+      "loss": 0.2591,
+      "step": 764
+    },
+    {
+      "epoch": 0.2554683586575388,
+      "grad_norm": 0.9083573270282176,
+      "learning_rate": 8.509454949944384e-06,
+      "loss": 0.2521,
+      "step": 765
+    },
+    {
+      "epoch": 0.2558023042244114,
+      "grad_norm": 0.767552861904764,
+      "learning_rate": 8.520578420467186e-06,
+      "loss": 0.2625,
+      "step": 766
+    },
+    {
+      "epoch": 0.256136249791284,
+      "grad_norm": 0.7629866167443453,
+      "learning_rate": 8.531701890989989e-06,
+      "loss": 0.251,
+      "step": 767
+    },
+    {
+      "epoch": 0.2564701953581566,
+      "grad_norm": 0.7377761813541851,
+      "learning_rate": 8.542825361512793e-06,
+      "loss": 0.2404,
+      "step": 768
+    },
+    {
+      "epoch": 0.25680414092502923,
+      "grad_norm": 0.7352772762034555,
+      "learning_rate": 8.553948832035596e-06,
+      "loss": 0.2509,
+      "step": 769
+    },
+    {
+      "epoch": 0.25713808649190184,
+      "grad_norm": 0.7212760539284759,
+      "learning_rate": 8.565072302558399e-06,
+      "loss": 0.2418,
+      "step": 770
+    },
+    {
+      "epoch": 0.25747203205877445,
+      "grad_norm": 0.6973041703150606,
+      "learning_rate": 8.576195773081202e-06,
+      "loss": 0.2378,
+      "step": 771
+    },
+    {
+      "epoch": 0.257805977625647,
+      "grad_norm": 0.8275720116284327,
+      "learning_rate": 8.587319243604006e-06,
+      "loss": 0.2453,
+      "step": 772
+    },
+    {
+      "epoch": 0.2581399231925196,
+      "grad_norm": 0.7089874078732612,
+      "learning_rate": 8.598442714126807e-06,
+      "loss": 0.2363,
+      "step": 773
+    },
+    {
+      "epoch": 0.2584738687593922,
+      "grad_norm": 0.6667510504135715,
+      "learning_rate": 8.609566184649611e-06,
+      "loss": 0.2404,
+      "step": 774
+    },
+    {
+      "epoch": 0.2588078143262648,
+      "grad_norm": 0.8122294668554957,
+      "learning_rate": 8.620689655172414e-06,
+      "loss": 0.2622,
+      "step": 775
+    },
+    {
+      "epoch": 0.2591417598931374,
+      "grad_norm": 0.823450700107754,
+      "learning_rate": 8.631813125695217e-06,
+      "loss": 0.2417,
+      "step": 776
+    },
+    {
+      "epoch": 0.25947570546001003,
+      "grad_norm": 0.7726092316119784,
+      "learning_rate": 8.642936596218021e-06,
+      "loss": 0.2474,
+      "step": 777
+    },
+    {
+      "epoch": 0.25980965102688264,
+      "grad_norm": 0.6746381988563429,
+      "learning_rate": 8.654060066740824e-06,
+      "loss": 0.2445,
+      "step": 778
+    },
+    {
+      "epoch": 0.26014359659375524,
+      "grad_norm": 0.7622637807060025,
+      "learning_rate": 8.665183537263627e-06,
+      "loss": 0.2593,
+      "step": 779
+    },
+    {
+      "epoch": 0.2604775421606278,
+      "grad_norm": 0.7660043938587235,
+      "learning_rate": 8.67630700778643e-06,
+      "loss": 0.2612,
+      "step": 780
+    },
+    {
+      "epoch": 0.2608114877275004,
+      "grad_norm": 0.733656969442999,
+      "learning_rate": 8.687430478309234e-06,
+      "loss": 0.2649,
+      "step": 781
+    },
+    {
+      "epoch": 0.261145433294373,
+      "grad_norm": 0.8112312167102349,
+      "learning_rate": 8.698553948832036e-06,
+      "loss": 0.2637,
+      "step": 782
+    },
+    {
+      "epoch": 0.2614793788612456,
+      "grad_norm": 0.7552777944668807,
+      "learning_rate": 8.70967741935484e-06,
+      "loss": 0.2511,
+      "step": 783
+    },
+    {
+      "epoch": 0.2618133244281182,
+      "grad_norm": 0.6705632857664593,
+      "learning_rate": 8.720800889877644e-06,
+      "loss": 0.2517,
+      "step": 784
+    },
+    {
+      "epoch": 0.26214726999499083,
+      "grad_norm": 0.7991435702186575,
+      "learning_rate": 8.731924360400446e-06,
+      "loss": 0.2798,
+      "step": 785
+    },
+    {
+      "epoch": 0.26248121556186343,
+      "grad_norm": 0.7428406935602541,
+      "learning_rate": 8.743047830923249e-06,
+      "loss": 0.2669,
+      "step": 786
+    },
+    {
+      "epoch": 0.26281516112873604,
+      "grad_norm": 0.7460196154653285,
+      "learning_rate": 8.754171301446052e-06,
+      "loss": 0.2314,
+      "step": 787
+    },
+    {
+      "epoch": 0.2631491066956086,
+      "grad_norm": 0.6813560165573437,
+      "learning_rate": 8.765294771968854e-06,
+      "loss": 0.2387,
+      "step": 788
+    },
+    {
+      "epoch": 0.2634830522624812,
+      "grad_norm": 1.1679094658128664,
+      "learning_rate": 8.776418242491657e-06,
+      "loss": 0.2642,
+      "step": 789
+    },
+    {
+      "epoch": 0.2638169978293538,
+      "grad_norm": 0.6998489054574449,
+      "learning_rate": 8.787541713014462e-06,
+      "loss": 0.2493,
+      "step": 790
+    },
+    {
+      "epoch": 0.2641509433962264,
+      "grad_norm": 0.7233905242204927,
+      "learning_rate": 8.798665183537264e-06,
+      "loss": 0.2765,
+      "step": 791
+    },
+    {
+      "epoch": 0.264484888963099,
+      "grad_norm": 0.8130295529317806,
+      "learning_rate": 8.809788654060067e-06,
+      "loss": 0.2572,
+      "step": 792
+    },
+    {
+      "epoch": 0.2648188345299716,
+      "grad_norm": 0.734643637678886,
+      "learning_rate": 8.820912124582871e-06,
+      "loss": 0.2764,
+      "step": 793
+    },
+    {
+      "epoch": 0.26515278009684423,
+      "grad_norm": 0.8129757879046563,
+      "learning_rate": 8.832035595105674e-06,
+      "loss": 0.2597,
+      "step": 794
+    },
+    {
+      "epoch": 0.26548672566371684,
+      "grad_norm": 0.7165953475826065,
+      "learning_rate": 8.843159065628477e-06,
+      "loss": 0.2338,
+      "step": 795
+    },
+    {
+      "epoch": 0.2658206712305894,
+      "grad_norm": 0.7259835911679116,
+      "learning_rate": 8.85428253615128e-06,
+      "loss": 0.2579,
+      "step": 796
+    },
+    {
+      "epoch": 0.266154616797462,
+      "grad_norm": 0.6619475009098943,
+      "learning_rate": 8.865406006674084e-06,
+      "loss": 0.2119,
+      "step": 797
+    },
+    {
+      "epoch": 0.2664885623643346,
+      "grad_norm": 0.7625539062454664,
+      "learning_rate": 8.876529477196885e-06,
+      "loss": 0.2281,
+      "step": 798
+    },
+    {
+      "epoch": 0.2668225079312072,
+      "grad_norm": 0.7790875438082882,
+      "learning_rate": 8.88765294771969e-06,
+      "loss": 0.264,
+      "step": 799
+    },
+    {
+      "epoch": 0.2671564534980798,
+      "grad_norm": 0.7937898709271423,
+      "learning_rate": 8.898776418242492e-06,
+      "loss": 0.2448,
+      "step": 800
+    },
+    {
+      "epoch": 0.2674903990649524,
+      "grad_norm": 0.7534922554712896,
+      "learning_rate": 8.909899888765295e-06,
+      "loss": 0.2601,
+      "step": 801
+    },
+    {
+      "epoch": 0.26782434463182503,
+      "grad_norm": 0.709328797539281,
+      "learning_rate": 8.921023359288099e-06,
+      "loss": 0.2378,
+      "step": 802
+    },
+    {
+      "epoch": 0.26815829019869764,
+      "grad_norm": 0.6700322828506419,
+      "learning_rate": 8.932146829810902e-06,
+      "loss": 0.2377,
+      "step": 803
+    },
+    {
+      "epoch": 0.2684922357655702,
+      "grad_norm": 0.7397483140494081,
+      "learning_rate": 8.943270300333705e-06,
+      "loss": 0.2504,
+      "step": 804
+    },
+    {
+      "epoch": 0.2688261813324428,
+      "grad_norm": 0.6236769417814907,
+      "learning_rate": 8.954393770856507e-06,
+      "loss": 0.2407,
+      "step": 805
+    },
+    {
+      "epoch": 0.2691601268993154,
+      "grad_norm": 0.6903960276905347,
+      "learning_rate": 8.965517241379312e-06,
+      "loss": 0.2449,
+      "step": 806
+    },
+    {
+      "epoch": 0.269494072466188,
+      "grad_norm": 0.6618661572934079,
+      "learning_rate": 8.976640711902114e-06,
+      "loss": 0.2322,
+      "step": 807
+    },
+    {
+      "epoch": 0.2698280180330606,
+      "grad_norm": 0.6450814858368458,
+      "learning_rate": 8.987764182424917e-06,
+      "loss": 0.2527,
+      "step": 808
+    },
+    {
+      "epoch": 0.2701619635999332,
+      "grad_norm": 0.6732302794328731,
+      "learning_rate": 8.998887652947721e-06,
+      "loss": 0.2492,
+      "step": 809
+    },
+    {
+      "epoch": 0.2704959091668058,
+      "grad_norm": 0.7193330108510266,
+      "learning_rate": 9.010011123470524e-06,
+      "loss": 0.2521,
+      "step": 810
+    },
+    {
+      "epoch": 0.27082985473367843,
+      "grad_norm": 0.6638055458342996,
+      "learning_rate": 9.021134593993327e-06,
+      "loss": 0.2558,
+      "step": 811
+    },
+    {
+      "epoch": 0.271163800300551,
+      "grad_norm": 0.6932796625861776,
+      "learning_rate": 9.03225806451613e-06,
+      "loss": 0.2685,
+      "step": 812
+    },
+    {
+      "epoch": 0.2714977458674236,
+      "grad_norm": 0.669882215821802,
+      "learning_rate": 9.043381535038932e-06,
+      "loss": 0.2511,
+      "step": 813
+    },
+    {
+      "epoch": 0.2718316914342962,
+      "grad_norm": 0.738016797517266,
+      "learning_rate": 9.054505005561735e-06,
+      "loss": 0.2508,
+      "step": 814
+    },
+    {
+      "epoch": 0.2721656370011688,
+      "grad_norm": 0.6579473564630967,
+      "learning_rate": 9.06562847608454e-06,
+      "loss": 0.2428,
+      "step": 815
+    },
+    {
+      "epoch": 0.2724995825680414,
+      "grad_norm": 0.6601416633554115,
+      "learning_rate": 9.076751946607342e-06,
+      "loss": 0.2334,
+      "step": 816
+    },
+    {
+      "epoch": 0.272833528134914,
+      "grad_norm": 0.8681212873817967,
+      "learning_rate": 9.087875417130145e-06,
+      "loss": 0.2649,
+      "step": 817
+    },
+    {
+      "epoch": 0.2731674737017866,
+      "grad_norm": 0.6871613484359503,
+      "learning_rate": 9.09899888765295e-06,
+      "loss": 0.2488,
+      "step": 818
+    },
+    {
+      "epoch": 0.27350141926865923,
+      "grad_norm": 0.8565684394856233,
+      "learning_rate": 9.110122358175752e-06,
+      "loss": 0.2566,
+      "step": 819
+    },
+    {
+      "epoch": 0.27383536483553184,
+      "grad_norm": 0.8255322571537959,
+      "learning_rate": 9.121245828698555e-06,
+      "loss": 0.2493,
+      "step": 820
+    },
+    {
+      "epoch": 0.2741693104024044,
+      "grad_norm": 0.8229306366015684,
+      "learning_rate": 9.132369299221357e-06,
+      "loss": 0.252,
+      "step": 821
+    },
+    {
+      "epoch": 0.274503255969277,
+      "grad_norm": 0.8463175261291924,
+      "learning_rate": 9.143492769744162e-06,
+      "loss": 0.2567,
+      "step": 822
+    },
+    {
+      "epoch": 0.2748372015361496,
+      "grad_norm": 0.6728403338031879,
+      "learning_rate": 9.154616240266963e-06,
+      "loss": 0.2441,
+      "step": 823
+    },
+    {
+      "epoch": 0.2751711471030222,
+      "grad_norm": 0.8767027013992857,
+      "learning_rate": 9.165739710789767e-06,
+      "loss": 0.2473,
+      "step": 824
+    },
+    {
+      "epoch": 0.2755050926698948,
+      "grad_norm": 0.8201243130499303,
+      "learning_rate": 9.176863181312572e-06,
+      "loss": 0.2563,
+      "step": 825
+    },
+    {
+      "epoch": 0.2758390382367674,
+      "grad_norm": 0.6872232291309764,
+      "learning_rate": 9.187986651835373e-06,
+      "loss": 0.2485,
+      "step": 826
+    },
+    {
+      "epoch": 0.27617298380364,
+      "grad_norm": 0.6930371900328928,
+      "learning_rate": 9.199110122358177e-06,
+      "loss": 0.236,
+      "step": 827
+    },
+    {
+      "epoch": 0.27650692937051263,
+      "grad_norm": 0.7768765519692938,
+      "learning_rate": 9.21023359288098e-06,
+      "loss": 0.2268,
+      "step": 828
+    },
+    {
+      "epoch": 0.2768408749373852,
+      "grad_norm": 0.641650918415613,
+      "learning_rate": 9.221357063403782e-06,
+      "loss": 0.2458,
+      "step": 829
+    },
+    {
+      "epoch": 0.2771748205042578,
+      "grad_norm": 0.7847607928855562,
+      "learning_rate": 9.232480533926585e-06,
+      "loss": 0.2561,
+      "step": 830
+    },
+    {
+      "epoch": 0.2775087660711304,
+      "grad_norm": 0.7522962047924147,
+      "learning_rate": 9.24360400444939e-06,
+      "loss": 0.2462,
+      "step": 831
+    },
+    {
+      "epoch": 0.277842711638003,
+      "grad_norm": 0.7856412167874145,
+      "learning_rate": 9.254727474972192e-06,
+      "loss": 0.2453,
+      "step": 832
+    },
+    {
+      "epoch": 0.2781766572048756,
+      "grad_norm": 0.6379368244033592,
+      "learning_rate": 9.265850945494995e-06,
+      "loss": 0.2319,
+      "step": 833
+    },
+    {
+      "epoch": 0.2785106027717482,
+      "grad_norm": 0.6923372665425509,
+      "learning_rate": 9.2769744160178e-06,
+      "loss": 0.2521,
+      "step": 834
+    },
+    {
+      "epoch": 0.2788445483386208,
+      "grad_norm": 0.6285185522634985,
+      "learning_rate": 9.288097886540602e-06,
+      "loss": 0.2414,
+      "step": 835
+    },
+    {
+      "epoch": 0.27917849390549343,
+      "grad_norm": 0.7659134854937873,
+      "learning_rate": 9.299221357063405e-06,
+      "loss": 0.2617,
+      "step": 836
+    },
+    {
+      "epoch": 0.279512439472366,
+      "grad_norm": 0.651210591819511,
+      "learning_rate": 9.310344827586207e-06,
+      "loss": 0.2412,
+      "step": 837
+    },
+    {
+      "epoch": 0.2798463850392386,
+      "grad_norm": 0.604891742779754,
+      "learning_rate": 9.32146829810901e-06,
+      "loss": 0.2361,
+      "step": 838
+    },
+    {
+      "epoch": 0.2801803306061112,
+      "grad_norm": 0.6371826775462133,
+      "learning_rate": 9.332591768631813e-06,
+      "loss": 0.2512,
+      "step": 839
+    },
+    {
+      "epoch": 0.2805142761729838,
+      "grad_norm": 0.6340690502327005,
+      "learning_rate": 9.343715239154617e-06,
+      "loss": 0.2416,
+      "step": 840
+    },
+    {
+      "epoch": 0.2808482217398564,
+      "grad_norm": 0.7342239300077404,
+      "learning_rate": 9.35483870967742e-06,
+      "loss": 0.2448,
+      "step": 841
+    },
+    {
+      "epoch": 0.281182167306729,
+      "grad_norm": 0.6549609966808125,
+      "learning_rate": 9.365962180200223e-06,
+      "loss": 0.2458,
+      "step": 842
+    },
+    {
+      "epoch": 0.2815161128736016,
+      "grad_norm": 0.7299241942323713,
+      "learning_rate": 9.377085650723027e-06,
+      "loss": 0.2318,
+      "step": 843
+    },
+    {
+      "epoch": 0.2818500584404742,
+      "grad_norm": 0.6845739065601146,
+      "learning_rate": 9.38820912124583e-06,
+      "loss": 0.2428,
+      "step": 844
+    },
+    {
+      "epoch": 0.2821840040073468,
+      "grad_norm": 0.677968498084514,
+      "learning_rate": 9.399332591768633e-06,
+      "loss": 0.2594,
+      "step": 845
+    },
+    {
+      "epoch": 0.2825179495742194,
+      "grad_norm": 0.682427111871113,
+      "learning_rate": 9.410456062291435e-06,
+      "loss": 0.2504,
+      "step": 846
+    },
+    {
+      "epoch": 0.282851895141092,
+      "grad_norm": 0.6035000880972591,
+      "learning_rate": 9.42157953281424e-06,
+      "loss": 0.2399,
+      "step": 847
+    },
+    {
+      "epoch": 0.2831858407079646,
+      "grad_norm": 0.7419050323765086,
+      "learning_rate": 9.43270300333704e-06,
+      "loss": 0.2724,
+      "step": 848
+    },
+    {
+      "epoch": 0.2835197862748372,
+      "grad_norm": 0.6133376838559707,
+      "learning_rate": 9.443826473859845e-06,
+      "loss": 0.2289,
+      "step": 849
+    },
+    {
+      "epoch": 0.2838537318417098,
+      "grad_norm": 0.687952283763828,
+      "learning_rate": 9.45494994438265e-06,
+      "loss": 0.2427,
+      "step": 850
+    },
+    {
+      "epoch": 0.2841876774085824,
+      "grad_norm": 0.7028605452057839,
+      "learning_rate": 9.46607341490545e-06,
+      "loss": 0.2476,
+      "step": 851
+    },
+    {
+      "epoch": 0.284521622975455,
+      "grad_norm": 0.6623344875228663,
+      "learning_rate": 9.477196885428255e-06,
+      "loss": 0.2452,
+      "step": 852
+    },
+    {
+      "epoch": 0.2848555685423276,
+      "grad_norm": 0.6332298537339767,
+      "learning_rate": 9.488320355951058e-06,
+      "loss": 0.2399,
+      "step": 853
+    },
+    {
+      "epoch": 0.2851895141092002,
+      "grad_norm": 0.6645743882820965,
+      "learning_rate": 9.49944382647386e-06,
+      "loss": 0.2447,
+      "step": 854
+    },
+    {
+      "epoch": 0.2855234596760728,
+      "grad_norm": 0.6662259006288558,
+      "learning_rate": 9.510567296996663e-06,
+      "loss": 0.2296,
+      "step": 855
+    },
+    {
+      "epoch": 0.2858574052429454,
+      "grad_norm": 0.6919974903102527,
+      "learning_rate": 9.521690767519467e-06,
+      "loss": 0.2537,
+      "step": 856
+    },
+    {
+      "epoch": 0.286191350809818,
+      "grad_norm": 0.6454163080434254,
+      "learning_rate": 9.53281423804227e-06,
+      "loss": 0.2514,
+      "step": 857
+    },
+    {
+      "epoch": 0.2865252963766906,
+      "grad_norm": 0.6480489173178399,
+      "learning_rate": 9.543937708565073e-06,
+      "loss": 0.2422,
+      "step": 858
+    },
+    {
+      "epoch": 0.2868592419435632,
+      "grad_norm": 0.7353022665236786,
+      "learning_rate": 9.555061179087877e-06,
+      "loss": 0.2351,
+      "step": 859
+    },
+    {
+      "epoch": 0.2871931875104358,
+      "grad_norm": 0.661065913724645,
+      "learning_rate": 9.56618464961068e-06,
+      "loss": 0.2479,
+      "step": 860
+    },
+    {
+      "epoch": 0.28752713307730837,
+      "grad_norm": 0.6763453703327771,
+      "learning_rate": 9.577308120133483e-06,
+      "loss": 0.2226,
+      "step": 861
+    },
+    {
+      "epoch": 0.287861078644181,
+      "grad_norm": 0.6727786879376527,
+      "learning_rate": 9.588431590656285e-06,
+      "loss": 0.2548,
+      "step": 862
+    },
+    {
+      "epoch": 0.2881950242110536,
+      "grad_norm": 0.9197408709347626,
+      "learning_rate": 9.599555061179088e-06,
+      "loss": 0.2489,
+      "step": 863
+    },
+    {
+      "epoch": 0.2885289697779262,
+      "grad_norm": 0.6646129153644628,
+      "learning_rate": 9.61067853170189e-06,
+      "loss": 0.2541,
+      "step": 864
+    },
+    {
+      "epoch": 0.2888629153447988,
+      "grad_norm": 0.7687613388007959,
+      "learning_rate": 9.621802002224695e-06,
+      "loss": 0.2549,
+      "step": 865
+    },
+    {
+      "epoch": 0.2891968609116714,
+      "grad_norm": 0.663099474914373,
+      "learning_rate": 9.632925472747498e-06,
+      "loss": 0.255,
+      "step": 866
+    },
+    {
+      "epoch": 0.289530806478544,
+      "grad_norm": 0.6373838203654694,
+      "learning_rate": 9.6440489432703e-06,
+      "loss": 0.2274,
+      "step": 867
+    },
+    {
+      "epoch": 0.2898647520454166,
+      "grad_norm": 0.8958296860718516,
+      "learning_rate": 9.655172413793105e-06,
+      "loss": 0.2744,
+      "step": 868
+    },
+    {
+      "epoch": 0.29019869761228917,
+      "grad_norm": 0.6689521121922187,
+      "learning_rate": 9.666295884315908e-06,
+      "loss": 0.2486,
+      "step": 869
+    },
+    {
+      "epoch": 0.2905326431791618,
+      "grad_norm": 0.7051581520556306,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 0.2325,
+      "step": 870
+    },
+    {
+      "epoch": 0.2908665887460344,
+      "grad_norm": 0.6723499452278224,
+      "learning_rate": 9.688542825361513e-06,
+      "loss": 0.2494,
+      "step": 871
+    },
+    {
+      "epoch": 0.291200534312907,
+      "grad_norm": 0.7127882744746751,
+      "learning_rate": 9.699666295884318e-06,
+      "loss": 0.2492,
+      "step": 872
+    },
+    {
+      "epoch": 0.2915344798797796,
+      "grad_norm": 0.6707104286364339,
+      "learning_rate": 9.710789766407119e-06,
+      "loss": 0.251,
+      "step": 873
+    },
+    {
+      "epoch": 0.2918684254466522,
+      "grad_norm": 0.781136986166393,
+      "learning_rate": 9.721913236929923e-06,
+      "loss": 0.2484,
+      "step": 874
+    },
+    {
+      "epoch": 0.2922023710135248,
+      "grad_norm": 0.6755229208947209,
+      "learning_rate": 9.733036707452727e-06,
+      "loss": 0.2545,
+      "step": 875
+    },
+    {
+      "epoch": 0.2925363165803974,
+      "grad_norm": 0.663707156055788,
+      "learning_rate": 9.744160177975528e-06,
+      "loss": 0.235,
+      "step": 876
+    },
+    {
+      "epoch": 0.29287026214727,
+      "grad_norm": 0.6679337491435877,
+      "learning_rate": 9.755283648498333e-06,
+      "loss": 0.2409,
+      "step": 877
+    },
+    {
+      "epoch": 0.2932042077141426,
+      "grad_norm": 0.8073669301897611,
+      "learning_rate": 9.766407119021135e-06,
+      "loss": 0.2554,
+      "step": 878
+    },
+    {
+      "epoch": 0.2935381532810152,
+      "grad_norm": 0.7198972500299666,
+      "learning_rate": 9.777530589543938e-06,
+      "loss": 0.2601,
+      "step": 879
+    },
+    {
+      "epoch": 0.2938720988478878,
+      "grad_norm": 0.7486411039843818,
+      "learning_rate": 9.788654060066741e-06,
+      "loss": 0.2717,
+      "step": 880
+    },
+    {
+      "epoch": 0.2942060444147604,
+      "grad_norm": 0.7554055472380966,
+      "learning_rate": 9.799777530589545e-06,
+      "loss": 0.2536,
+      "step": 881
+    },
+    {
+      "epoch": 0.294539989981633,
+      "grad_norm": 0.6769414341242086,
+      "learning_rate": 9.810901001112348e-06,
+      "loss": 0.2459,
+      "step": 882
+    },
+    {
+      "epoch": 0.2948739355485056,
+      "grad_norm": 0.6873429312658788,
+      "learning_rate": 9.82202447163515e-06,
+      "loss": 0.2353,
+      "step": 883
+    },
+    {
+      "epoch": 0.2952078811153782,
+      "grad_norm": 0.7752597910074831,
+      "learning_rate": 9.833147942157955e-06,
+      "loss": 0.2535,
+      "step": 884
+    },
+    {
+      "epoch": 0.2955418266822508,
+      "grad_norm": 0.6943618227859522,
+      "learning_rate": 9.844271412680758e-06,
+      "loss": 0.2689,
+      "step": 885
+    },
+    {
+      "epoch": 0.29587577224912337,
+      "grad_norm": 0.6801838220701963,
+      "learning_rate": 9.85539488320356e-06,
+      "loss": 0.2509,
+      "step": 886
+    },
+    {
+      "epoch": 0.296209717815996,
+      "grad_norm": 0.7339744312012771,
+      "learning_rate": 9.866518353726363e-06,
+      "loss": 0.2506,
+      "step": 887
+    },
+    {
+      "epoch": 0.2965436633828686,
+      "grad_norm": 0.7467491221803867,
+      "learning_rate": 9.877641824249166e-06,
+      "loss": 0.2566,
+      "step": 888
+    },
+    {
+      "epoch": 0.2968776089497412,
+      "grad_norm": 0.6412331225573681,
+      "learning_rate": 9.888765294771969e-06,
+      "loss": 0.2474,
+      "step": 889
+    },
+    {
+      "epoch": 0.2972115545166138,
+      "grad_norm": 0.7697158729899859,
+      "learning_rate": 9.899888765294773e-06,
+      "loss": 0.247,
+      "step": 890
+    },
+    {
+      "epoch": 0.2975455000834864,
+      "grad_norm": 0.7524145415460668,
+      "learning_rate": 9.911012235817576e-06,
+      "loss": 0.2567,
+      "step": 891
+    },
+    {
+      "epoch": 0.297879445650359,
+      "grad_norm": 0.6734182049066624,
+      "learning_rate": 9.922135706340378e-06,
+      "loss": 0.2549,
+      "step": 892
+    },
+    {
+      "epoch": 0.2982133912172316,
+      "grad_norm": 0.7218829257352565,
+      "learning_rate": 9.933259176863183e-06,
+      "loss": 0.2406,
+      "step": 893
+    },
+    {
+      "epoch": 0.29854733678410417,
+      "grad_norm": 0.6704757464416423,
+      "learning_rate": 9.944382647385986e-06,
+      "loss": 0.2441,
+      "step": 894
+    },
+    {
+      "epoch": 0.2988812823509768,
+      "grad_norm": 0.7404615308151165,
+      "learning_rate": 9.955506117908788e-06,
+      "loss": 0.2708,
+      "step": 895
+    },
+    {
+      "epoch": 0.2992152279178494,
+      "grad_norm": 0.7370776669079746,
+      "learning_rate": 9.966629588431591e-06,
+      "loss": 0.2554,
+      "step": 896
+    },
+    {
+      "epoch": 0.299549173484722,
+      "grad_norm": 0.7453068037060375,
+      "learning_rate": 9.977753058954395e-06,
+      "loss": 0.2522,
+      "step": 897
+    },
+    {
+      "epoch": 0.2998831190515946,
+      "grad_norm": 0.7537765886662524,
+      "learning_rate": 9.988876529477196e-06,
+      "loss": 0.2532,
+      "step": 898
+    },
+    {
+      "epoch": 0.3002170646184672,
+      "grad_norm": 0.7151572032080934,
+      "learning_rate": 1e-05,
+      "loss": 0.2534,
+      "step": 899
+    },
+    {
+      "epoch": 0.3005510101853398,
+      "grad_norm": 0.714477700987338,
+      "learning_rate": 9.999999622345564e-06,
+      "loss": 0.2404,
+      "step": 900
+    },
+    {
+      "epoch": 0.3008849557522124,
+      "grad_norm": 0.77324691413999,
+      "learning_rate": 9.999998489382312e-06,
+      "loss": 0.2426,
+      "step": 901
+    },
+    {
+      "epoch": 0.30121890131908496,
+      "grad_norm": 0.6128610447991886,
+      "learning_rate": 9.999996601110414e-06,
+      "loss": 0.2368,
+      "step": 902
+    },
+    {
+      "epoch": 0.30155284688595757,
+      "grad_norm": 0.8103821519751393,
+      "learning_rate": 9.999993957530157e-06,
+      "loss": 0.2486,
+      "step": 903
+    },
+    {
+      "epoch": 0.3018867924528302,
+      "grad_norm": 0.7509864144087558,
+      "learning_rate": 9.999990558641939e-06,
+      "loss": 0.2538,
+      "step": 904
+    },
+    {
+      "epoch": 0.3022207380197028,
+      "grad_norm": 0.6455618999464463,
+      "learning_rate": 9.999986404446276e-06,
+      "loss": 0.2622,
+      "step": 905
+    },
+    {
+      "epoch": 0.3025546835865754,
+      "grad_norm": 0.7642363495501507,
+      "learning_rate": 9.999981494943791e-06,
+      "loss": 0.2579,
+      "step": 906
+    },
+    {
+      "epoch": 0.302888629153448,
+      "grad_norm": 0.7516753216991904,
+      "learning_rate": 9.99997583013523e-06,
+      "loss": 0.2634,
+      "step": 907
+    },
+    {
+      "epoch": 0.3032225747203206,
+      "grad_norm": 0.5702441441339939,
+      "learning_rate": 9.999969410021447e-06,
+      "loss": 0.2212,
+      "step": 908
+    },
+    {
+      "epoch": 0.3035565202871932,
+      "grad_norm": 0.7489930131914292,
+      "learning_rate": 9.999962234603412e-06,
+      "loss": 0.245,
+      "step": 909
+    },
+    {
+      "epoch": 0.30389046585406576,
+      "grad_norm": 0.6082830635429192,
+      "learning_rate": 9.99995430388221e-06,
+      "loss": 0.2325,
+      "step": 910
+    },
+    {
+      "epoch": 0.30422441142093837,
+      "grad_norm": 0.620352857608966,
+      "learning_rate": 9.999945617859034e-06,
+      "loss": 0.2396,
+      "step": 911
+    },
+    {
+      "epoch": 0.304558356987811,
+      "grad_norm": 0.7221173476692679,
+      "learning_rate": 9.999936176535203e-06,
+      "loss": 0.2305,
+      "step": 912
+    },
+    {
+      "epoch": 0.3048923025546836,
+      "grad_norm": 0.8190610193603814,
+      "learning_rate": 9.99992597991214e-06,
+      "loss": 0.2528,
+      "step": 913
+    },
+    {
+      "epoch": 0.3052262481215562,
+      "grad_norm": 0.6278371889811301,
+      "learning_rate": 9.999915027991384e-06,
+      "loss": 0.227,
+      "step": 914
+    },
+    {
+      "epoch": 0.3055601936884288,
+      "grad_norm": 0.6759595522501595,
+      "learning_rate": 9.999903320774593e-06,
+      "loss": 0.2519,
+      "step": 915
+    },
+    {
+      "epoch": 0.3058941392553014,
+      "grad_norm": 0.6733098529721205,
+      "learning_rate": 9.999890858263532e-06,
+      "loss": 0.2299,
+      "step": 916
+    },
+    {
+      "epoch": 0.306228084822174,
+      "grad_norm": 0.6706970692688264,
+      "learning_rate": 9.999877640460085e-06,
+      "loss": 0.248,
+      "step": 917
+    },
+    {
+      "epoch": 0.30656203038904656,
+      "grad_norm": 0.7033323944048314,
+      "learning_rate": 9.999863667366249e-06,
+      "loss": 0.2528,
+      "step": 918
+    },
+    {
+      "epoch": 0.30689597595591916,
+      "grad_norm": 0.7882000097369829,
+      "learning_rate": 9.999848938984135e-06,
+      "loss": 0.2462,
+      "step": 919
+    },
+    {
+      "epoch": 0.30722992152279177,
+      "grad_norm": 0.663692017284394,
+      "learning_rate": 9.999833455315966e-06,
+      "loss": 0.2401,
+      "step": 920
+    },
+    {
+      "epoch": 0.3075638670896644,
+      "grad_norm": 0.7406267967602063,
+      "learning_rate": 9.999817216364085e-06,
+      "loss": 0.2426,
+      "step": 921
+    },
+    {
+      "epoch": 0.307897812656537,
+      "grad_norm": 0.719538796129936,
+      "learning_rate": 9.99980022213094e-06,
+      "loss": 0.259,
+      "step": 922
+    },
+    {
+      "epoch": 0.3082317582234096,
+      "grad_norm": 0.7198059211802099,
+      "learning_rate": 9.999782472619102e-06,
+      "loss": 0.2356,
+      "step": 923
+    },
+    {
+      "epoch": 0.3085657037902822,
+      "grad_norm": 0.7748449854954533,
+      "learning_rate": 9.99976396783125e-06,
+      "loss": 0.2512,
+      "step": 924
+    },
+    {
+      "epoch": 0.3088996493571548,
+      "grad_norm": 0.6743289212320538,
+      "learning_rate": 9.999744707770182e-06,
+      "loss": 0.2295,
+      "step": 925
+    },
+    {
+      "epoch": 0.3092335949240274,
+      "grad_norm": 0.8585415752543676,
+      "learning_rate": 9.999724692438805e-06,
+      "loss": 0.2855,
+      "step": 926
+    },
+    {
+      "epoch": 0.30956754049089996,
+      "grad_norm": 0.6469196243149617,
+      "learning_rate": 9.999703921840143e-06,
+      "loss": 0.2527,
+      "step": 927
+    },
+    {
+      "epoch": 0.30990148605777257,
+      "grad_norm": 0.6798611050791047,
+      "learning_rate": 9.999682395977334e-06,
+      "loss": 0.2469,
+      "step": 928
+    },
+    {
+      "epoch": 0.3102354316246452,
+      "grad_norm": 0.7112677389326599,
+      "learning_rate": 9.999660114853631e-06,
+      "loss": 0.243,
+      "step": 929
+    },
+    {
+      "epoch": 0.3105693771915178,
+      "grad_norm": 0.5521283993542538,
+      "learning_rate": 9.999637078472398e-06,
+      "loss": 0.2228,
+      "step": 930
+    },
+    {
+      "epoch": 0.3109033227583904,
+      "grad_norm": 0.6084209475663657,
+      "learning_rate": 9.999613286837115e-06,
+      "loss": 0.2429,
+      "step": 931
+    },
+    {
+      "epoch": 0.311237268325263,
+      "grad_norm": 0.6229978352267757,
+      "learning_rate": 9.999588739951376e-06,
+      "loss": 0.2676,
+      "step": 932
+    },
+    {
+      "epoch": 0.3115712138921356,
+      "grad_norm": 0.6220393849687876,
+      "learning_rate": 9.99956343781889e-06,
+      "loss": 0.2396,
+      "step": 933
+    },
+    {
+      "epoch": 0.3119051594590082,
+      "grad_norm": 0.6661414030839978,
+      "learning_rate": 9.999537380443479e-06,
+      "loss": 0.2464,
+      "step": 934
+    },
+    {
+      "epoch": 0.31223910502588076,
+      "grad_norm": 0.657272661420604,
+      "learning_rate": 9.999510567829079e-06,
+      "loss": 0.2318,
+      "step": 935
+    },
+    {
+      "epoch": 0.31257305059275337,
+      "grad_norm": 0.6663711093096606,
+      "learning_rate": 9.999482999979739e-06,
+      "loss": 0.2738,
+      "step": 936
+    },
+    {
+      "epoch": 0.31290699615962597,
+      "grad_norm": 1.3839502383145643,
+      "learning_rate": 9.999454676899628e-06,
+      "loss": 0.2663,
+      "step": 937
+    },
+    {
+      "epoch": 0.3132409417264986,
+      "grad_norm": 0.7046860437125939,
+      "learning_rate": 9.999425598593018e-06,
+      "loss": 0.2601,
+      "step": 938
+    },
+    {
+      "epoch": 0.3135748872933712,
+      "grad_norm": 0.7271393767781734,
+      "learning_rate": 9.999395765064308e-06,
+      "loss": 0.2669,
+      "step": 939
+    },
+    {
+      "epoch": 0.3139088328602438,
+      "grad_norm": 0.6277865590931718,
+      "learning_rate": 9.999365176318e-06,
+      "loss": 0.24,
+      "step": 940
+    },
+    {
+      "epoch": 0.3142427784271164,
+      "grad_norm": 0.7623469016255324,
+      "learning_rate": 9.999333832358716e-06,
+      "loss": 0.2474,
+      "step": 941
+    },
+    {
+      "epoch": 0.314576723993989,
+      "grad_norm": 0.6244292989040466,
+      "learning_rate": 9.999301733191193e-06,
+      "loss": 0.2476,
+      "step": 942
+    },
+    {
+      "epoch": 0.31491066956086156,
+      "grad_norm": 0.6679513544214297,
+      "learning_rate": 9.999268878820278e-06,
+      "loss": 0.2566,
+      "step": 943
+    },
+    {
+      "epoch": 0.31524461512773416,
+      "grad_norm": 0.9118400082357937,
+      "learning_rate": 9.999235269250933e-06,
+      "loss": 0.2472,
+      "step": 944
+    },
+    {
+      "epoch": 0.31557856069460677,
+      "grad_norm": 0.5506843638558178,
+      "learning_rate": 9.999200904488238e-06,
+      "loss": 0.2239,
+      "step": 945
+    },
+    {
+      "epoch": 0.3159125062614794,
+      "grad_norm": 0.6977183679119197,
+      "learning_rate": 9.999165784537381e-06,
+      "loss": 0.2522,
+      "step": 946
+    },
+    {
+      "epoch": 0.316246451828352,
+      "grad_norm": 0.6229267335231975,
+      "learning_rate": 9.999129909403671e-06,
+      "loss": 0.2385,
+      "step": 947
+    },
+    {
+      "epoch": 0.3165803973952246,
+      "grad_norm": 0.6346107999834576,
+      "learning_rate": 9.999093279092524e-06,
+      "loss": 0.2464,
+      "step": 948
+    },
+    {
+      "epoch": 0.3169143429620972,
+      "grad_norm": 0.7147362148938331,
+      "learning_rate": 9.999055893609475e-06,
+      "loss": 0.2519,
+      "step": 949
+    },
+    {
+      "epoch": 0.3172482885289698,
+      "grad_norm": 0.681378538809813,
+      "learning_rate": 9.999017752960172e-06,
+      "loss": 0.2497,
+      "step": 950
+    },
+    {
+      "epoch": 0.31758223409584235,
+      "grad_norm": 0.7257313818749964,
+      "learning_rate": 9.998978857150375e-06,
+      "loss": 0.2586,
+      "step": 951
+    },
+    {
+      "epoch": 0.31791617966271496,
+      "grad_norm": 0.6552438164802247,
+      "learning_rate": 9.99893920618596e-06,
+      "loss": 0.2486,
+      "step": 952
+    },
+    {
+      "epoch": 0.31825012522958757,
+      "grad_norm": 0.7499319030931288,
+      "learning_rate": 9.998898800072919e-06,
+      "loss": 0.2519,
+      "step": 953
+    },
+    {
+      "epoch": 0.3185840707964602,
+      "grad_norm": 0.623600497254021,
+      "learning_rate": 9.998857638817354e-06,
+      "loss": 0.2435,
+      "step": 954
+    },
+    {
+      "epoch": 0.3189180163633328,
+      "grad_norm": 0.7248155860321855,
+      "learning_rate": 9.99881572242548e-06,
+      "loss": 0.2376,
+      "step": 955
+    },
+    {
+      "epoch": 0.3192519619302054,
+      "grad_norm": 0.5991586377037108,
+      "learning_rate": 9.998773050903637e-06,
+      "loss": 0.2376,
+      "step": 956
+    },
+    {
+      "epoch": 0.319585907497078,
+      "grad_norm": 0.7346806017257687,
+      "learning_rate": 9.998729624258262e-06,
+      "loss": 0.2633,
+      "step": 957
+    },
+    {
+      "epoch": 0.3199198530639506,
+      "grad_norm": 0.563428634742872,
+      "learning_rate": 9.998685442495921e-06,
+      "loss": 0.2382,
+      "step": 958
+    },
+    {
+      "epoch": 0.32025379863082315,
+      "grad_norm": 0.6736707339352784,
+      "learning_rate": 9.998640505623284e-06,
+      "loss": 0.2523,
+      "step": 959
+    },
+    {
+      "epoch": 0.32058774419769576,
+      "grad_norm": 0.6334538052099089,
+      "learning_rate": 9.998594813647145e-06,
+      "loss": 0.2382,
+      "step": 960
+    },
+    {
+      "epoch": 0.32092168976456836,
+      "grad_norm": 0.6355713277943253,
+      "learning_rate": 9.998548366574401e-06,
+      "loss": 0.245,
+      "step": 961
+    },
+    {
+      "epoch": 0.32125563533144097,
+      "grad_norm": 0.5993057051640902,
+      "learning_rate": 9.99850116441207e-06,
+      "loss": 0.2418,
+      "step": 962
+    },
+    {
+      "epoch": 0.3215895808983136,
+      "grad_norm": 0.5920496868805775,
+      "learning_rate": 9.998453207167282e-06,
+      "loss": 0.2565,
+      "step": 963
+    },
+    {
+      "epoch": 0.3219235264651862,
+      "grad_norm": 0.61198413975063,
+      "learning_rate": 9.998404494847285e-06,
+      "loss": 0.2585,
+      "step": 964
+    },
+    {
+      "epoch": 0.3222574720320588,
+      "grad_norm": 0.6369440770274445,
+      "learning_rate": 9.998355027459432e-06,
+      "loss": 0.2706,
+      "step": 965
+    },
+    {
+      "epoch": 0.3225914175989314,
+      "grad_norm": 0.5872473237923148,
+      "learning_rate": 9.998304805011199e-06,
+      "loss": 0.2486,
+      "step": 966
+    },
+    {
+      "epoch": 0.32292536316580395,
+      "grad_norm": 0.620263995911841,
+      "learning_rate": 9.998253827510173e-06,
+      "loss": 0.2547,
+      "step": 967
+    },
+    {
+      "epoch": 0.32325930873267655,
+      "grad_norm": 0.7511215575831635,
+      "learning_rate": 9.998202094964053e-06,
+      "loss": 0.2556,
+      "step": 968
+    },
+    {
+      "epoch": 0.32359325429954916,
+      "grad_norm": 0.6305888434992816,
+      "learning_rate": 9.998149607380654e-06,
+      "loss": 0.2356,
+      "step": 969
+    },
+    {
+      "epoch": 0.32392719986642177,
+      "grad_norm": 0.6474250636445457,
+      "learning_rate": 9.998096364767906e-06,
+      "loss": 0.2409,
+      "step": 970
+    },
+    {
+      "epoch": 0.3242611454332944,
+      "grad_norm": 0.6404737722389929,
+      "learning_rate": 9.998042367133854e-06,
+      "loss": 0.2418,
+      "step": 971
+    },
+    {
+      "epoch": 0.324595091000167,
+      "grad_norm": 0.6255070104433448,
+      "learning_rate": 9.997987614486648e-06,
+      "loss": 0.2421,
+      "step": 972
+    },
+    {
+      "epoch": 0.3249290365670396,
+      "grad_norm": 0.5688015945545797,
+      "learning_rate": 9.997932106834567e-06,
+      "loss": 0.2349,
+      "step": 973
+    },
+    {
+      "epoch": 0.3252629821339122,
+      "grad_norm": 0.6829656370856075,
+      "learning_rate": 9.997875844185991e-06,
+      "loss": 0.2501,
+      "step": 974
+    },
+    {
+      "epoch": 0.3255969277007848,
+      "grad_norm": 0.6365251336709236,
+      "learning_rate": 9.99781882654942e-06,
+      "loss": 0.2408,
+      "step": 975
+    },
+    {
+      "epoch": 0.32593087326765735,
+      "grad_norm": 0.6327216943145679,
+      "learning_rate": 9.997761053933469e-06,
+      "loss": 0.2532,
+      "step": 976
+    },
+    {
+      "epoch": 0.32626481883452996,
+      "grad_norm": 0.5741706691051712,
+      "learning_rate": 9.997702526346864e-06,
+      "loss": 0.2517,
+      "step": 977
+    },
+    {
+      "epoch": 0.32659876440140256,
+      "grad_norm": 0.634532616958001,
+      "learning_rate": 9.997643243798446e-06,
+      "loss": 0.2561,
+      "step": 978
+    },
+    {
+      "epoch": 0.32693270996827517,
+      "grad_norm": 0.6627572201424764,
+      "learning_rate": 9.99758320629717e-06,
+      "loss": 0.2212,
+      "step": 979
+    },
+    {
+      "epoch": 0.3272666555351478,
+      "grad_norm": 0.6562948613513276,
+      "learning_rate": 9.997522413852108e-06,
+      "loss": 0.2503,
+      "step": 980
+    },
+    {
+      "epoch": 0.3276006011020204,
+      "grad_norm": 0.7777578566787204,
+      "learning_rate": 9.997460866472439e-06,
+      "loss": 0.2431,
+      "step": 981
+    },
+    {
+      "epoch": 0.327934546668893,
+      "grad_norm": 0.7143909349751347,
+      "learning_rate": 9.997398564167465e-06,
+      "loss": 0.261,
+      "step": 982
+    },
+    {
+      "epoch": 0.3282684922357656,
+      "grad_norm": 0.6511654646356778,
+      "learning_rate": 9.997335506946596e-06,
+      "loss": 0.2605,
+      "step": 983
+    },
+    {
+      "epoch": 0.32860243780263815,
+      "grad_norm": 0.8594408758077905,
+      "learning_rate": 9.997271694819354e-06,
+      "loss": 0.2563,
+      "step": 984
+    },
+    {
+      "epoch": 0.32893638336951075,
+      "grad_norm": 0.630895825476141,
+      "learning_rate": 9.997207127795383e-06,
+      "loss": 0.2357,
+      "step": 985
+    },
+    {
+      "epoch": 0.32927032893638336,
+      "grad_norm": 0.6193761330050703,
+      "learning_rate": 9.997141805884436e-06,
+      "loss": 0.2395,
+      "step": 986
+    },
+    {
+      "epoch": 0.32960427450325597,
+      "grad_norm": 0.611311370706388,
+      "learning_rate": 9.997075729096379e-06,
+      "loss": 0.2506,
+      "step": 987
+    },
+    {
+      "epoch": 0.3299382200701286,
+      "grad_norm": 0.7033542886635047,
+      "learning_rate": 9.997008897441194e-06,
+      "loss": 0.2526,
+      "step": 988
+    },
+    {
+      "epoch": 0.3302721656370012,
+      "grad_norm": 0.6230286822204218,
+      "learning_rate": 9.996941310928978e-06,
+      "loss": 0.2395,
+      "step": 989
+    },
+    {
+      "epoch": 0.3306061112038738,
+      "grad_norm": 0.6362704854163252,
+      "learning_rate": 9.99687296956994e-06,
+      "loss": 0.2378,
+      "step": 990
+    },
+    {
+      "epoch": 0.3309400567707464,
+      "grad_norm": 0.5800914414032446,
+      "learning_rate": 9.996803873374402e-06,
+      "loss": 0.2254,
+      "step": 991
+    },
+    {
+      "epoch": 0.33127400233761894,
+      "grad_norm": 0.7204698593726746,
+      "learning_rate": 9.996734022352805e-06,
+      "loss": 0.2527,
+      "step": 992
+    },
+    {
+      "epoch": 0.33160794790449155,
+      "grad_norm": 0.634627877169416,
+      "learning_rate": 9.9966634165157e-06,
+      "loss": 0.2449,
+      "step": 993
+    },
+    {
+      "epoch": 0.33194189347136416,
+      "grad_norm": 0.6111123727859595,
+      "learning_rate": 9.99659205587375e-06,
+      "loss": 0.2474,
+      "step": 994
+    },
+    {
+      "epoch": 0.33227583903823676,
+      "grad_norm": 0.6755862324184183,
+      "learning_rate": 9.996519940437737e-06,
+      "loss": 0.2505,
+      "step": 995
+    },
+    {
+      "epoch": 0.33260978460510937,
+      "grad_norm": 0.6460412308879759,
+      "learning_rate": 9.996447070218557e-06,
+      "loss": 0.2473,
+      "step": 996
+    },
+    {
+      "epoch": 0.332943730171982,
+      "grad_norm": 0.6346454236316136,
+      "learning_rate": 9.996373445227215e-06,
+      "loss": 0.239,
+      "step": 997
+    },
+    {
+      "epoch": 0.3332776757388546,
+      "grad_norm": 0.6415377886338729,
+      "learning_rate": 9.996299065474832e-06,
+      "loss": 0.2445,
+      "step": 998
+    },
+    {
+      "epoch": 0.3336116213057272,
+      "grad_norm": 0.6555194675483528,
+      "learning_rate": 9.996223930972649e-06,
+      "loss": 0.2401,
+      "step": 999
+    },
+    {
+      "epoch": 0.33394556687259974,
+      "grad_norm": 0.5713457684437109,
+      "learning_rate": 9.99614804173201e-06,
+      "loss": 0.2244,
+      "step": 1000
+    },
+    {
+      "epoch": 0.33427951243947235,
+      "grad_norm": 0.5893092973518516,
+      "learning_rate": 9.996071397764381e-06,
+      "loss": 0.2602,
+      "step": 1001
+    },
+    {
+      "epoch": 0.33461345800634495,
+      "grad_norm": 0.5906004788656943,
+      "learning_rate": 9.995993999081343e-06,
+      "loss": 0.2298,
+      "step": 1002
+    },
+    {
+      "epoch": 0.33494740357321756,
+      "grad_norm": 0.6462480077770343,
+      "learning_rate": 9.995915845694584e-06,
+      "loss": 0.2573,
+      "step": 1003
+    },
+    {
+      "epoch": 0.33528134914009017,
+      "grad_norm": 0.9424314763458541,
+      "learning_rate": 9.995836937615913e-06,
+      "loss": 0.2275,
+      "step": 1004
+    },
+    {
+      "epoch": 0.3356152947069628,
+      "grad_norm": 0.5975148522005419,
+      "learning_rate": 9.995757274857246e-06,
+      "loss": 0.252,
+      "step": 1005
+    },
+    {
+      "epoch": 0.3359492402738354,
+      "grad_norm": 0.6746114929233912,
+      "learning_rate": 9.995676857430621e-06,
+      "loss": 0.2296,
+      "step": 1006
+    },
+    {
+      "epoch": 0.336283185840708,
+      "grad_norm": 0.6043959986784809,
+      "learning_rate": 9.995595685348186e-06,
+      "loss": 0.2418,
+      "step": 1007
+    },
+    {
+      "epoch": 0.33661713140758054,
+      "grad_norm": 0.6561996117107064,
+      "learning_rate": 9.995513758622198e-06,
+      "loss": 0.2411,
+      "step": 1008
+    },
+    {
+      "epoch": 0.33695107697445315,
+      "grad_norm": 0.5768734030397459,
+      "learning_rate": 9.995431077265038e-06,
+      "loss": 0.2397,
+      "step": 1009
+    },
+    {
+      "epoch": 0.33728502254132575,
+      "grad_norm": 0.5644019192557386,
+      "learning_rate": 9.995347641289194e-06,
+      "loss": 0.2308,
+      "step": 1010
+    },
+    {
+      "epoch": 0.33761896810819836,
+      "grad_norm": 0.7155920321425914,
+      "learning_rate": 9.995263450707273e-06,
+      "loss": 0.2483,
+      "step": 1011
+    },
+    {
+      "epoch": 0.33795291367507097,
+      "grad_norm": 0.5820377404778101,
+      "learning_rate": 9.995178505531989e-06,
+      "loss": 0.2361,
+      "step": 1012
+    },
+    {
+      "epoch": 0.33828685924194357,
+      "grad_norm": 0.6074649479328393,
+      "learning_rate": 9.995092805776175e-06,
+      "loss": 0.2501,
+      "step": 1013
+    },
+    {
+      "epoch": 0.3386208048088162,
+      "grad_norm": 0.6808463616772454,
+      "learning_rate": 9.995006351452775e-06,
+      "loss": 0.2514,
+      "step": 1014
+    },
+    {
+      "epoch": 0.3389547503756888,
+      "grad_norm": 0.5357571508833467,
+      "learning_rate": 9.994919142574854e-06,
+      "loss": 0.2361,
+      "step": 1015
+    },
+    {
+      "epoch": 0.33928869594256134,
+      "grad_norm": 0.6020487192909576,
+      "learning_rate": 9.994831179155584e-06,
+      "loss": 0.2419,
+      "step": 1016
+    },
+    {
+      "epoch": 0.33962264150943394,
+      "grad_norm": 0.7283194863543118,
+      "learning_rate": 9.994742461208251e-06,
+      "loss": 0.2607,
+      "step": 1017
+    },
+    {
+      "epoch": 0.33995658707630655,
+      "grad_norm": 0.5782631325584849,
+      "learning_rate": 9.994652988746258e-06,
+      "loss": 0.238,
+      "step": 1018
+    },
+    {
+      "epoch": 0.34029053264317916,
+      "grad_norm": 0.7864013734784921,
+      "learning_rate": 9.994562761783122e-06,
+      "loss": 0.2497,
+      "step": 1019
+    },
+    {
+      "epoch": 0.34062447821005176,
+      "grad_norm": 0.6293240854325619,
+      "learning_rate": 9.99447178033247e-06,
+      "loss": 0.2423,
+      "step": 1020
+    },
+    {
+      "epoch": 0.34095842377692437,
+      "grad_norm": 0.6532327987330085,
+      "learning_rate": 9.99438004440805e-06,
+      "loss": 0.2511,
+      "step": 1021
+    },
+    {
+      "epoch": 0.341292369343797,
+      "grad_norm": 0.5993467036575328,
+      "learning_rate": 9.994287554023717e-06,
+      "loss": 0.2241,
+      "step": 1022
+    },
+    {
+      "epoch": 0.3416263149106696,
+      "grad_norm": 0.6048208268550589,
+      "learning_rate": 9.994194309193442e-06,
+      "loss": 0.2263,
+      "step": 1023
+    },
+    {
+      "epoch": 0.3419602604775422,
+      "grad_norm": 0.620928708879154,
+      "learning_rate": 9.99410030993131e-06,
+      "loss": 0.2333,
+      "step": 1024
+    },
+    {
+      "epoch": 0.34229420604441474,
+      "grad_norm": 0.5811126029360281,
+      "learning_rate": 9.994005556251525e-06,
+      "loss": 0.2369,
+      "step": 1025
+    },
+    {
+      "epoch": 0.34262815161128735,
+      "grad_norm": 0.6238825923972041,
+      "learning_rate": 9.993910048168399e-06,
+      "loss": 0.2273,
+      "step": 1026
+    },
+    {
+      "epoch": 0.34296209717815995,
+      "grad_norm": 0.6765552725580103,
+      "learning_rate": 9.993813785696355e-06,
+      "loss": 0.2254,
+      "step": 1027
+    },
+    {
+      "epoch": 0.34329604274503256,
+      "grad_norm": 0.6187685905117865,
+      "learning_rate": 9.993716768849942e-06,
+      "loss": 0.2408,
+      "step": 1028
+    },
+    {
+      "epoch": 0.34362998831190517,
+      "grad_norm": 0.6097073883952683,
+      "learning_rate": 9.99361899764381e-06,
+      "loss": 0.25,
+      "step": 1029
+    },
+    {
+      "epoch": 0.3439639338787778,
+      "grad_norm": 0.6545748380846143,
+      "learning_rate": 9.993520472092732e-06,
+      "loss": 0.2434,
+      "step": 1030
+    },
+    {
+      "epoch": 0.3442978794456504,
+      "grad_norm": 0.5477963426981712,
+      "learning_rate": 9.99342119221159e-06,
+      "loss": 0.2346,
+      "step": 1031
+    },
+    {
+      "epoch": 0.344631825012523,
+      "grad_norm": 0.5510391995222212,
+      "learning_rate": 9.993321158015379e-06,
+      "loss": 0.2305,
+      "step": 1032
+    },
+    {
+      "epoch": 0.34496577057939554,
+      "grad_norm": 0.6324909697234063,
+      "learning_rate": 9.993220369519215e-06,
+      "loss": 0.2467,
+      "step": 1033
+    },
+    {
+      "epoch": 0.34529971614626814,
+      "grad_norm": 0.6142047956027233,
+      "learning_rate": 9.99311882673832e-06,
+      "loss": 0.2333,
+      "step": 1034
+    },
+    {
+      "epoch": 0.34563366171314075,
+      "grad_norm": 0.6555126781240093,
+      "learning_rate": 9.993016529688033e-06,
+      "loss": 0.2383,
+      "step": 1035
+    },
+    {
+      "epoch": 0.34596760728001336,
+      "grad_norm": 0.6360668156570694,
+      "learning_rate": 9.99291347838381e-06,
+      "loss": 0.2418,
+      "step": 1036
+    },
+    {
+      "epoch": 0.34630155284688596,
+      "grad_norm": 0.6241515101912609,
+      "learning_rate": 9.992809672841218e-06,
+      "loss": 0.2362,
+      "step": 1037
+    },
+    {
+      "epoch": 0.34663549841375857,
+      "grad_norm": 0.6300889467982095,
+      "learning_rate": 9.992705113075933e-06,
+      "loss": 0.237,
+      "step": 1038
+    },
+    {
+      "epoch": 0.3469694439806312,
+      "grad_norm": 0.564235622274707,
+      "learning_rate": 9.992599799103754e-06,
+      "loss": 0.2354,
+      "step": 1039
+    },
+    {
+      "epoch": 0.3473033895475038,
+      "grad_norm": 0.6842460803144252,
+      "learning_rate": 9.99249373094059e-06,
+      "loss": 0.2416,
+      "step": 1040
+    },
+    {
+      "epoch": 0.34763733511437633,
+      "grad_norm": 0.7151896495749245,
+      "learning_rate": 9.992386908602466e-06,
+      "loss": 0.2572,
+      "step": 1041
+    },
+    {
+      "epoch": 0.34797128068124894,
+      "grad_norm": 0.5719922134860118,
+      "learning_rate": 9.992279332105512e-06,
+      "loss": 0.2325,
+      "step": 1042
+    },
+    {
+      "epoch": 0.34830522624812155,
+      "grad_norm": 0.7231166488003365,
+      "learning_rate": 9.992171001465985e-06,
+      "loss": 0.2549,
+      "step": 1043
+    },
+    {
+      "epoch": 0.34863917181499415,
+      "grad_norm": 0.592629753373637,
+      "learning_rate": 9.992061916700247e-06,
+      "loss": 0.235,
+      "step": 1044
+    },
+    {
+      "epoch": 0.34897311738186676,
+      "grad_norm": 0.6764193734512775,
+      "learning_rate": 9.991952077824776e-06,
+      "loss": 0.2413,
+      "step": 1045
+    },
+    {
+      "epoch": 0.34930706294873937,
+      "grad_norm": 0.7115915690879145,
+      "learning_rate": 9.991841484856166e-06,
+      "loss": 0.2521,
+      "step": 1046
+    },
+    {
+      "epoch": 0.349641008515612,
+      "grad_norm": 0.5797754264927091,
+      "learning_rate": 9.991730137811122e-06,
+      "loss": 0.2476,
+      "step": 1047
+    },
+    {
+      "epoch": 0.3499749540824846,
+      "grad_norm": 0.6078645313605627,
+      "learning_rate": 9.991618036706464e-06,
+      "loss": 0.2429,
+      "step": 1048
+    },
+    {
+      "epoch": 0.35030889964935713,
+      "grad_norm": 0.810806694264766,
+      "learning_rate": 9.99150518155913e-06,
+      "loss": 0.2475,
+      "step": 1049
+    },
+    {
+      "epoch": 0.35064284521622974,
+      "grad_norm": 0.7023437648349548,
+      "learning_rate": 9.991391572386162e-06,
+      "loss": 0.2565,
+      "step": 1050
+    },
+    {
+      "epoch": 0.35097679078310234,
+      "grad_norm": 0.7282652300633451,
+      "learning_rate": 9.991277209204728e-06,
+      "loss": 0.244,
+      "step": 1051
+    },
+    {
+      "epoch": 0.35131073634997495,
+      "grad_norm": 0.6241402490643777,
+      "learning_rate": 9.991162092032101e-06,
+      "loss": 0.2614,
+      "step": 1052
+    },
+    {
+      "epoch": 0.35164468191684756,
+      "grad_norm": 0.6258752658980473,
+      "learning_rate": 9.99104622088567e-06,
+      "loss": 0.2461,
+      "step": 1053
+    },
+    {
+      "epoch": 0.35197862748372016,
+      "grad_norm": 0.6658730176732505,
+      "learning_rate": 9.990929595782938e-06,
+      "loss": 0.2402,
+      "step": 1054
+    },
+    {
+      "epoch": 0.35231257305059277,
+      "grad_norm": 0.5923285109590797,
+      "learning_rate": 9.990812216741529e-06,
+      "loss": 0.2275,
+      "step": 1055
+    },
+    {
+      "epoch": 0.3526465186174654,
+      "grad_norm": 0.6257275221441978,
+      "learning_rate": 9.990694083779166e-06,
+      "loss": 0.2396,
+      "step": 1056
+    },
+    {
+      "epoch": 0.3529804641843379,
+      "grad_norm": 0.6557840979079784,
+      "learning_rate": 9.990575196913699e-06,
+      "loss": 0.2337,
+      "step": 1057
+    },
+    {
+      "epoch": 0.35331440975121053,
+      "grad_norm": 0.7708982420567367,
+      "learning_rate": 9.990455556163086e-06,
+      "loss": 0.251,
+      "step": 1058
+    },
+    {
+      "epoch": 0.35364835531808314,
+      "grad_norm": 0.7583771899024517,
+      "learning_rate": 9.990335161545401e-06,
+      "loss": 0.2584,
+      "step": 1059
+    },
+    {
+      "epoch": 0.35398230088495575,
+      "grad_norm": 0.6318386182926147,
+      "learning_rate": 9.99021401307883e-06,
+      "loss": 0.2519,
+      "step": 1060
+    },
+    {
+      "epoch": 0.35431624645182835,
+      "grad_norm": 0.769122863891848,
+      "learning_rate": 9.990092110781675e-06,
+      "loss": 0.2404,
+      "step": 1061
+    },
+    {
+      "epoch": 0.35465019201870096,
+      "grad_norm": 0.6258454284206358,
+      "learning_rate": 9.98996945467235e-06,
+      "loss": 0.2406,
+      "step": 1062
+    },
+    {
+      "epoch": 0.35498413758557357,
+      "grad_norm": 0.5571971449891453,
+      "learning_rate": 9.989846044769384e-06,
+      "loss": 0.2342,
+      "step": 1063
+    },
+    {
+      "epoch": 0.3553180831524462,
+      "grad_norm": 0.6702068647047834,
+      "learning_rate": 9.98972188109142e-06,
+      "loss": 0.2532,
+      "step": 1064
+    },
+    {
+      "epoch": 0.3556520287193187,
+      "grad_norm": 0.709957113892022,
+      "learning_rate": 9.989596963657213e-06,
+      "loss": 0.2501,
+      "step": 1065
+    },
+    {
+      "epoch": 0.35598597428619133,
+      "grad_norm": 0.7180875116372593,
+      "learning_rate": 9.989471292485636e-06,
+      "loss": 0.2767,
+      "step": 1066
+    },
+    {
+      "epoch": 0.35631991985306394,
+      "grad_norm": 0.6068909083103505,
+      "learning_rate": 9.989344867595668e-06,
+      "loss": 0.2425,
+      "step": 1067
+    },
+    {
+      "epoch": 0.35665386541993654,
+      "grad_norm": 0.7137465058277483,
+      "learning_rate": 9.989217689006412e-06,
+      "loss": 0.2531,
+      "step": 1068
+    },
+    {
+      "epoch": 0.35698781098680915,
+      "grad_norm": 0.6022049439905831,
+      "learning_rate": 9.989089756737077e-06,
+      "loss": 0.2427,
+      "step": 1069
+    },
+    {
+      "epoch": 0.35732175655368176,
+      "grad_norm": 0.6083150772695914,
+      "learning_rate": 9.988961070806991e-06,
+      "loss": 0.248,
+      "step": 1070
+    },
+    {
+      "epoch": 0.35765570212055436,
+      "grad_norm": 0.6471745928158773,
+      "learning_rate": 9.988831631235591e-06,
+      "loss": 0.238,
+      "step": 1071
+    },
+    {
+      "epoch": 0.35798964768742697,
+      "grad_norm": 0.5615061566854802,
+      "learning_rate": 9.98870143804243e-06,
+      "loss": 0.2149,
+      "step": 1072
+    },
+    {
+      "epoch": 0.3583235932542995,
+      "grad_norm": 0.6199663578980865,
+      "learning_rate": 9.988570491247179e-06,
+      "loss": 0.2493,
+      "step": 1073
+    },
+    {
+      "epoch": 0.35865753882117213,
+      "grad_norm": 0.6728065943159997,
+      "learning_rate": 9.988438790869616e-06,
+      "loss": 0.2547,
+      "step": 1074
+    },
+    {
+      "epoch": 0.35899148438804473,
+      "grad_norm": 0.7194963403788357,
+      "learning_rate": 9.988306336929637e-06,
+      "loss": 0.2482,
+      "step": 1075
+    },
+    {
+      "epoch": 0.35932542995491734,
+      "grad_norm": 0.6633897906991787,
+      "learning_rate": 9.988173129447251e-06,
+      "loss": 0.257,
+      "step": 1076
+    },
+    {
+      "epoch": 0.35965937552178995,
+      "grad_norm": 0.6374583029418619,
+      "learning_rate": 9.98803916844258e-06,
+      "loss": 0.2498,
+      "step": 1077
+    },
+    {
+      "epoch": 0.35999332108866255,
+      "grad_norm": 0.5644829103003917,
+      "learning_rate": 9.98790445393586e-06,
+      "loss": 0.2219,
+      "step": 1078
+    },
+    {
+      "epoch": 0.36032726665553516,
+      "grad_norm": 0.7465815493447656,
+      "learning_rate": 9.98776898594744e-06,
+      "loss": 0.2516,
+      "step": 1079
+    },
+    {
+      "epoch": 0.36066121222240777,
+      "grad_norm": 0.6137141055718743,
+      "learning_rate": 9.987632764497787e-06,
+      "loss": 0.2307,
+      "step": 1080
+    },
+    {
+      "epoch": 0.3609951577892804,
+      "grad_norm": 0.6606373030819324,
+      "learning_rate": 9.987495789607478e-06,
+      "loss": 0.2372,
+      "step": 1081
+    },
+    {
+      "epoch": 0.3613291033561529,
+      "grad_norm": 0.6144078594991942,
+      "learning_rate": 9.987358061297203e-06,
+      "loss": 0.2552,
+      "step": 1082
+    },
+    {
+      "epoch": 0.36166304892302553,
+      "grad_norm": 0.6034573546156843,
+      "learning_rate": 9.987219579587768e-06,
+      "loss": 0.2271,
+      "step": 1083
+    },
+    {
+      "epoch": 0.36199699448989814,
+      "grad_norm": 0.6365200936320912,
+      "learning_rate": 9.987080344500094e-06,
+      "loss": 0.2305,
+      "step": 1084
+    },
+    {
+      "epoch": 0.36233094005677075,
+      "grad_norm": 0.6381373323927708,
+      "learning_rate": 9.986940356055212e-06,
+      "loss": 0.2371,
+      "step": 1085
+    },
+    {
+      "epoch": 0.36266488562364335,
+      "grad_norm": 0.612493000628613,
+      "learning_rate": 9.986799614274271e-06,
+      "loss": 0.238,
+      "step": 1086
+    },
+    {
+      "epoch": 0.36299883119051596,
+      "grad_norm": 0.6157534327216453,
+      "learning_rate": 9.986658119178532e-06,
+      "loss": 0.2597,
+      "step": 1087
+    },
+    {
+      "epoch": 0.36333277675738856,
+      "grad_norm": 0.6729028218692379,
+      "learning_rate": 9.986515870789366e-06,
+      "loss": 0.2415,
+      "step": 1088
+    },
+    {
+      "epoch": 0.36366672232426117,
+      "grad_norm": 0.6283943072436228,
+      "learning_rate": 9.986372869128264e-06,
+      "loss": 0.2563,
+      "step": 1089
+    },
+    {
+      "epoch": 0.3640006678911337,
+      "grad_norm": 0.6576049042218921,
+      "learning_rate": 9.986229114216828e-06,
+      "loss": 0.2427,
+      "step": 1090
+    },
+    {
+      "epoch": 0.36433461345800633,
+      "grad_norm": 0.5841024132825844,
+      "learning_rate": 9.986084606076772e-06,
+      "loss": 0.242,
+      "step": 1091
+    },
+    {
+      "epoch": 0.36466855902487894,
+      "grad_norm": 0.5877810096206579,
+      "learning_rate": 9.985939344729926e-06,
+      "loss": 0.2214,
+      "step": 1092
+    },
+    {
+      "epoch": 0.36500250459175154,
+      "grad_norm": 0.590923371403508,
+      "learning_rate": 9.985793330198237e-06,
+      "loss": 0.2376,
+      "step": 1093
+    },
+    {
+      "epoch": 0.36533645015862415,
+      "grad_norm": 0.5462432642092312,
+      "learning_rate": 9.98564656250376e-06,
+      "loss": 0.2239,
+      "step": 1094
+    },
+    {
+      "epoch": 0.36567039572549676,
+      "grad_norm": 0.649410508856008,
+      "learning_rate": 9.985499041668664e-06,
+      "loss": 0.2589,
+      "step": 1095
+    },
+    {
+      "epoch": 0.36600434129236936,
+      "grad_norm": 0.5597787325842369,
+      "learning_rate": 9.985350767715236e-06,
+      "loss": 0.2265,
+      "step": 1096
+    },
+    {
+      "epoch": 0.36633828685924197,
+      "grad_norm": 0.5506515525737525,
+      "learning_rate": 9.985201740665873e-06,
+      "loss": 0.2254,
+      "step": 1097
+    },
+    {
+      "epoch": 0.3666722324261145,
+      "grad_norm": 0.599236470388504,
+      "learning_rate": 9.98505196054309e-06,
+      "loss": 0.2164,
+      "step": 1098
+    },
+    {
+      "epoch": 0.3670061779929871,
+      "grad_norm": 0.6214018418575783,
+      "learning_rate": 9.98490142736951e-06,
+      "loss": 0.258,
+      "step": 1099
+    },
+    {
+      "epoch": 0.36734012355985973,
+      "grad_norm": 0.6032134090942005,
+      "learning_rate": 9.984750141167874e-06,
+      "loss": 0.2412,
+      "step": 1100
+    },
+    {
+      "epoch": 0.36767406912673234,
+      "grad_norm": 0.6180438639636011,
+      "learning_rate": 9.984598101961036e-06,
+      "loss": 0.2426,
+      "step": 1101
+    },
+    {
+      "epoch": 0.36800801469360495,
+      "grad_norm": 0.9315847932687742,
+      "learning_rate": 9.984445309771963e-06,
+      "loss": 0.2613,
+      "step": 1102
+    },
+    {
+      "epoch": 0.36834196026047755,
+      "grad_norm": 0.5753009194815258,
+      "learning_rate": 9.984291764623735e-06,
+      "loss": 0.2298,
+      "step": 1103
+    },
+    {
+      "epoch": 0.36867590582735016,
+      "grad_norm": 0.7200885040091543,
+      "learning_rate": 9.98413746653955e-06,
+      "loss": 0.2487,
+      "step": 1104
+    },
+    {
+      "epoch": 0.36900985139422277,
+      "grad_norm": 0.6844842815753607,
+      "learning_rate": 9.983982415542713e-06,
+      "loss": 0.2229,
+      "step": 1105
+    },
+    {
+      "epoch": 0.3693437969610953,
+      "grad_norm": 0.610318931691899,
+      "learning_rate": 9.983826611656649e-06,
+      "loss": 0.2417,
+      "step": 1106
+    },
+    {
+      "epoch": 0.3696777425279679,
+      "grad_norm": 0.6735146550314316,
+      "learning_rate": 9.983670054904891e-06,
+      "loss": 0.2455,
+      "step": 1107
+    },
+    {
+      "epoch": 0.37001168809484053,
+      "grad_norm": 0.6404413794547794,
+      "learning_rate": 9.98351274531109e-06,
+      "loss": 0.2393,
+      "step": 1108
+    },
+    {
+      "epoch": 0.37034563366171314,
+      "grad_norm": 0.590966335203484,
+      "learning_rate": 9.983354682899012e-06,
+      "loss": 0.2316,
+      "step": 1109
+    },
+    {
+      "epoch": 0.37067957922858574,
+      "grad_norm": 0.6989245308286045,
+      "learning_rate": 9.98319586769253e-06,
+      "loss": 0.2408,
+      "step": 1110
+    },
+    {
+      "epoch": 0.37101352479545835,
+      "grad_norm": 0.641463921195796,
+      "learning_rate": 9.983036299715637e-06,
+      "loss": 0.2358,
+      "step": 1111
+    },
+    {
+      "epoch": 0.37134747036233096,
+      "grad_norm": 0.7275199738126983,
+      "learning_rate": 9.98287597899244e-06,
+      "loss": 0.2494,
+      "step": 1112
+    },
+    {
+      "epoch": 0.37168141592920356,
+      "grad_norm": 0.6809370125852048,
+      "learning_rate": 9.982714905547152e-06,
+      "loss": 0.2322,
+      "step": 1113
+    },
+    {
+      "epoch": 0.3720153614960761,
+      "grad_norm": 0.6312324387169239,
+      "learning_rate": 9.982553079404109e-06,
+      "loss": 0.2458,
+      "step": 1114
+    },
+    {
+      "epoch": 0.3723493070629487,
+      "grad_norm": 0.5856517142062221,
+      "learning_rate": 9.982390500587755e-06,
+      "loss": 0.2344,
+      "step": 1115
+    },
+    {
+      "epoch": 0.3726832526298213,
+      "grad_norm": 0.6576879620189332,
+      "learning_rate": 9.982227169122652e-06,
+      "loss": 0.2406,
+      "step": 1116
+    },
+    {
+      "epoch": 0.37301719819669393,
+      "grad_norm": 0.6109211179006137,
+      "learning_rate": 9.98206308503347e-06,
+      "loss": 0.2407,
+      "step": 1117
+    },
+    {
+      "epoch": 0.37335114376356654,
+      "grad_norm": 0.5360857677658837,
+      "learning_rate": 9.981898248344996e-06,
+      "loss": 0.2291,
+      "step": 1118
+    },
+    {
+      "epoch": 0.37368508933043915,
+      "grad_norm": 0.6350541721035942,
+      "learning_rate": 9.981732659082136e-06,
+      "loss": 0.2351,
+      "step": 1119
+    },
+    {
+      "epoch": 0.37401903489731175,
+      "grad_norm": 0.5564505418095035,
+      "learning_rate": 9.981566317269895e-06,
+      "loss": 0.2199,
+      "step": 1120
+    },
+    {
+      "epoch": 0.37435298046418436,
+      "grad_norm": 0.5869314057324901,
+      "learning_rate": 9.981399222933408e-06,
+      "loss": 0.2459,
+      "step": 1121
+    },
+    {
+      "epoch": 0.3746869260310569,
+      "grad_norm": 0.7657771256449448,
+      "learning_rate": 9.981231376097914e-06,
+      "loss": 0.2407,
+      "step": 1122
+    },
+    {
+      "epoch": 0.3750208715979295,
+      "grad_norm": 0.5795690536586283,
+      "learning_rate": 9.981062776788769e-06,
+      "loss": 0.2347,
+      "step": 1123
+    },
+    {
+      "epoch": 0.3753548171648021,
+      "grad_norm": 0.5891710443110899,
+      "learning_rate": 9.98089342503144e-06,
+      "loss": 0.2382,
+      "step": 1124
+    },
+    {
+      "epoch": 0.37568876273167473,
+      "grad_norm": 0.621199984594368,
+      "learning_rate": 9.980723320851512e-06,
+      "loss": 0.2347,
+      "step": 1125
+    },
+    {
+      "epoch": 0.37602270829854734,
+      "grad_norm": 0.6517502567095506,
+      "learning_rate": 9.98055246427468e-06,
+      "loss": 0.2462,
+      "step": 1126
+    },
+    {
+      "epoch": 0.37635665386541994,
+      "grad_norm": 0.6881884763822493,
+      "learning_rate": 9.980380855326754e-06,
+      "loss": 0.2504,
+      "step": 1127
+    },
+    {
+      "epoch": 0.37669059943229255,
+      "grad_norm": 0.5745203456313879,
+      "learning_rate": 9.980208494033659e-06,
+      "loss": 0.2404,
+      "step": 1128
+    },
+    {
+      "epoch": 0.37702454499916516,
+      "grad_norm": 0.6050555371100264,
+      "learning_rate": 9.98003538042143e-06,
+      "loss": 0.2531,
+      "step": 1129
+    },
+    {
+      "epoch": 0.37735849056603776,
+      "grad_norm": 0.9254994222499924,
+      "learning_rate": 9.979861514516217e-06,
+      "loss": 0.2427,
+      "step": 1130
+    },
+    {
+      "epoch": 0.3776924361329103,
+      "grad_norm": 0.6004163946203853,
+      "learning_rate": 9.979686896344289e-06,
+      "loss": 0.2199,
+      "step": 1131
+    },
+    {
+      "epoch": 0.3780263816997829,
+      "grad_norm": 0.6102051176431205,
+      "learning_rate": 9.97951152593202e-06,
+      "loss": 0.2514,
+      "step": 1132
+    },
+    {
+      "epoch": 0.3783603272666555,
+      "grad_norm": 0.6680363132138403,
+      "learning_rate": 9.979335403305904e-06,
+      "loss": 0.2491,
+      "step": 1133
+    },
+    {
+      "epoch": 0.37869427283352813,
+      "grad_norm": 0.6372805133933099,
+      "learning_rate": 9.979158528492546e-06,
+      "loss": 0.2553,
+      "step": 1134
+    },
+    {
+      "epoch": 0.37902821840040074,
+      "grad_norm": 0.7266606127441678,
+      "learning_rate": 9.978980901518663e-06,
+      "loss": 0.2435,
+      "step": 1135
+    },
+    {
+      "epoch": 0.37936216396727335,
+      "grad_norm": 0.5668609743092723,
+      "learning_rate": 9.978802522411091e-06,
+      "loss": 0.2361,
+      "step": 1136
+    },
+    {
+      "epoch": 0.37969610953414595,
+      "grad_norm": 0.6140023762446175,
+      "learning_rate": 9.978623391196774e-06,
+      "loss": 0.2581,
+      "step": 1137
+    },
+    {
+      "epoch": 0.38003005510101856,
+      "grad_norm": 0.6739052795637623,
+      "learning_rate": 9.978443507902772e-06,
+      "loss": 0.2446,
+      "step": 1138
+    },
+    {
+      "epoch": 0.3803640006678911,
+      "grad_norm": 0.5882555204439429,
+      "learning_rate": 9.978262872556257e-06,
+      "loss": 0.2407,
+      "step": 1139
+    },
+    {
+      "epoch": 0.3806979462347637,
+      "grad_norm": 0.7887496587693288,
+      "learning_rate": 9.97808148518452e-06,
+      "loss": 0.2564,
+      "step": 1140
+    },
+    {
+      "epoch": 0.3810318918016363,
+      "grad_norm": 0.7154651655569723,
+      "learning_rate": 9.977899345814959e-06,
+      "loss": 0.2677,
+      "step": 1141
+    },
+    {
+      "epoch": 0.38136583736850893,
+      "grad_norm": 0.5884603812414391,
+      "learning_rate": 9.977716454475089e-06,
+      "loss": 0.243,
+      "step": 1142
+    },
+    {
+      "epoch": 0.38169978293538154,
+      "grad_norm": 0.5670137381925265,
+      "learning_rate": 9.977532811192539e-06,
+      "loss": 0.2321,
+      "step": 1143
+    },
+    {
+      "epoch": 0.38203372850225414,
+      "grad_norm": 0.6540687344399836,
+      "learning_rate": 9.977348415995048e-06,
+      "loss": 0.2382,
+      "step": 1144
+    },
+    {
+      "epoch": 0.38236767406912675,
+      "grad_norm": 0.6062895960141702,
+      "learning_rate": 9.977163268910472e-06,
+      "loss": 0.2467,
+      "step": 1145
+    },
+    {
+      "epoch": 0.38270161963599936,
+      "grad_norm": 0.6354013496136728,
+      "learning_rate": 9.976977369966781e-06,
+      "loss": 0.2568,
+      "step": 1146
+    },
+    {
+      "epoch": 0.3830355652028719,
+      "grad_norm": 0.6291640800274567,
+      "learning_rate": 9.976790719192055e-06,
+      "loss": 0.239,
+      "step": 1147
+    },
+    {
+      "epoch": 0.3833695107697445,
+      "grad_norm": 0.6401219097828764,
+      "learning_rate": 9.976603316614492e-06,
+      "loss": 0.2528,
+      "step": 1148
+    },
+    {
+      "epoch": 0.3837034563366171,
+      "grad_norm": 0.6010786738863266,
+      "learning_rate": 9.976415162262401e-06,
+      "loss": 0.2328,
+      "step": 1149
+    },
+    {
+      "epoch": 0.38403740190348973,
+      "grad_norm": 0.606105846014851,
+      "learning_rate": 9.976226256164204e-06,
+      "loss": 0.249,
+      "step": 1150
+    },
+    {
+      "epoch": 0.38437134747036233,
+      "grad_norm": 0.6021670703179877,
+      "learning_rate": 9.976036598348437e-06,
+      "loss": 0.2464,
+      "step": 1151
+    },
+    {
+      "epoch": 0.38470529303723494,
+      "grad_norm": 0.5488912740197421,
+      "learning_rate": 9.975846188843754e-06,
+      "loss": 0.2343,
+      "step": 1152
+    },
+    {
+      "epoch": 0.38503923860410755,
+      "grad_norm": 0.6489026795753373,
+      "learning_rate": 9.975655027678913e-06,
+      "loss": 0.2383,
+      "step": 1153
+    },
+    {
+      "epoch": 0.38537318417098015,
+      "grad_norm": 0.5352509454624307,
+      "learning_rate": 9.975463114882792e-06,
+      "loss": 0.2385,
+      "step": 1154
+    },
+    {
+      "epoch": 0.3857071297378527,
+      "grad_norm": 0.5557378389098141,
+      "learning_rate": 9.975270450484385e-06,
+      "loss": 0.2331,
+      "step": 1155
+    },
+    {
+      "epoch": 0.3860410753047253,
+      "grad_norm": 0.6592063946537283,
+      "learning_rate": 9.975077034512795e-06,
+      "loss": 0.228,
+      "step": 1156
+    },
+    {
+      "epoch": 0.3863750208715979,
+      "grad_norm": 0.6561305895606129,
+      "learning_rate": 9.97488286699724e-06,
+      "loss": 0.2507,
+      "step": 1157
+    },
+    {
+      "epoch": 0.3867089664384705,
+      "grad_norm": 0.5344971723442756,
+      "learning_rate": 9.974687947967047e-06,
+      "loss": 0.2347,
+      "step": 1158
+    },
+    {
+      "epoch": 0.38704291200534313,
+      "grad_norm": 0.5740634729457469,
+      "learning_rate": 9.974492277451668e-06,
+      "loss": 0.2402,
+      "step": 1159
+    },
+    {
+      "epoch": 0.38737685757221574,
+      "grad_norm": 0.6784228325087824,
+      "learning_rate": 9.974295855480658e-06,
+      "loss": 0.2468,
+      "step": 1160
+    },
+    {
+      "epoch": 0.38771080313908834,
+      "grad_norm": 0.5351257365785597,
+      "learning_rate": 9.974098682083687e-06,
+      "loss": 0.236,
+      "step": 1161
+    },
+    {
+      "epoch": 0.38804474870596095,
+      "grad_norm": 0.5153552023195942,
+      "learning_rate": 9.973900757290541e-06,
+      "loss": 0.2096,
+      "step": 1162
+    },
+    {
+      "epoch": 0.3883786942728335,
+      "grad_norm": 0.5687511104588763,
+      "learning_rate": 9.97370208113112e-06,
+      "loss": 0.2368,
+      "step": 1163
+    },
+    {
+      "epoch": 0.3887126398397061,
+      "grad_norm": 0.6394062683395763,
+      "learning_rate": 9.973502653635438e-06,
+      "loss": 0.2521,
+      "step": 1164
+    },
+    {
+      "epoch": 0.3890465854065787,
+      "grad_norm": 0.6513149391308338,
+      "learning_rate": 9.97330247483362e-06,
+      "loss": 0.2421,
+      "step": 1165
+    },
+    {
+      "epoch": 0.3893805309734513,
+      "grad_norm": 0.5407808834505893,
+      "learning_rate": 9.973101544755901e-06,
+      "loss": 0.2295,
+      "step": 1166
+    },
+    {
+      "epoch": 0.38971447654032393,
+      "grad_norm": 0.6398170225716239,
+      "learning_rate": 9.97289986343264e-06,
+      "loss": 0.2346,
+      "step": 1167
+    },
+    {
+      "epoch": 0.39004842210719654,
+      "grad_norm": 0.6043234167832073,
+      "learning_rate": 9.972697430894299e-06,
+      "loss": 0.2338,
+      "step": 1168
+    },
+    {
+      "epoch": 0.39038236767406914,
+      "grad_norm": 0.6333600734362572,
+      "learning_rate": 9.97249424717146e-06,
+      "loss": 0.2472,
+      "step": 1169
+    },
+    {
+      "epoch": 0.39071631324094175,
+      "grad_norm": 0.6017998882821272,
+      "learning_rate": 9.972290312294816e-06,
+      "loss": 0.2488,
+      "step": 1170
+    },
+    {
+      "epoch": 0.3910502588078143,
+      "grad_norm": 0.6311509148718895,
+      "learning_rate": 9.972085626295173e-06,
+      "loss": 0.2426,
+      "step": 1171
+    },
+    {
+      "epoch": 0.3913842043746869,
+      "grad_norm": 0.6141756863914073,
+      "learning_rate": 9.971880189203452e-06,
+      "loss": 0.2537,
+      "step": 1172
+    },
+    {
+      "epoch": 0.3917181499415595,
+      "grad_norm": 0.5793351876901642,
+      "learning_rate": 9.971674001050687e-06,
+      "loss": 0.2432,
+      "step": 1173
+    },
+    {
+      "epoch": 0.3920520955084321,
+      "grad_norm": 0.7048398913255026,
+      "learning_rate": 9.971467061868022e-06,
+      "loss": 0.2774,
+      "step": 1174
+    },
+    {
+      "epoch": 0.3923860410753047,
+      "grad_norm": 0.5894804335467615,
+      "learning_rate": 9.971259371686724e-06,
+      "loss": 0.2264,
+      "step": 1175
+    },
+    {
+      "epoch": 0.39271998664217733,
+      "grad_norm": 0.6305881343504768,
+      "learning_rate": 9.971050930538161e-06,
+      "loss": 0.2515,
+      "step": 1176
+    },
+    {
+      "epoch": 0.39305393220904994,
+      "grad_norm": 0.6169670939348186,
+      "learning_rate": 9.970841738453823e-06,
+      "loss": 0.2474,
+      "step": 1177
+    },
+    {
+      "epoch": 0.39338787777592255,
+      "grad_norm": 0.6203195420477706,
+      "learning_rate": 9.970631795465311e-06,
+      "loss": 0.2446,
+      "step": 1178
+    },
+    {
+      "epoch": 0.39372182334279515,
+      "grad_norm": 0.5285532423016363,
+      "learning_rate": 9.970421101604339e-06,
+      "loss": 0.2168,
+      "step": 1179
+    },
+    {
+      "epoch": 0.3940557689096677,
+      "grad_norm": 0.6175748683107137,
+      "learning_rate": 9.970209656902734e-06,
+      "loss": 0.2421,
+      "step": 1180
+    },
+    {
+      "epoch": 0.3943897144765403,
+      "grad_norm": 0.5780266651964711,
+      "learning_rate": 9.969997461392439e-06,
+      "loss": 0.2334,
+      "step": 1181
+    },
+    {
+      "epoch": 0.3947236600434129,
+      "grad_norm": 0.7028863790206683,
+      "learning_rate": 9.969784515105508e-06,
+      "loss": 0.2619,
+      "step": 1182
+    },
+    {
+      "epoch": 0.3950576056102855,
+      "grad_norm": 0.5744613731332193,
+      "learning_rate": 9.969570818074109e-06,
+      "loss": 0.2273,
+      "step": 1183
+    },
+    {
+      "epoch": 0.39539155117715813,
+      "grad_norm": 0.7222715318962293,
+      "learning_rate": 9.96935637033052e-06,
+      "loss": 0.2282,
+      "step": 1184
+    },
+    {
+      "epoch": 0.39572549674403074,
+      "grad_norm": 0.6124101572650357,
+      "learning_rate": 9.969141171907142e-06,
+      "loss": 0.2348,
+      "step": 1185
+    },
+    {
+      "epoch": 0.39605944231090334,
+      "grad_norm": 0.5960876102003673,
+      "learning_rate": 9.968925222836478e-06,
+      "loss": 0.235,
+      "step": 1186
+    },
+    {
+      "epoch": 0.39639338787777595,
+      "grad_norm": 0.574113404593996,
+      "learning_rate": 9.968708523151154e-06,
+      "loss": 0.2494,
+      "step": 1187
+    },
+    {
+      "epoch": 0.3967273334446485,
+      "grad_norm": 0.660656988924486,
+      "learning_rate": 9.968491072883902e-06,
+      "loss": 0.2437,
+      "step": 1188
+    },
+    {
+      "epoch": 0.3970612790115211,
+      "grad_norm": 0.6305428349461573,
+      "learning_rate": 9.968272872067571e-06,
+      "loss": 0.2477,
+      "step": 1189
+    },
+    {
+      "epoch": 0.3973952245783937,
+      "grad_norm": 0.6256118019755806,
+      "learning_rate": 9.968053920735124e-06,
+      "loss": 0.2616,
+      "step": 1190
+    },
+    {
+      "epoch": 0.3977291701452663,
+      "grad_norm": 0.5724515150428571,
+      "learning_rate": 9.967834218919634e-06,
+      "loss": 0.2399,
+      "step": 1191
+    },
+    {
+      "epoch": 0.3980631157121389,
+      "grad_norm": 0.7058668997385279,
+      "learning_rate": 9.967613766654293e-06,
+      "loss": 0.2488,
+      "step": 1192
+    },
+    {
+      "epoch": 0.39839706127901153,
+      "grad_norm": 0.5837497840997278,
+      "learning_rate": 9.967392563972399e-06,
+      "loss": 0.2404,
+      "step": 1193
+    },
+    {
+      "epoch": 0.39873100684588414,
+      "grad_norm": 0.5877505859460367,
+      "learning_rate": 9.96717061090737e-06,
+      "loss": 0.2427,
+      "step": 1194
+    },
+    {
+      "epoch": 0.39906495241275675,
+      "grad_norm": 0.6767793531128838,
+      "learning_rate": 9.966947907492734e-06,
+      "loss": 0.2423,
+      "step": 1195
+    },
+    {
+      "epoch": 0.3993988979796293,
+      "grad_norm": 0.7876764633998403,
+      "learning_rate": 9.966724453762131e-06,
+      "loss": 0.2593,
+      "step": 1196
+    },
+    {
+      "epoch": 0.3997328435465019,
+      "grad_norm": 0.664535756576125,
+      "learning_rate": 9.96650024974932e-06,
+      "loss": 0.2668,
+      "step": 1197
+    },
+    {
+      "epoch": 0.4000667891133745,
+      "grad_norm": 0.6748545214684532,
+      "learning_rate": 9.966275295488165e-06,
+      "loss": 0.2307,
+      "step": 1198
+    },
+    {
+      "epoch": 0.4004007346802471,
+      "grad_norm": 0.5847120767540386,
+      "learning_rate": 9.966049591012651e-06,
+      "loss": 0.25,
+      "step": 1199
+    },
+    {
+      "epoch": 0.4007346802471197,
+      "grad_norm": 0.5593088473361937,
+      "learning_rate": 9.965823136356877e-06,
+      "loss": 0.2512,
+      "step": 1200
+    },
+    {
+      "epoch": 0.40106862581399233,
+      "grad_norm": 0.5979036171420881,
+      "learning_rate": 9.965595931555043e-06,
+      "loss": 0.2344,
+      "step": 1201
+    },
+    {
+      "epoch": 0.40140257138086494,
+      "grad_norm": 0.5610390171317386,
+      "learning_rate": 9.965367976641478e-06,
+      "loss": 0.2297,
+      "step": 1202
+    },
+    {
+      "epoch": 0.40173651694773754,
+      "grad_norm": 0.6833684414411917,
+      "learning_rate": 9.965139271650614e-06,
+      "loss": 0.2459,
+      "step": 1203
+    },
+    {
+      "epoch": 0.4020704625146101,
+      "grad_norm": 0.728364535501137,
+      "learning_rate": 9.964909816617002e-06,
+      "loss": 0.2544,
+      "step": 1204
+    },
+    {
+      "epoch": 0.4024044080814827,
+      "grad_norm": 0.5693930955364505,
+      "learning_rate": 9.964679611575298e-06,
+      "loss": 0.2378,
+      "step": 1205
+    },
+    {
+      "epoch": 0.4027383536483553,
+      "grad_norm": 0.6133102401066323,
+      "learning_rate": 9.964448656560286e-06,
+      "loss": 0.2279,
+      "step": 1206
+    },
+    {
+      "epoch": 0.4030722992152279,
+      "grad_norm": 0.563392679821793,
+      "learning_rate": 9.964216951606848e-06,
+      "loss": 0.2397,
+      "step": 1207
+    },
+    {
+      "epoch": 0.4034062447821005,
+      "grad_norm": 0.5672909007940137,
+      "learning_rate": 9.963984496749988e-06,
+      "loss": 0.2267,
+      "step": 1208
+    },
+    {
+      "epoch": 0.4037401903489731,
+      "grad_norm": 0.6488344228564302,
+      "learning_rate": 9.96375129202482e-06,
+      "loss": 0.2494,
+      "step": 1209
+    },
+    {
+      "epoch": 0.40407413591584573,
+      "grad_norm": 0.5139669587828634,
+      "learning_rate": 9.963517337466575e-06,
+      "loss": 0.207,
+      "step": 1210
+    },
+    {
+      "epoch": 0.40440808148271834,
+      "grad_norm": 0.5287676891678358,
+      "learning_rate": 9.963282633110591e-06,
+      "loss": 0.2303,
+      "step": 1211
+    },
+    {
+      "epoch": 0.4047420270495909,
+      "grad_norm": 0.5058680387315015,
+      "learning_rate": 9.963047178992324e-06,
+      "loss": 0.22,
+      "step": 1212
+    },
+    {
+      "epoch": 0.4050759726164635,
+      "grad_norm": 0.5445840438834135,
+      "learning_rate": 9.962810975147344e-06,
+      "loss": 0.2397,
+      "step": 1213
+    },
+    {
+      "epoch": 0.4054099181833361,
+      "grad_norm": 0.6382046509955335,
+      "learning_rate": 9.96257402161133e-06,
+      "loss": 0.2715,
+      "step": 1214
+    },
+    {
+      "epoch": 0.4057438637502087,
+      "grad_norm": 0.5413823853664697,
+      "learning_rate": 9.962336318420078e-06,
+      "loss": 0.2261,
+      "step": 1215
+    },
+    {
+      "epoch": 0.4060778093170813,
+      "grad_norm": 0.6202996949474051,
+      "learning_rate": 9.962097865609495e-06,
+      "loss": 0.2283,
+      "step": 1216
+    },
+    {
+      "epoch": 0.4064117548839539,
+      "grad_norm": 0.60696612940717,
+      "learning_rate": 9.961858663215604e-06,
+      "loss": 0.2401,
+      "step": 1217
+    },
+    {
+      "epoch": 0.40674570045082653,
+      "grad_norm": 0.649264221755663,
+      "learning_rate": 9.961618711274537e-06,
+      "loss": 0.2472,
+      "step": 1218
+    },
+    {
+      "epoch": 0.40707964601769914,
+      "grad_norm": 0.5501090472861203,
+      "learning_rate": 9.961378009822542e-06,
+      "loss": 0.2407,
+      "step": 1219
+    },
+    {
+      "epoch": 0.4074135915845717,
+      "grad_norm": 0.5345854297532595,
+      "learning_rate": 9.961136558895981e-06,
+      "loss": 0.228,
+      "step": 1220
+    },
+    {
+      "epoch": 0.4077475371514443,
+      "grad_norm": 0.5125851264368648,
+      "learning_rate": 9.960894358531328e-06,
+      "loss": 0.2167,
+      "step": 1221
+    },
+    {
+      "epoch": 0.4080814827183169,
+      "grad_norm": 0.5920584268375877,
+      "learning_rate": 9.960651408765168e-06,
+      "loss": 0.2367,
+      "step": 1222
+    },
+    {
+      "epoch": 0.4084154282851895,
+      "grad_norm": 0.716453784192867,
+      "learning_rate": 9.960407709634203e-06,
+      "loss": 0.2477,
+      "step": 1223
+    },
+    {
+      "epoch": 0.4087493738520621,
+      "grad_norm": 0.7562990534172225,
+      "learning_rate": 9.960163261175247e-06,
+      "loss": 0.2332,
+      "step": 1224
+    },
+    {
+      "epoch": 0.4090833194189347,
+      "grad_norm": 0.5231312350431248,
+      "learning_rate": 9.959918063425228e-06,
+      "loss": 0.2259,
+      "step": 1225
+    },
+    {
+      "epoch": 0.4094172649858073,
+      "grad_norm": 0.5921963859667128,
+      "learning_rate": 9.959672116421181e-06,
+      "loss": 0.2263,
+      "step": 1226
+    },
+    {
+      "epoch": 0.40975121055267993,
+      "grad_norm": 0.6028966599262293,
+      "learning_rate": 9.959425420200267e-06,
+      "loss": 0.2489,
+      "step": 1227
+    },
+    {
+      "epoch": 0.4100851561195525,
+      "grad_norm": 0.686672288403263,
+      "learning_rate": 9.959177974799742e-06,
+      "loss": 0.2508,
+      "step": 1228
+    },
+    {
+      "epoch": 0.4104191016864251,
+      "grad_norm": 0.5673389018187519,
+      "learning_rate": 9.958929780256996e-06,
+      "loss": 0.2207,
+      "step": 1229
+    },
+    {
+      "epoch": 0.4107530472532977,
+      "grad_norm": 0.5753969316911537,
+      "learning_rate": 9.958680836609516e-06,
+      "loss": 0.239,
+      "step": 1230
+    },
+    {
+      "epoch": 0.4110869928201703,
+      "grad_norm": 0.676883776899676,
+      "learning_rate": 9.95843114389491e-06,
+      "loss": 0.2225,
+      "step": 1231
+    },
+    {
+      "epoch": 0.4114209383870429,
+      "grad_norm": 0.6238678967525173,
+      "learning_rate": 9.958180702150895e-06,
+      "loss": 0.2278,
+      "step": 1232
+    },
+    {
+      "epoch": 0.4117548839539155,
+      "grad_norm": 0.6316307542043934,
+      "learning_rate": 9.957929511415304e-06,
+      "loss": 0.2485,
+      "step": 1233
+    },
+    {
+      "epoch": 0.4120888295207881,
+      "grad_norm": 0.5602900296016604,
+      "learning_rate": 9.957677571726084e-06,
+      "loss": 0.2288,
+      "step": 1234
+    },
+    {
+      "epoch": 0.41242277508766073,
+      "grad_norm": 0.6275850836013913,
+      "learning_rate": 9.95742488312129e-06,
+      "loss": 0.2393,
+      "step": 1235
+    },
+    {
+      "epoch": 0.41275672065453334,
+      "grad_norm": 0.6298629000187451,
+      "learning_rate": 9.957171445639096e-06,
+      "loss": 0.2456,
+      "step": 1236
+    },
+    {
+      "epoch": 0.4130906662214059,
+      "grad_norm": 0.5989836124208895,
+      "learning_rate": 9.956917259317788e-06,
+      "loss": 0.2349,
+      "step": 1237
+    },
+    {
+      "epoch": 0.4134246117882785,
+      "grad_norm": 0.6067541086981411,
+      "learning_rate": 9.95666232419576e-06,
+      "loss": 0.2377,
+      "step": 1238
+    },
+    {
+      "epoch": 0.4137585573551511,
+      "grad_norm": 0.5754404112239477,
+      "learning_rate": 9.956406640311527e-06,
+      "loss": 0.2555,
+      "step": 1239
+    },
+    {
+      "epoch": 0.4140925029220237,
+      "grad_norm": 0.5621508892107175,
+      "learning_rate": 9.956150207703712e-06,
+      "loss": 0.2386,
+      "step": 1240
+    },
+    {
+      "epoch": 0.4144264484888963,
+      "grad_norm": 0.5959704813881286,
+      "learning_rate": 9.955893026411048e-06,
+      "loss": 0.2469,
+      "step": 1241
+    },
+    {
+      "epoch": 0.4147603940557689,
+      "grad_norm": 0.6373037754433803,
+      "learning_rate": 9.955635096472391e-06,
+      "loss": 0.2326,
+      "step": 1242
+    },
+    {
+      "epoch": 0.41509433962264153,
+      "grad_norm": 0.8712757922100017,
+      "learning_rate": 9.9553764179267e-06,
+      "loss": 0.2408,
+      "step": 1243
+    },
+    {
+      "epoch": 0.41542828518951413,
+      "grad_norm": 0.5794440921487206,
+      "learning_rate": 9.955116990813056e-06,
+      "loss": 0.2515,
+      "step": 1244
+    },
+    {
+      "epoch": 0.4157622307563867,
+      "grad_norm": 0.5417617912815341,
+      "learning_rate": 9.954856815170644e-06,
+      "loss": 0.2257,
+      "step": 1245
+    },
+    {
+      "epoch": 0.4160961763232593,
+      "grad_norm": 0.608261531031807,
+      "learning_rate": 9.95459589103877e-06,
+      "loss": 0.2422,
+      "step": 1246
+    },
+    {
+      "epoch": 0.4164301218901319,
+      "grad_norm": 0.5882276462833476,
+      "learning_rate": 9.954334218456846e-06,
+      "loss": 0.237,
+      "step": 1247
+    },
+    {
+      "epoch": 0.4167640674570045,
+      "grad_norm": 0.9285636581320588,
+      "learning_rate": 9.954071797464405e-06,
+      "loss": 0.249,
+      "step": 1248
+    },
+    {
+      "epoch": 0.4170980130238771,
+      "grad_norm": 0.6268984023382586,
+      "learning_rate": 9.953808628101086e-06,
+      "loss": 0.2415,
+      "step": 1249
+    },
+    {
+      "epoch": 0.4174319585907497,
+      "grad_norm": 0.6069222808256296,
+      "learning_rate": 9.953544710406646e-06,
+      "loss": 0.2545,
+      "step": 1250
+    },
+    {
+      "epoch": 0.4177659041576223,
+      "grad_norm": 0.5778817193778154,
+      "learning_rate": 9.95328004442095e-06,
+      "loss": 0.241,
+      "step": 1251
+    },
+    {
+      "epoch": 0.41809984972449493,
+      "grad_norm": 0.61922600341147,
+      "learning_rate": 9.953014630183979e-06,
+      "loss": 0.2244,
+      "step": 1252
+    },
+    {
+      "epoch": 0.4184337952913675,
+      "grad_norm": 0.6687820682983995,
+      "learning_rate": 9.95274846773583e-06,
+      "loss": 0.2337,
+      "step": 1253
+    },
+    {
+      "epoch": 0.4187677408582401,
+      "grad_norm": 0.5042432866684735,
+      "learning_rate": 9.952481557116708e-06,
+      "loss": 0.2169,
+      "step": 1254
+    },
+    {
+      "epoch": 0.4191016864251127,
+      "grad_norm": 0.6258367926926472,
+      "learning_rate": 9.952213898366932e-06,
+      "loss": 0.249,
+      "step": 1255
+    },
+    {
+      "epoch": 0.4194356319919853,
+      "grad_norm": 0.5585811866612296,
+      "learning_rate": 9.951945491526938e-06,
+      "loss": 0.226,
+      "step": 1256
+    },
+    {
+      "epoch": 0.4197695775588579,
+      "grad_norm": 0.6218677388521249,
+      "learning_rate": 9.951676336637267e-06,
+      "loss": 0.2318,
+      "step": 1257
+    },
+    {
+      "epoch": 0.4201035231257305,
+      "grad_norm": 0.5375786545701925,
+      "learning_rate": 9.951406433738587e-06,
+      "loss": 0.2295,
+      "step": 1258
+    },
+    {
+      "epoch": 0.4204374686926031,
+      "grad_norm": 0.5316795089983514,
+      "learning_rate": 9.95113578287166e-06,
+      "loss": 0.2278,
+      "step": 1259
+    },
+    {
+      "epoch": 0.42077141425947573,
+      "grad_norm": 0.6387497516438277,
+      "learning_rate": 9.950864384077376e-06,
+      "loss": 0.2513,
+      "step": 1260
+    },
+    {
+      "epoch": 0.4211053598263483,
+      "grad_norm": 0.7334294291090919,
+      "learning_rate": 9.950592237396732e-06,
+      "loss": 0.2376,
+      "step": 1261
+    },
+    {
+      "epoch": 0.4214393053932209,
+      "grad_norm": 0.47638791811568987,
+      "learning_rate": 9.95031934287084e-06,
+      "loss": 0.2203,
+      "step": 1262
+    },
+    {
+      "epoch": 0.4217732509600935,
+      "grad_norm": 0.4970617194044711,
+      "learning_rate": 9.950045700540923e-06,
+      "loss": 0.2322,
+      "step": 1263
+    },
+    {
+      "epoch": 0.4221071965269661,
+      "grad_norm": 0.5481044534689685,
+      "learning_rate": 9.949771310448317e-06,
+      "loss": 0.2348,
+      "step": 1264
+    },
+    {
+      "epoch": 0.4224411420938387,
+      "grad_norm": 0.5249211204705201,
+      "learning_rate": 9.949496172634474e-06,
+      "loss": 0.2237,
+      "step": 1265
+    },
+    {
+      "epoch": 0.4227750876607113,
+      "grad_norm": 0.6158684467073959,
+      "learning_rate": 9.949220287140955e-06,
+      "loss": 0.2528,
+      "step": 1266
+    },
+    {
+      "epoch": 0.4231090332275839,
+      "grad_norm": 0.5751869229573902,
+      "learning_rate": 9.948943654009438e-06,
+      "loss": 0.2437,
+      "step": 1267
+    },
+    {
+      "epoch": 0.4234429787944565,
+      "grad_norm": 0.5278534477612402,
+      "learning_rate": 9.948666273281708e-06,
+      "loss": 0.2281,
+      "step": 1268
+    },
+    {
+      "epoch": 0.4237769243613291,
+      "grad_norm": 0.6021004635115851,
+      "learning_rate": 9.94838814499967e-06,
+      "loss": 0.2273,
+      "step": 1269
+    },
+    {
+      "epoch": 0.4241108699282017,
+      "grad_norm": 0.5556933684548845,
+      "learning_rate": 9.948109269205338e-06,
+      "loss": 0.2458,
+      "step": 1270
+    },
+    {
+      "epoch": 0.4244448154950743,
+      "grad_norm": 0.581592239372877,
+      "learning_rate": 9.947829645940836e-06,
+      "loss": 0.2316,
+      "step": 1271
+    },
+    {
+      "epoch": 0.4247787610619469,
+      "grad_norm": 0.5489943931236657,
+      "learning_rate": 9.94754927524841e-06,
+      "loss": 0.2313,
+      "step": 1272
+    },
+    {
+      "epoch": 0.4251127066288195,
+      "grad_norm": 0.6603804956451033,
+      "learning_rate": 9.947268157170409e-06,
+      "loss": 0.2503,
+      "step": 1273
+    },
+    {
+      "epoch": 0.4254466521956921,
+      "grad_norm": 0.5408603570247881,
+      "learning_rate": 9.9469862917493e-06,
+      "loss": 0.2385,
+      "step": 1274
+    },
+    {
+      "epoch": 0.4257805977625647,
+      "grad_norm": 0.506151945318377,
+      "learning_rate": 9.946703679027664e-06,
+      "loss": 0.2129,
+      "step": 1275
+    },
+    {
+      "epoch": 0.4261145433294373,
+      "grad_norm": 0.6388174252564417,
+      "learning_rate": 9.946420319048192e-06,
+      "loss": 0.2522,
+      "step": 1276
+    },
+    {
+      "epoch": 0.4264484888963099,
+      "grad_norm": 0.6479722260293415,
+      "learning_rate": 9.946136211853689e-06,
+      "loss": 0.2502,
+      "step": 1277
+    },
+    {
+      "epoch": 0.4267824344631825,
+      "grad_norm": 0.5423191804013884,
+      "learning_rate": 9.94585135748707e-06,
+      "loss": 0.2243,
+      "step": 1278
+    },
+    {
+      "epoch": 0.4271163800300551,
+      "grad_norm": 0.6023773989376456,
+      "learning_rate": 9.94556575599137e-06,
+      "loss": 0.2388,
+      "step": 1279
+    },
+    {
+      "epoch": 0.4274503255969277,
+      "grad_norm": 0.6403445586636296,
+      "learning_rate": 9.94527940740973e-06,
+      "loss": 0.253,
+      "step": 1280
+    },
+    {
+      "epoch": 0.4277842711638003,
+      "grad_norm": 0.5765601105471051,
+      "learning_rate": 9.944992311785406e-06,
+      "loss": 0.2351,
+      "step": 1281
+    },
+    {
+      "epoch": 0.4281182167306729,
+      "grad_norm": 0.5943838786454438,
+      "learning_rate": 9.94470446916177e-06,
+      "loss": 0.2477,
+      "step": 1282
+    },
+    {
+      "epoch": 0.4284521622975455,
+      "grad_norm": 0.5967603952647004,
+      "learning_rate": 9.9444158795823e-06,
+      "loss": 0.2286,
+      "step": 1283
+    },
+    {
+      "epoch": 0.4287861078644181,
+      "grad_norm": 0.5889203354692991,
+      "learning_rate": 9.944126543090593e-06,
+      "loss": 0.2318,
+      "step": 1284
+    },
+    {
+      "epoch": 0.4291200534312907,
+      "grad_norm": 0.5778644128403275,
+      "learning_rate": 9.943836459730356e-06,
+      "loss": 0.2385,
+      "step": 1285
+    },
+    {
+      "epoch": 0.4294539989981633,
+      "grad_norm": 0.829101133307546,
+      "learning_rate": 9.943545629545412e-06,
+      "loss": 0.2493,
+      "step": 1286
+    },
+    {
+      "epoch": 0.4297879445650359,
+      "grad_norm": 0.5682692079123255,
+      "learning_rate": 9.94325405257969e-06,
+      "loss": 0.2594,
+      "step": 1287
+    },
+    {
+      "epoch": 0.4301218901319085,
+      "grad_norm": 0.5616769757489756,
+      "learning_rate": 9.94296172887724e-06,
+      "loss": 0.2407,
+      "step": 1288
+    },
+    {
+      "epoch": 0.4304558356987811,
+      "grad_norm": 0.5082241382193762,
+      "learning_rate": 9.942668658482219e-06,
+      "loss": 0.2247,
+      "step": 1289
+    },
+    {
+      "epoch": 0.4307897812656537,
+      "grad_norm": 0.5720871139751971,
+      "learning_rate": 9.942374841438898e-06,
+      "loss": 0.2412,
+      "step": 1290
+    },
+    {
+      "epoch": 0.4311237268325263,
+      "grad_norm": 0.5571165598994844,
+      "learning_rate": 9.942080277791663e-06,
+      "loss": 0.2318,
+      "step": 1291
+    },
+    {
+      "epoch": 0.4314576723993989,
+      "grad_norm": 0.6092865735409773,
+      "learning_rate": 9.941784967585012e-06,
+      "loss": 0.2472,
+      "step": 1292
+    },
+    {
+      "epoch": 0.4317916179662715,
+      "grad_norm": 0.47701447633800187,
+      "learning_rate": 9.941488910863553e-06,
+      "loss": 0.2055,
+      "step": 1293
+    },
+    {
+      "epoch": 0.4321255635331441,
+      "grad_norm": 0.5174504669044425,
+      "learning_rate": 9.941192107672011e-06,
+      "loss": 0.2353,
+      "step": 1294
+    },
+    {
+      "epoch": 0.4324595091000167,
+      "grad_norm": 0.5998296572054426,
+      "learning_rate": 9.940894558055218e-06,
+      "loss": 0.2497,
+      "step": 1295
+    },
+    {
+      "epoch": 0.4327934546668893,
+      "grad_norm": 0.5960708917312382,
+      "learning_rate": 9.940596262058128e-06,
+      "loss": 0.2369,
+      "step": 1296
+    },
+    {
+      "epoch": 0.4331274002337619,
+      "grad_norm": 0.5601394501830662,
+      "learning_rate": 9.940297219725797e-06,
+      "loss": 0.2363,
+      "step": 1297
+    },
+    {
+      "epoch": 0.4334613458006345,
+      "grad_norm": 0.5181219900429102,
+      "learning_rate": 9.939997431103402e-06,
+      "loss": 0.2327,
+      "step": 1298
+    },
+    {
+      "epoch": 0.4337952913675071,
+      "grad_norm": 0.5893205151841744,
+      "learning_rate": 9.939696896236229e-06,
+      "loss": 0.2463,
+      "step": 1299
+    },
+    {
+      "epoch": 0.4341292369343797,
+      "grad_norm": 0.5527429152911172,
+      "learning_rate": 9.939395615169673e-06,
+      "loss": 0.2401,
+      "step": 1300
+    },
+    {
+      "epoch": 0.4344631825012523,
+      "grad_norm": 0.5109471022299947,
+      "learning_rate": 9.939093587949254e-06,
+      "loss": 0.2243,
+      "step": 1301
+    },
+    {
+      "epoch": 0.43479712806812487,
+      "grad_norm": 0.5656116960222147,
+      "learning_rate": 9.938790814620591e-06,
+      "loss": 0.2325,
+      "step": 1302
+    },
+    {
+      "epoch": 0.4351310736349975,
+      "grad_norm": 0.7838636481919538,
+      "learning_rate": 9.938487295229423e-06,
+      "loss": 0.2528,
+      "step": 1303
+    },
+    {
+      "epoch": 0.4354650192018701,
+      "grad_norm": 0.5427979756356961,
+      "learning_rate": 9.9381830298216e-06,
+      "loss": 0.2288,
+      "step": 1304
+    },
+    {
+      "epoch": 0.4357989647687427,
+      "grad_norm": 0.5688915141922664,
+      "learning_rate": 9.937878018443085e-06,
+      "loss": 0.2196,
+      "step": 1305
+    },
+    {
+      "epoch": 0.4361329103356153,
+      "grad_norm": 0.49786316106132467,
+      "learning_rate": 9.937572261139956e-06,
+      "loss": 0.2376,
+      "step": 1306
+    },
+    {
+      "epoch": 0.4364668559024879,
+      "grad_norm": 0.6767510753934641,
+      "learning_rate": 9.937265757958397e-06,
+      "loss": 0.2521,
+      "step": 1307
+    },
+    {
+      "epoch": 0.4368008014693605,
+      "grad_norm": 0.542449196038807,
+      "learning_rate": 9.93695850894471e-06,
+      "loss": 0.2305,
+      "step": 1308
+    },
+    {
+      "epoch": 0.4371347470362331,
+      "grad_norm": 0.5329012111090545,
+      "learning_rate": 9.93665051414531e-06,
+      "loss": 0.2241,
+      "step": 1309
+    },
+    {
+      "epoch": 0.43746869260310567,
+      "grad_norm": 0.5644124677867142,
+      "learning_rate": 9.936341773606723e-06,
+      "loss": 0.2418,
+      "step": 1310
+    },
+    {
+      "epoch": 0.4378026381699783,
+      "grad_norm": 0.8167083316266492,
+      "learning_rate": 9.936032287375587e-06,
+      "loss": 0.2533,
+      "step": 1311
+    },
+    {
+      "epoch": 0.4381365837368509,
+      "grad_norm": 0.5474885438610664,
+      "learning_rate": 9.935722055498655e-06,
+      "loss": 0.2368,
+      "step": 1312
+    },
+    {
+      "epoch": 0.4384705293037235,
+      "grad_norm": 0.6208297848511186,
+      "learning_rate": 9.935411078022791e-06,
+      "loss": 0.2388,
+      "step": 1313
+    },
+    {
+      "epoch": 0.4388044748705961,
+      "grad_norm": 0.6282302035769642,
+      "learning_rate": 9.93509935499497e-06,
+      "loss": 0.2493,
+      "step": 1314
+    },
+    {
+      "epoch": 0.4391384204374687,
+      "grad_norm": 0.5820840951158709,
+      "learning_rate": 9.934786886462282e-06,
+      "loss": 0.2263,
+      "step": 1315
+    },
+    {
+      "epoch": 0.4394723660043413,
+      "grad_norm": 0.582239493200357,
+      "learning_rate": 9.934473672471931e-06,
+      "loss": 0.2248,
+      "step": 1316
+    },
+    {
+      "epoch": 0.4398063115712139,
+      "grad_norm": 0.5473573028008037,
+      "learning_rate": 9.934159713071229e-06,
+      "loss": 0.2353,
+      "step": 1317
+    },
+    {
+      "epoch": 0.44014025713808647,
+      "grad_norm": 0.59739267317348,
+      "learning_rate": 9.933845008307605e-06,
+      "loss": 0.2398,
+      "step": 1318
+    },
+    {
+      "epoch": 0.4404742027049591,
+      "grad_norm": 0.6216076077529502,
+      "learning_rate": 9.933529558228599e-06,
+      "loss": 0.2263,
+      "step": 1319
+    },
+    {
+      "epoch": 0.4408081482718317,
+      "grad_norm": 0.5900409644511551,
+      "learning_rate": 9.933213362881861e-06,
+      "loss": 0.235,
+      "step": 1320
+    },
+    {
+      "epoch": 0.4411420938387043,
+      "grad_norm": 0.5318503294685462,
+      "learning_rate": 9.932896422315159e-06,
+      "loss": 0.2183,
+      "step": 1321
+    },
+    {
+      "epoch": 0.4414760394055769,
+      "grad_norm": 0.5992427234414971,
+      "learning_rate": 9.93257873657637e-06,
+      "loss": 0.2416,
+      "step": 1322
+    },
+    {
+      "epoch": 0.4418099849724495,
+      "grad_norm": 0.6067601416763263,
+      "learning_rate": 9.932260305713481e-06,
+      "loss": 0.2391,
+      "step": 1323
+    },
+    {
+      "epoch": 0.4421439305393221,
+      "grad_norm": 0.5311726881325887,
+      "learning_rate": 9.9319411297746e-06,
+      "loss": 0.2463,
+      "step": 1324
+    },
+    {
+      "epoch": 0.4424778761061947,
+      "grad_norm": 0.6287448890448014,
+      "learning_rate": 9.931621208807939e-06,
+      "loss": 0.2494,
+      "step": 1325
+    },
+    {
+      "epoch": 0.44281182167306726,
+      "grad_norm": 0.6170893282731494,
+      "learning_rate": 9.931300542861826e-06,
+      "loss": 0.2418,
+      "step": 1326
+    },
+    {
+      "epoch": 0.44314576723993987,
+      "grad_norm": 0.5783810354471404,
+      "learning_rate": 9.930979131984702e-06,
+      "loss": 0.2418,
+      "step": 1327
+    },
+    {
+      "epoch": 0.4434797128068125,
+      "grad_norm": 0.5475904806111974,
+      "learning_rate": 9.93065697622512e-06,
+      "loss": 0.2315,
+      "step": 1328
+    },
+    {
+      "epoch": 0.4438136583736851,
+      "grad_norm": 0.5659886586228791,
+      "learning_rate": 9.930334075631745e-06,
+      "loss": 0.223,
+      "step": 1329
+    },
+    {
+      "epoch": 0.4441476039405577,
+      "grad_norm": 0.5751459012303536,
+      "learning_rate": 9.930010430253356e-06,
+      "loss": 0.2215,
+      "step": 1330
+    },
+    {
+      "epoch": 0.4444815495074303,
+      "grad_norm": 0.6063177897359333,
+      "learning_rate": 9.92968604013884e-06,
+      "loss": 0.2563,
+      "step": 1331
+    },
+    {
+      "epoch": 0.4448154950743029,
+      "grad_norm": 0.7243143461095843,
+      "learning_rate": 9.929360905337204e-06,
+      "loss": 0.2327,
+      "step": 1332
+    },
+    {
+      "epoch": 0.4451494406411755,
+      "grad_norm": 0.570687190416024,
+      "learning_rate": 9.929035025897561e-06,
+      "loss": 0.2226,
+      "step": 1333
+    },
+    {
+      "epoch": 0.4454833862080481,
+      "grad_norm": 0.5940808216197645,
+      "learning_rate": 9.928708401869143e-06,
+      "loss": 0.2221,
+      "step": 1334
+    },
+    {
+      "epoch": 0.44581733177492067,
+      "grad_norm": 0.6811536558542676,
+      "learning_rate": 9.928381033301284e-06,
+      "loss": 0.2463,
+      "step": 1335
+    },
+    {
+      "epoch": 0.4461512773417933,
+      "grad_norm": 0.6824905822313252,
+      "learning_rate": 9.928052920243443e-06,
+      "loss": 0.2545,
+      "step": 1336
+    },
+    {
+      "epoch": 0.4464852229086659,
+      "grad_norm": 0.6637186339127504,
+      "learning_rate": 9.927724062745179e-06,
+      "loss": 0.2662,
+      "step": 1337
+    },
+    {
+      "epoch": 0.4468191684755385,
+      "grad_norm": 0.5809812855523707,
+      "learning_rate": 9.927394460856174e-06,
+      "loss": 0.2335,
+      "step": 1338
+    },
+    {
+      "epoch": 0.4471531140424111,
+      "grad_norm": 0.5700895995361696,
+      "learning_rate": 9.92706411462622e-06,
+      "loss": 0.2137,
+      "step": 1339
+    },
+    {
+      "epoch": 0.4474870596092837,
+      "grad_norm": 0.5735937174321404,
+      "learning_rate": 9.926733024105216e-06,
+      "loss": 0.2401,
+      "step": 1340
+    },
+    {
+      "epoch": 0.4478210051761563,
+      "grad_norm": 0.5128813606990313,
+      "learning_rate": 9.926401189343177e-06,
+      "loss": 0.2335,
+      "step": 1341
+    },
+    {
+      "epoch": 0.4481549507430289,
+      "grad_norm": 0.5468742678506229,
+      "learning_rate": 9.926068610390231e-06,
+      "loss": 0.2232,
+      "step": 1342
+    },
+    {
+      "epoch": 0.44848889630990146,
+      "grad_norm": 0.5713584153616417,
+      "learning_rate": 9.925735287296621e-06,
+      "loss": 0.2331,
+      "step": 1343
+    },
+    {
+      "epoch": 0.44882284187677407,
+      "grad_norm": 0.6096003519856071,
+      "learning_rate": 9.925401220112698e-06,
+      "loss": 0.2303,
+      "step": 1344
+    },
+    {
+      "epoch": 0.4491567874436467,
+      "grad_norm": 0.58393370045869,
+      "learning_rate": 9.925066408888924e-06,
+      "loss": 0.2401,
+      "step": 1345
+    },
+    {
+      "epoch": 0.4494907330105193,
+      "grad_norm": 0.5503546449887823,
+      "learning_rate": 9.92473085367588e-06,
+      "loss": 0.2422,
+      "step": 1346
+    },
+    {
+      "epoch": 0.4498246785773919,
+      "grad_norm": 0.5633846354206201,
+      "learning_rate": 9.924394554524252e-06,
+      "loss": 0.245,
+      "step": 1347
+    },
+    {
+      "epoch": 0.4501586241442645,
+      "grad_norm": 0.5534055213955522,
+      "learning_rate": 9.924057511484844e-06,
+      "loss": 0.2315,
+      "step": 1348
+    },
+    {
+      "epoch": 0.4504925697111371,
+      "grad_norm": 0.6078240341950537,
+      "learning_rate": 9.92371972460857e-06,
+      "loss": 0.2341,
+      "step": 1349
+    },
+    {
+      "epoch": 0.4508265152780097,
+      "grad_norm": 0.7758411012157077,
+      "learning_rate": 9.923381193946457e-06,
+      "loss": 0.2513,
+      "step": 1350
+    },
+    {
+      "epoch": 0.45116046084488226,
+      "grad_norm": 0.590993603668986,
+      "learning_rate": 9.923041919549644e-06,
+      "loss": 0.2259,
+      "step": 1351
+    },
+    {
+      "epoch": 0.45149440641175487,
+      "grad_norm": 0.5771541634320295,
+      "learning_rate": 9.92270190146938e-06,
+      "loss": 0.237,
+      "step": 1352
+    },
+    {
+      "epoch": 0.4518283519786275,
+      "grad_norm": 0.5921642270486791,
+      "learning_rate": 9.922361139757033e-06,
+      "loss": 0.2391,
+      "step": 1353
+    },
+    {
+      "epoch": 0.4521622975455001,
+      "grad_norm": 0.5803176539415612,
+      "learning_rate": 9.922019634464077e-06,
+      "loss": 0.237,
+      "step": 1354
+    },
+    {
+      "epoch": 0.4524962431123727,
+      "grad_norm": 0.5689236619706025,
+      "learning_rate": 9.9216773856421e-06,
+      "loss": 0.2329,
+      "step": 1355
+    },
+    {
+      "epoch": 0.4528301886792453,
+      "grad_norm": 0.4940682334608807,
+      "learning_rate": 9.921334393342803e-06,
+      "loss": 0.2341,
+      "step": 1356
+    },
+    {
+      "epoch": 0.4531641342461179,
+      "grad_norm": 0.5752961824097249,
+      "learning_rate": 9.920990657617998e-06,
+      "loss": 0.254,
+      "step": 1357
+    },
+    {
+      "epoch": 0.4534980798129905,
+      "grad_norm": 0.5364989105256935,
+      "learning_rate": 9.920646178519612e-06,
+      "loss": 0.2203,
+      "step": 1358
+    },
+    {
+      "epoch": 0.45383202537986306,
+      "grad_norm": 0.5052033735914636,
+      "learning_rate": 9.920300956099682e-06,
+      "loss": 0.2328,
+      "step": 1359
+    },
+    {
+      "epoch": 0.45416597094673566,
+      "grad_norm": 0.47346441786960736,
+      "learning_rate": 9.919954990410359e-06,
+      "loss": 0.2226,
+      "step": 1360
+    },
+    {
+      "epoch": 0.45449991651360827,
+      "grad_norm": 0.4934375959384355,
+      "learning_rate": 9.919608281503903e-06,
+      "loss": 0.2348,
+      "step": 1361
+    },
+    {
+      "epoch": 0.4548338620804809,
+      "grad_norm": 0.511312257771701,
+      "learning_rate": 9.91926082943269e-06,
+      "loss": 0.2153,
+      "step": 1362
+    },
+    {
+      "epoch": 0.4551678076473535,
+      "grad_norm": 0.5543164186583251,
+      "learning_rate": 9.918912634249206e-06,
+      "loss": 0.2454,
+      "step": 1363
+    },
+    {
+      "epoch": 0.4555017532142261,
+      "grad_norm": 0.5212508239813047,
+      "learning_rate": 9.91856369600605e-06,
+      "loss": 0.2232,
+      "step": 1364
+    },
+    {
+      "epoch": 0.4558356987810987,
+      "grad_norm": 0.527429149205606,
+      "learning_rate": 9.918214014755935e-06,
+      "loss": 0.2342,
+      "step": 1365
+    },
+    {
+      "epoch": 0.4561696443479713,
+      "grad_norm": 0.5562734152529801,
+      "learning_rate": 9.917863590551682e-06,
+      "loss": 0.2185,
+      "step": 1366
+    },
+    {
+      "epoch": 0.45650358991484385,
+      "grad_norm": 0.5093451096437928,
+      "learning_rate": 9.917512423446226e-06,
+      "loss": 0.2252,
+      "step": 1367
+    },
+    {
+      "epoch": 0.45683753548171646,
+      "grad_norm": 0.5376789383435711,
+      "learning_rate": 9.917160513492619e-06,
+      "loss": 0.2326,
+      "step": 1368
+    },
+    {
+      "epoch": 0.45717148104858907,
+      "grad_norm": 0.5160005226617597,
+      "learning_rate": 9.916807860744017e-06,
+      "loss": 0.2458,
+      "step": 1369
+    },
+    {
+      "epoch": 0.4575054266154617,
+      "grad_norm": 0.5243392738518032,
+      "learning_rate": 9.916454465253695e-06,
+      "loss": 0.2208,
+      "step": 1370
+    },
+    {
+      "epoch": 0.4578393721823343,
+      "grad_norm": 0.5454097822750845,
+      "learning_rate": 9.916100327075038e-06,
+      "loss": 0.2388,
+      "step": 1371
+    },
+    {
+      "epoch": 0.4581733177492069,
+      "grad_norm": 0.5311013984757362,
+      "learning_rate": 9.91574544626154e-06,
+      "loss": 0.2374,
+      "step": 1372
+    },
+    {
+      "epoch": 0.4585072633160795,
+      "grad_norm": 0.5081063384150541,
+      "learning_rate": 9.915389822866811e-06,
+      "loss": 0.2221,
+      "step": 1373
+    },
+    {
+      "epoch": 0.4588412088829521,
+      "grad_norm": 0.529213615144114,
+      "learning_rate": 9.915033456944572e-06,
+      "loss": 0.2264,
+      "step": 1374
+    },
+    {
+      "epoch": 0.45917515444982465,
+      "grad_norm": 0.5216114427240847,
+      "learning_rate": 9.914676348548658e-06,
+      "loss": 0.2343,
+      "step": 1375
+    },
+    {
+      "epoch": 0.45950910001669726,
+      "grad_norm": 0.5098198767095327,
+      "learning_rate": 9.914318497733013e-06,
+      "loss": 0.2199,
+      "step": 1376
+    },
+    {
+      "epoch": 0.45984304558356986,
+      "grad_norm": 0.5772594585549337,
+      "learning_rate": 9.913959904551695e-06,
+      "loss": 0.2444,
+      "step": 1377
+    },
+    {
+      "epoch": 0.46017699115044247,
+      "grad_norm": 0.5692684244936813,
+      "learning_rate": 9.913600569058871e-06,
+      "loss": 0.2325,
+      "step": 1378
+    },
+    {
+      "epoch": 0.4605109367173151,
+      "grad_norm": 0.5528985886777825,
+      "learning_rate": 9.913240491308828e-06,
+      "loss": 0.2268,
+      "step": 1379
+    },
+    {
+      "epoch": 0.4608448822841877,
+      "grad_norm": 0.5417272692609387,
+      "learning_rate": 9.912879671355956e-06,
+      "loss": 0.2171,
+      "step": 1380
+    },
+    {
+      "epoch": 0.4611788278510603,
+      "grad_norm": 0.5302491893848811,
+      "learning_rate": 9.912518109254763e-06,
+      "loss": 0.2215,
+      "step": 1381
+    },
+    {
+      "epoch": 0.4615127734179329,
+      "grad_norm": 0.5350805468842509,
+      "learning_rate": 9.912155805059866e-06,
+      "loss": 0.2351,
+      "step": 1382
+    },
+    {
+      "epoch": 0.4618467189848055,
+      "grad_norm": 0.5606286743790297,
+      "learning_rate": 9.911792758825996e-06,
+      "loss": 0.2267,
+      "step": 1383
+    },
+    {
+      "epoch": 0.46218066455167806,
+      "grad_norm": 0.5599706835878967,
+      "learning_rate": 9.911428970607995e-06,
+      "loss": 0.2361,
+      "step": 1384
+    },
+    {
+      "epoch": 0.46251461011855066,
+      "grad_norm": 0.615424046565781,
+      "learning_rate": 9.911064440460818e-06,
+      "loss": 0.2298,
+      "step": 1385
+    },
+    {
+      "epoch": 0.46284855568542327,
+      "grad_norm": 0.5649976414245927,
+      "learning_rate": 9.91069916843953e-06,
+      "loss": 0.232,
+      "step": 1386
+    },
+    {
+      "epoch": 0.4631825012522959,
+      "grad_norm": 0.5508577983404023,
+      "learning_rate": 9.910333154599314e-06,
+      "loss": 0.2436,
+      "step": 1387
+    },
+    {
+      "epoch": 0.4635164468191685,
+      "grad_norm": 0.5205854960323886,
+      "learning_rate": 9.909966398995456e-06,
+      "loss": 0.2362,
+      "step": 1388
+    },
+    {
+      "epoch": 0.4638503923860411,
+      "grad_norm": 0.5757688708297488,
+      "learning_rate": 9.909598901683361e-06,
+      "loss": 0.2491,
+      "step": 1389
+    },
+    {
+      "epoch": 0.4641843379529137,
+      "grad_norm": 0.5869979821978962,
+      "learning_rate": 9.909230662718543e-06,
+      "loss": 0.2324,
+      "step": 1390
+    },
+    {
+      "epoch": 0.4645182835197863,
+      "grad_norm": 0.5625640594323424,
+      "learning_rate": 9.908861682156628e-06,
+      "loss": 0.236,
+      "step": 1391
+    },
+    {
+      "epoch": 0.46485222908665885,
+      "grad_norm": 0.7152995547677362,
+      "learning_rate": 9.908491960053357e-06,
+      "loss": 0.2418,
+      "step": 1392
+    },
+    {
+      "epoch": 0.46518617465353146,
+      "grad_norm": 0.4895348021939775,
+      "learning_rate": 9.90812149646458e-06,
+      "loss": 0.2246,
+      "step": 1393
+    },
+    {
+      "epoch": 0.46552012022040407,
+      "grad_norm": 0.5909081784606736,
+      "learning_rate": 9.907750291446258e-06,
+      "loss": 0.2381,
+      "step": 1394
+    },
+    {
+      "epoch": 0.46585406578727667,
+      "grad_norm": 0.5724895641003475,
+      "learning_rate": 9.907378345054471e-06,
+      "loss": 0.2385,
+      "step": 1395
+    },
+    {
+      "epoch": 0.4661880113541493,
+      "grad_norm": 0.4977242293200873,
+      "learning_rate": 9.9070056573454e-06,
+      "loss": 0.2221,
+      "step": 1396
+    },
+    {
+      "epoch": 0.4665219569210219,
+      "grad_norm": 0.7104851590746769,
+      "learning_rate": 9.906632228375346e-06,
+      "loss": 0.2439,
+      "step": 1397
+    },
+    {
+      "epoch": 0.4668559024878945,
+      "grad_norm": 0.5366132433315185,
+      "learning_rate": 9.906258058200722e-06,
+      "loss": 0.2334,
+      "step": 1398
+    },
+    {
+      "epoch": 0.4671898480547671,
+      "grad_norm": 0.603863208324995,
+      "learning_rate": 9.905883146878049e-06,
+      "loss": 0.2469,
+      "step": 1399
+    },
+    {
+      "epoch": 0.46752379362163965,
+      "grad_norm": 0.6712903315298903,
+      "learning_rate": 9.90550749446396e-06,
+      "loss": 0.2525,
+      "step": 1400
+    },
+    {
+      "epoch": 0.46785773918851226,
+      "grad_norm": 0.5599933585459058,
+      "learning_rate": 9.905131101015204e-06,
+      "loss": 0.2298,
+      "step": 1401
+    },
+    {
+      "epoch": 0.46819168475538486,
+      "grad_norm": 0.6742593776670335,
+      "learning_rate": 9.904753966588638e-06,
+      "loss": 0.2249,
+      "step": 1402
+    },
+    {
+      "epoch": 0.46852563032225747,
+      "grad_norm": 0.5966659500589226,
+      "learning_rate": 9.904376091241236e-06,
+      "loss": 0.2406,
+      "step": 1403
+    },
+    {
+      "epoch": 0.4688595758891301,
+      "grad_norm": 0.5086157470594107,
+      "learning_rate": 9.903997475030077e-06,
+      "loss": 0.2239,
+      "step": 1404
+    },
+    {
+      "epoch": 0.4691935214560027,
+      "grad_norm": 0.5688250377303993,
+      "learning_rate": 9.903618118012358e-06,
+      "loss": 0.2499,
+      "step": 1405
+    },
+    {
+      "epoch": 0.4695274670228753,
+      "grad_norm": 0.5001870005825596,
+      "learning_rate": 9.903238020245383e-06,
+      "loss": 0.228,
+      "step": 1406
+    },
+    {
+      "epoch": 0.4698614125897479,
+      "grad_norm": 0.6111840293711042,
+      "learning_rate": 9.902857181786571e-06,
+      "loss": 0.2388,
+      "step": 1407
+    },
+    {
+      "epoch": 0.47019535815662045,
+      "grad_norm": 0.620871773232944,
+      "learning_rate": 9.902475602693451e-06,
+      "loss": 0.2176,
+      "step": 1408
+    },
+    {
+      "epoch": 0.47052930372349305,
+      "grad_norm": 0.4920698441065501,
+      "learning_rate": 9.90209328302367e-06,
+      "loss": 0.2375,
+      "step": 1409
+    },
+    {
+      "epoch": 0.47086324929036566,
+      "grad_norm": 0.5528573354567796,
+      "learning_rate": 9.901710222834976e-06,
+      "loss": 0.2424,
+      "step": 1410
+    },
+    {
+      "epoch": 0.47119719485723827,
+      "grad_norm": 0.5208706668107904,
+      "learning_rate": 9.901326422185238e-06,
+      "loss": 0.2298,
+      "step": 1411
+    },
+    {
+      "epoch": 0.4715311404241109,
+      "grad_norm": 0.5579864808318521,
+      "learning_rate": 9.900941881132431e-06,
+      "loss": 0.2303,
+      "step": 1412
+    },
+    {
+      "epoch": 0.4718650859909835,
+      "grad_norm": 0.7348092396729472,
+      "learning_rate": 9.900556599734647e-06,
+      "loss": 0.2589,
+      "step": 1413
+    },
+    {
+      "epoch": 0.4721990315578561,
+      "grad_norm": 0.5613570516628159,
+      "learning_rate": 9.900170578050088e-06,
+      "loss": 0.2541,
+      "step": 1414
+    },
+    {
+      "epoch": 0.4725329771247287,
+      "grad_norm": 0.5573029926646937,
+      "learning_rate": 9.899783816137065e-06,
+      "loss": 0.2491,
+      "step": 1415
+    },
+    {
+      "epoch": 0.47286692269160124,
+      "grad_norm": 0.5737576301390739,
+      "learning_rate": 9.899396314054002e-06,
+      "loss": 0.2165,
+      "step": 1416
+    },
+    {
+      "epoch": 0.47320086825847385,
+      "grad_norm": 0.49834479477003774,
+      "learning_rate": 9.89900807185944e-06,
+      "loss": 0.2188,
+      "step": 1417
+    },
+    {
+      "epoch": 0.47353481382534646,
+      "grad_norm": 0.5684392075108803,
+      "learning_rate": 9.89861908961202e-06,
+      "loss": 0.2361,
+      "step": 1418
+    },
+    {
+      "epoch": 0.47386875939221906,
+      "grad_norm": 0.5847270713464249,
+      "learning_rate": 9.89822936737051e-06,
+      "loss": 0.2402,
+      "step": 1419
+    },
+    {
+      "epoch": 0.47420270495909167,
+      "grad_norm": 0.5831201748397754,
+      "learning_rate": 9.897838905193781e-06,
+      "loss": 0.2264,
+      "step": 1420
+    },
+    {
+      "epoch": 0.4745366505259643,
+      "grad_norm": 0.6718561947154861,
+      "learning_rate": 9.897447703140813e-06,
+      "loss": 0.2483,
+      "step": 1421
+    },
+    {
+      "epoch": 0.4748705960928369,
+      "grad_norm": 0.5212579182437543,
+      "learning_rate": 9.897055761270705e-06,
+      "loss": 0.2264,
+      "step": 1422
+    },
+    {
+      "epoch": 0.4752045416597095,
+      "grad_norm": 0.5544409675912124,
+      "learning_rate": 9.896663079642663e-06,
+      "loss": 0.2447,
+      "step": 1423
+    },
+    {
+      "epoch": 0.47553848722658204,
+      "grad_norm": 0.5946504988614014,
+      "learning_rate": 9.896269658316006e-06,
+      "loss": 0.2322,
+      "step": 1424
+    },
+    {
+      "epoch": 0.47587243279345465,
+      "grad_norm": 0.5325243574651551,
+      "learning_rate": 9.895875497350165e-06,
+      "loss": 0.2138,
+      "step": 1425
+    },
+    {
+      "epoch": 0.47620637836032725,
+      "grad_norm": 0.7211936693494748,
+      "learning_rate": 9.895480596804684e-06,
+      "loss": 0.2187,
+      "step": 1426
+    },
+    {
+      "epoch": 0.47654032392719986,
+      "grad_norm": 0.5542697823075144,
+      "learning_rate": 9.895084956739215e-06,
+      "loss": 0.2251,
+      "step": 1427
+    },
+    {
+      "epoch": 0.47687426949407247,
+      "grad_norm": 0.5701293308719834,
+      "learning_rate": 9.894688577213527e-06,
+      "loss": 0.2416,
+      "step": 1428
+    },
+    {
+      "epoch": 0.4772082150609451,
+      "grad_norm": 0.5347630793631123,
+      "learning_rate": 9.894291458287496e-06,
+      "loss": 0.2292,
+      "step": 1429
+    },
+    {
+      "epoch": 0.4775421606278177,
+      "grad_norm": 0.5973098372961372,
+      "learning_rate": 9.893893600021112e-06,
+      "loss": 0.2455,
+      "step": 1430
+    },
+    {
+      "epoch": 0.4778761061946903,
+      "grad_norm": 0.5641760379191003,
+      "learning_rate": 9.893495002474475e-06,
+      "loss": 0.231,
+      "step": 1431
+    },
+    {
+      "epoch": 0.47821005176156284,
+      "grad_norm": 0.5541774904431316,
+      "learning_rate": 9.893095665707801e-06,
+      "loss": 0.2501,
+      "step": 1432
+    },
+    {
+      "epoch": 0.47854399732843544,
+      "grad_norm": 0.5136825487190733,
+      "learning_rate": 9.89269558978141e-06,
+      "loss": 0.2241,
+      "step": 1433
+    },
+    {
+      "epoch": 0.47887794289530805,
+      "grad_norm": 0.6526585039569316,
+      "learning_rate": 9.892294774755741e-06,
+      "loss": 0.25,
+      "step": 1434
+    },
+    {
+      "epoch": 0.47921188846218066,
+      "grad_norm": 0.5924925924061507,
+      "learning_rate": 9.891893220691343e-06,
+      "loss": 0.245,
+      "step": 1435
+    },
+    {
+      "epoch": 0.47954583402905326,
+      "grad_norm": 0.605221439278321,
+      "learning_rate": 9.891490927648872e-06,
+      "loss": 0.2365,
+      "step": 1436
+    },
+    {
+      "epoch": 0.47987977959592587,
+      "grad_norm": 0.5890164596674092,
+      "learning_rate": 9.891087895689102e-06,
+      "loss": 0.2298,
+      "step": 1437
+    },
+    {
+      "epoch": 0.4802137251627985,
+      "grad_norm": 0.5321381950410381,
+      "learning_rate": 9.890684124872914e-06,
+      "loss": 0.2202,
+      "step": 1438
+    },
+    {
+      "epoch": 0.4805476707296711,
+      "grad_norm": 0.7746352115111984,
+      "learning_rate": 9.890279615261302e-06,
+      "loss": 0.2241,
+      "step": 1439
+    },
+    {
+      "epoch": 0.4808816162965437,
+      "grad_norm": 0.5854514402548051,
+      "learning_rate": 9.889874366915374e-06,
+      "loss": 0.2265,
+      "step": 1440
+    },
+    {
+      "epoch": 0.48121556186341624,
+      "grad_norm": 0.5842152055041235,
+      "learning_rate": 9.889468379896347e-06,
+      "loss": 0.2398,
+      "step": 1441
+    },
+    {
+      "epoch": 0.48154950743028885,
+      "grad_norm": 0.5692166615610584,
+      "learning_rate": 9.88906165426555e-06,
+      "loss": 0.248,
+      "step": 1442
+    },
+    {
+      "epoch": 0.48188345299716145,
+      "grad_norm": 0.459765580131553,
+      "learning_rate": 9.888654190084422e-06,
+      "loss": 0.2071,
+      "step": 1443
+    },
+    {
+      "epoch": 0.48221739856403406,
+      "grad_norm": 0.6471093852235853,
+      "learning_rate": 9.888245987414517e-06,
+      "loss": 0.2368,
+      "step": 1444
+    },
+    {
+      "epoch": 0.48255134413090667,
+      "grad_norm": 0.5765295704173624,
+      "learning_rate": 9.8878370463175e-06,
+      "loss": 0.238,
+      "step": 1445
+    },
+    {
+      "epoch": 0.4828852896977793,
+      "grad_norm": 0.5371085420371692,
+      "learning_rate": 9.887427366855142e-06,
+      "loss": 0.2234,
+      "step": 1446
+    },
+    {
+      "epoch": 0.4832192352646519,
+      "grad_norm": 0.49845898630714686,
+      "learning_rate": 9.887016949089334e-06,
+      "loss": 0.2377,
+      "step": 1447
+    },
+    {
+      "epoch": 0.4835531808315245,
+      "grad_norm": 0.47885124555831643,
+      "learning_rate": 9.886605793082073e-06,
+      "loss": 0.2262,
+      "step": 1448
+    },
+    {
+      "epoch": 0.48388712639839704,
+      "grad_norm": 0.5622417184126103,
+      "learning_rate": 9.886193898895468e-06,
+      "loss": 0.2388,
+      "step": 1449
+    },
+    {
+      "epoch": 0.48422107196526964,
+      "grad_norm": 0.6126625007922746,
+      "learning_rate": 9.885781266591742e-06,
+      "loss": 0.2374,
+      "step": 1450
+    },
+    {
+      "epoch": 0.48455501753214225,
+      "grad_norm": 0.5754346810486999,
+      "learning_rate": 9.885367896233229e-06,
+      "loss": 0.2495,
+      "step": 1451
+    },
+    {
+      "epoch": 0.48488896309901486,
+      "grad_norm": 0.5679227654667943,
+      "learning_rate": 9.88495378788237e-06,
+      "loss": 0.232,
+      "step": 1452
+    },
+    {
+      "epoch": 0.48522290866588746,
+      "grad_norm": 0.523471620727347,
+      "learning_rate": 9.884538941601725e-06,
+      "loss": 0.2433,
+      "step": 1453
+    },
+    {
+      "epoch": 0.48555685423276007,
+      "grad_norm": 0.527665998819472,
+      "learning_rate": 9.884123357453959e-06,
+      "loss": 0.2264,
+      "step": 1454
+    },
+    {
+      "epoch": 0.4858907997996327,
+      "grad_norm": 0.507908698813618,
+      "learning_rate": 9.883707035501849e-06,
+      "loss": 0.2211,
+      "step": 1455
+    },
+    {
+      "epoch": 0.4862247453665053,
+      "grad_norm": 0.5790040972393914,
+      "learning_rate": 9.883289975808288e-06,
+      "loss": 0.2475,
+      "step": 1456
+    },
+    {
+      "epoch": 0.48655869093337784,
+      "grad_norm": 0.6259861327751014,
+      "learning_rate": 9.882872178436277e-06,
+      "loss": 0.2122,
+      "step": 1457
+    },
+    {
+      "epoch": 0.48689263650025044,
+      "grad_norm": 0.5921509956180295,
+      "learning_rate": 9.882453643448933e-06,
+      "loss": 0.2431,
+      "step": 1458
+    },
+    {
+      "epoch": 0.48722658206712305,
+      "grad_norm": 0.6237750475844797,
+      "learning_rate": 9.882034370909474e-06,
+      "loss": 0.2505,
+      "step": 1459
+    },
+    {
+      "epoch": 0.48756052763399566,
+      "grad_norm": 0.5668101845599637,
+      "learning_rate": 9.88161436088124e-06,
+      "loss": 0.2474,
+      "step": 1460
+    },
+    {
+      "epoch": 0.48789447320086826,
+      "grad_norm": 0.5512787179899852,
+      "learning_rate": 9.881193613427676e-06,
+      "loss": 0.2422,
+      "step": 1461
+    },
+    {
+      "epoch": 0.48822841876774087,
+      "grad_norm": 0.5569486602988732,
+      "learning_rate": 9.880772128612345e-06,
+      "loss": 0.2286,
+      "step": 1462
+    },
+    {
+      "epoch": 0.4885623643346135,
+      "grad_norm": 0.666951567922003,
+      "learning_rate": 9.880349906498914e-06,
+      "loss": 0.2386,
+      "step": 1463
+    },
+    {
+      "epoch": 0.4888963099014861,
+      "grad_norm": 0.6330185692498038,
+      "learning_rate": 9.879926947151164e-06,
+      "loss": 0.2461,
+      "step": 1464
+    },
+    {
+      "epoch": 0.48923025546835863,
+      "grad_norm": 0.5316986356744938,
+      "learning_rate": 9.879503250632991e-06,
+      "loss": 0.2424,
+      "step": 1465
+    },
+    {
+      "epoch": 0.48956420103523124,
+      "grad_norm": 0.8086554466946749,
+      "learning_rate": 9.879078817008395e-06,
+      "loss": 0.2381,
+      "step": 1466
+    },
+    {
+      "epoch": 0.48989814660210385,
+      "grad_norm": 0.6065660752029407,
+      "learning_rate": 9.878653646341498e-06,
+      "loss": 0.2331,
+      "step": 1467
+    },
+    {
+      "epoch": 0.49023209216897645,
+      "grad_norm": 0.557985159542511,
+      "learning_rate": 9.878227738696522e-06,
+      "loss": 0.2237,
+      "step": 1468
+    },
+    {
+      "epoch": 0.49056603773584906,
+      "grad_norm": 0.48904785077843377,
+      "learning_rate": 9.877801094137807e-06,
+      "loss": 0.2117,
+      "step": 1469
+    },
+    {
+      "epoch": 0.49089998330272167,
+      "grad_norm": 0.5750681520541658,
+      "learning_rate": 9.877373712729803e-06,
+      "loss": 0.2321,
+      "step": 1470
+    },
+    {
+      "epoch": 0.49123392886959427,
+      "grad_norm": 0.6187608854168688,
+      "learning_rate": 9.876945594537069e-06,
+      "loss": 0.2323,
+      "step": 1471
+    },
+    {
+      "epoch": 0.4915678744364669,
+      "grad_norm": 0.5914515580006867,
+      "learning_rate": 9.876516739624279e-06,
+      "loss": 0.2343,
+      "step": 1472
+    },
+    {
+      "epoch": 0.49190182000333943,
+      "grad_norm": 0.6132148683275411,
+      "learning_rate": 9.876087148056217e-06,
+      "loss": 0.2206,
+      "step": 1473
+    },
+    {
+      "epoch": 0.49223576557021204,
+      "grad_norm": 0.5779312466388966,
+      "learning_rate": 9.875656819897776e-06,
+      "loss": 0.2217,
+      "step": 1474
+    },
+    {
+      "epoch": 0.49256971113708464,
+      "grad_norm": 0.5934721585739572,
+      "learning_rate": 9.875225755213966e-06,
+      "loss": 0.2242,
+      "step": 1475
+    },
+    {
+      "epoch": 0.49290365670395725,
+      "grad_norm": 0.6090421834628303,
+      "learning_rate": 9.874793954069899e-06,
+      "loss": 0.2488,
+      "step": 1476
+    },
+    {
+      "epoch": 0.49323760227082986,
+      "grad_norm": 0.5063010044760622,
+      "learning_rate": 9.874361416530808e-06,
+      "loss": 0.2394,
+      "step": 1477
+    },
+    {
+      "epoch": 0.49357154783770246,
+      "grad_norm": 0.5439710083823438,
+      "learning_rate": 9.873928142662031e-06,
+      "loss": 0.2375,
+      "step": 1478
+    },
+    {
+      "epoch": 0.49390549340457507,
+      "grad_norm": 0.48631313293144357,
+      "learning_rate": 9.873494132529018e-06,
+      "loss": 0.2314,
+      "step": 1479
+    },
+    {
+      "epoch": 0.4942394389714477,
+      "grad_norm": 0.6393500940969762,
+      "learning_rate": 9.873059386197335e-06,
+      "loss": 0.2618,
+      "step": 1480
+    },
+    {
+      "epoch": 0.4945733845383202,
+      "grad_norm": 0.5827650304612245,
+      "learning_rate": 9.872623903732652e-06,
+      "loss": 0.2174,
+      "step": 1481
+    },
+    {
+      "epoch": 0.49490733010519283,
+      "grad_norm": 0.48147858509311164,
+      "learning_rate": 9.872187685200756e-06,
+      "loss": 0.2177,
+      "step": 1482
+    },
+    {
+      "epoch": 0.49524127567206544,
+      "grad_norm": 0.5515351930407174,
+      "learning_rate": 9.87175073066754e-06,
+      "loss": 0.24,
+      "step": 1483
+    },
+    {
+      "epoch": 0.49557522123893805,
+      "grad_norm": 0.565292330492309,
+      "learning_rate": 9.871313040199015e-06,
+      "loss": 0.2315,
+      "step": 1484
+    },
+    {
+      "epoch": 0.49590916680581065,
+      "grad_norm": 0.5179304120767915,
+      "learning_rate": 9.870874613861297e-06,
+      "loss": 0.2157,
+      "step": 1485
+    },
+    {
+      "epoch": 0.49624311237268326,
+      "grad_norm": 0.5509004443246928,
+      "learning_rate": 9.870435451720614e-06,
+      "loss": 0.2276,
+      "step": 1486
+    },
+    {
+      "epoch": 0.49657705793955587,
+      "grad_norm": 0.4892998005954115,
+      "learning_rate": 9.869995553843313e-06,
+      "loss": 0.2128,
+      "step": 1487
+    },
+    {
+      "epoch": 0.4969110035064285,
+      "grad_norm": 0.5037397253995877,
+      "learning_rate": 9.869554920295836e-06,
+      "loss": 0.2251,
+      "step": 1488
+    },
+    {
+      "epoch": 0.4972449490733011,
+      "grad_norm": 0.5227998261948305,
+      "learning_rate": 9.869113551144754e-06,
+      "loss": 0.2175,
+      "step": 1489
+    },
+    {
+      "epoch": 0.49757889464017363,
+      "grad_norm": 0.5995114670501726,
+      "learning_rate": 9.86867144645674e-06,
+      "loss": 0.2606,
+      "step": 1490
+    },
+    {
+      "epoch": 0.49791284020704624,
+      "grad_norm": 0.4795393090914556,
+      "learning_rate": 9.868228606298574e-06,
+      "loss": 0.2269,
+      "step": 1491
+    },
+    {
+      "epoch": 0.49824678577391884,
+      "grad_norm": 0.5497097986938829,
+      "learning_rate": 9.867785030737157e-06,
+      "loss": 0.2422,
+      "step": 1492
+    },
+    {
+      "epoch": 0.49858073134079145,
+      "grad_norm": 0.5503594416571509,
+      "learning_rate": 9.867340719839494e-06,
+      "loss": 0.241,
+      "step": 1493
+    },
+    {
+      "epoch": 0.49891467690766406,
+      "grad_norm": 0.7026689122560772,
+      "learning_rate": 9.866895673672704e-06,
+      "loss": 0.2523,
+      "step": 1494
+    },
+    {
+      "epoch": 0.49924862247453666,
+      "grad_norm": 0.5290404576726735,
+      "learning_rate": 9.866449892304017e-06,
+      "loss": 0.2233,
+      "step": 1495
+    },
+    {
+      "epoch": 0.49958256804140927,
+      "grad_norm": 0.4789461584020222,
+      "learning_rate": 9.866003375800773e-06,
+      "loss": 0.2275,
+      "step": 1496
+    },
+    {
+      "epoch": 0.4999165136082819,
+      "grad_norm": 0.5496686648273017,
+      "learning_rate": 9.865556124230425e-06,
+      "loss": 0.2269,
+      "step": 1497
+    },
+    {
+      "epoch": 0.5002504591751544,
+      "grad_norm": 0.5178066142061346,
+      "learning_rate": 9.865108137660533e-06,
+      "loss": 0.2352,
+      "step": 1498
+    },
+    {
+      "epoch": 0.500584404742027,
+      "grad_norm": 0.5198841589298615,
+      "learning_rate": 9.864659416158773e-06,
+      "loss": 0.2312,
+      "step": 1499
+    },
+    {
+      "epoch": 0.5009183503088996,
+      "grad_norm": 0.5388111513331779,
+      "learning_rate": 9.864209959792927e-06,
+      "loss": 0.2384,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5012522958757722,
+      "grad_norm": 0.6553768022329063,
+      "learning_rate": 9.863759768630893e-06,
+      "loss": 0.2506,
+      "step": 1501
+    },
+    {
+      "epoch": 0.5015862414426449,
+      "grad_norm": 0.5610454814274192,
+      "learning_rate": 9.863308842740678e-06,
+      "loss": 0.2247,
+      "step": 1502
+    },
+    {
+      "epoch": 0.5019201870095175,
+      "grad_norm": 0.5502060297653726,
+      "learning_rate": 9.862857182190398e-06,
+      "loss": 0.2342,
+      "step": 1503
+    },
+    {
+      "epoch": 0.5022541325763901,
+      "grad_norm": 0.4949388903158392,
+      "learning_rate": 9.862404787048283e-06,
+      "loss": 0.2201,
+      "step": 1504
+    },
+    {
+      "epoch": 0.5025880781432627,
+      "grad_norm": 0.5240359024269955,
+      "learning_rate": 9.861951657382671e-06,
+      "loss": 0.2426,
+      "step": 1505
+    },
+    {
+      "epoch": 0.5029220237101353,
+      "grad_norm": 0.5132026830298406,
+      "learning_rate": 9.861497793262014e-06,
+      "loss": 0.2245,
+      "step": 1506
+    },
+    {
+      "epoch": 0.5032559692770079,
+      "grad_norm": 0.5198774578302745,
+      "learning_rate": 9.861043194754874e-06,
+      "loss": 0.2403,
+      "step": 1507
+    },
+    {
+      "epoch": 0.5035899148438805,
+      "grad_norm": 0.5039524096753569,
+      "learning_rate": 9.860587861929922e-06,
+      "loss": 0.2159,
+      "step": 1508
+    },
+    {
+      "epoch": 0.5039238604107531,
+      "grad_norm": 0.5448793630792352,
+      "learning_rate": 9.86013179485594e-06,
+      "loss": 0.2316,
+      "step": 1509
+    },
+    {
+      "epoch": 0.5042578059776256,
+      "grad_norm": 0.48213974652147173,
+      "learning_rate": 9.859674993601826e-06,
+      "loss": 0.2242,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5045917515444982,
+      "grad_norm": 0.590315746105308,
+      "learning_rate": 9.859217458236583e-06,
+      "loss": 0.2384,
+      "step": 1511
+    },
+    {
+      "epoch": 0.5049256971113708,
+      "grad_norm": 0.5104158501610313,
+      "learning_rate": 9.858759188829328e-06,
+      "loss": 0.2426,
+      "step": 1512
+    },
+    {
+      "epoch": 0.5052596426782434,
+      "grad_norm": 0.582674382044229,
+      "learning_rate": 9.858300185449287e-06,
+      "loss": 0.2401,
+      "step": 1513
+    },
+    {
+      "epoch": 0.505593588245116,
+      "grad_norm": 0.5203261108265982,
+      "learning_rate": 9.857840448165798e-06,
+      "loss": 0.2391,
+      "step": 1514
+    },
+    {
+      "epoch": 0.5059275338119886,
+      "grad_norm": 0.48683141373358735,
+      "learning_rate": 9.857379977048311e-06,
+      "loss": 0.2185,
+      "step": 1515
+    },
+    {
+      "epoch": 0.5062614793788612,
+      "grad_norm": 0.5561097986159383,
+      "learning_rate": 9.856918772166385e-06,
+      "loss": 0.2403,
+      "step": 1516
+    },
+    {
+      "epoch": 0.5065954249457338,
+      "grad_norm": 0.6235558405935004,
+      "learning_rate": 9.856456833589688e-06,
+      "loss": 0.2183,
+      "step": 1517
+    },
+    {
+      "epoch": 0.5069293705126064,
+      "grad_norm": 0.6446845091009377,
+      "learning_rate": 9.855994161388005e-06,
+      "loss": 0.2333,
+      "step": 1518
+    },
+    {
+      "epoch": 0.507263316079479,
+      "grad_norm": 0.5687674110037545,
+      "learning_rate": 9.855530755631226e-06,
+      "loss": 0.2428,
+      "step": 1519
+    },
+    {
+      "epoch": 0.5075972616463517,
+      "grad_norm": 0.5587656168664682,
+      "learning_rate": 9.855066616389356e-06,
+      "loss": 0.2369,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5079312072132243,
+      "grad_norm": 0.7359238587828697,
+      "learning_rate": 9.854601743732504e-06,
+      "loss": 0.2317,
+      "step": 1521
+    },
+    {
+      "epoch": 0.5082651527800969,
+      "grad_norm": 0.5066229049650518,
+      "learning_rate": 9.854136137730899e-06,
+      "loss": 0.2337,
+      "step": 1522
+    },
+    {
+      "epoch": 0.5085990983469695,
+      "grad_norm": 0.49537075038118694,
+      "learning_rate": 9.853669798454875e-06,
+      "loss": 0.2243,
+      "step": 1523
+    },
+    {
+      "epoch": 0.5089330439138421,
+      "grad_norm": 0.6022772076474969,
+      "learning_rate": 9.853202725974878e-06,
+      "loss": 0.2415,
+      "step": 1524
+    },
+    {
+      "epoch": 0.5092669894807147,
+      "grad_norm": 0.608749379967523,
+      "learning_rate": 9.852734920361465e-06,
+      "loss": 0.2345,
+      "step": 1525
+    },
+    {
+      "epoch": 0.5096009350475872,
+      "grad_norm": 0.6715781075920919,
+      "learning_rate": 9.8522663816853e-06,
+      "loss": 0.2641,
+      "step": 1526
+    },
+    {
+      "epoch": 0.5099348806144598,
+      "grad_norm": 0.5101715269421332,
+      "learning_rate": 9.851797110017167e-06,
+      "loss": 0.2196,
+      "step": 1527
+    },
+    {
+      "epoch": 0.5102688261813324,
+      "grad_norm": 0.511537007454956,
+      "learning_rate": 9.851327105427952e-06,
+      "loss": 0.2133,
+      "step": 1528
+    },
+    {
+      "epoch": 0.510602771748205,
+      "grad_norm": 0.5178737587676562,
+      "learning_rate": 9.850856367988657e-06,
+      "loss": 0.2185,
+      "step": 1529
+    },
+    {
+      "epoch": 0.5109367173150776,
+      "grad_norm": 0.5942221037087895,
+      "learning_rate": 9.850384897770388e-06,
+      "loss": 0.245,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5112706628819502,
+      "grad_norm": 0.5444327388738013,
+      "learning_rate": 9.84991269484437e-06,
+      "loss": 0.2412,
+      "step": 1531
+    },
+    {
+      "epoch": 0.5116046084488228,
+      "grad_norm": 0.5352070717510221,
+      "learning_rate": 9.849439759281934e-06,
+      "loss": 0.2304,
+      "step": 1532
+    },
+    {
+      "epoch": 0.5119385540156954,
+      "grad_norm": 0.5602568244179051,
+      "learning_rate": 9.848966091154522e-06,
+      "loss": 0.238,
+      "step": 1533
+    },
+    {
+      "epoch": 0.512272499582568,
+      "grad_norm": 0.5211858362853864,
+      "learning_rate": 9.848491690533686e-06,
+      "loss": 0.2376,
+      "step": 1534
+    },
+    {
+      "epoch": 0.5126064451494406,
+      "grad_norm": 0.5783858287499946,
+      "learning_rate": 9.848016557491092e-06,
+      "loss": 0.2582,
+      "step": 1535
+    },
+    {
+      "epoch": 0.5129403907163133,
+      "grad_norm": 0.5915548412171439,
+      "learning_rate": 9.847540692098513e-06,
+      "loss": 0.2462,
+      "step": 1536
+    },
+    {
+      "epoch": 0.5132743362831859,
+      "grad_norm": 0.5380120667594149,
+      "learning_rate": 9.847064094427835e-06,
+      "loss": 0.244,
+      "step": 1537
+    },
+    {
+      "epoch": 0.5136082818500585,
+      "grad_norm": 0.5931387539258556,
+      "learning_rate": 9.846586764551054e-06,
+      "loss": 0.241,
+      "step": 1538
+    },
+    {
+      "epoch": 0.5139422274169311,
+      "grad_norm": 0.6732547388174122,
+      "learning_rate": 9.846108702540274e-06,
+      "loss": 0.2453,
+      "step": 1539
+    },
+    {
+      "epoch": 0.5142761729838037,
+      "grad_norm": 0.5708516751959783,
+      "learning_rate": 9.845629908467714e-06,
+      "loss": 0.2253,
+      "step": 1540
+    },
+    {
+      "epoch": 0.5146101185506763,
+      "grad_norm": 0.5964930644824759,
+      "learning_rate": 9.8451503824057e-06,
+      "loss": 0.205,
+      "step": 1541
+    },
+    {
+      "epoch": 0.5149440641175489,
+      "grad_norm": 0.5433778021010894,
+      "learning_rate": 9.844670124426672e-06,
+      "loss": 0.2151,
+      "step": 1542
+    },
+    {
+      "epoch": 0.5152780096844214,
+      "grad_norm": 0.5567634491740848,
+      "learning_rate": 9.844189134603178e-06,
+      "loss": 0.2154,
+      "step": 1543
+    },
+    {
+      "epoch": 0.515611955251294,
+      "grad_norm": 0.4903197026217307,
+      "learning_rate": 9.843707413007874e-06,
+      "loss": 0.2137,
+      "step": 1544
+    },
+    {
+      "epoch": 0.5159459008181666,
+      "grad_norm": 0.5960697433572092,
+      "learning_rate": 9.843224959713535e-06,
+      "loss": 0.2527,
+      "step": 1545
+    },
+    {
+      "epoch": 0.5162798463850392,
+      "grad_norm": 0.4944143034523633,
+      "learning_rate": 9.842741774793038e-06,
+      "loss": 0.2307,
+      "step": 1546
+    },
+    {
+      "epoch": 0.5166137919519118,
+      "grad_norm": 0.49358314195918296,
+      "learning_rate": 9.842257858319375e-06,
+      "loss": 0.2179,
+      "step": 1547
+    },
+    {
+      "epoch": 0.5169477375187844,
+      "grad_norm": 0.5796108269031679,
+      "learning_rate": 9.841773210365646e-06,
+      "loss": 0.2519,
+      "step": 1548
+    },
+    {
+      "epoch": 0.517281683085657,
+      "grad_norm": 0.5138397804293885,
+      "learning_rate": 9.841287831005064e-06,
+      "loss": 0.2305,
+      "step": 1549
+    },
+    {
+      "epoch": 0.5176156286525296,
+      "grad_norm": 0.5250071600094292,
+      "learning_rate": 9.84080172031095e-06,
+      "loss": 0.2311,
+      "step": 1550
+    },
+    {
+      "epoch": 0.5179495742194022,
+      "grad_norm": 0.5574764925023964,
+      "learning_rate": 9.840314878356739e-06,
+      "loss": 0.2323,
+      "step": 1551
+    },
+    {
+      "epoch": 0.5182835197862748,
+      "grad_norm": 0.6140408300146574,
+      "learning_rate": 9.839827305215972e-06,
+      "loss": 0.2533,
+      "step": 1552
+    },
+    {
+      "epoch": 0.5186174653531475,
+      "grad_norm": 0.7516633182539431,
+      "learning_rate": 9.839339000962305e-06,
+      "loss": 0.226,
+      "step": 1553
+    },
+    {
+      "epoch": 0.5189514109200201,
+      "grad_norm": 0.5458694547214934,
+      "learning_rate": 9.838849965669499e-06,
+      "loss": 0.2325,
+      "step": 1554
+    },
+    {
+      "epoch": 0.5192853564868927,
+      "grad_norm": 0.5064986538497875,
+      "learning_rate": 9.83836019941143e-06,
+      "loss": 0.2114,
+      "step": 1555
+    },
+    {
+      "epoch": 0.5196193020537653,
+      "grad_norm": 0.61319519408008,
+      "learning_rate": 9.837869702262082e-06,
+      "loss": 0.2473,
+      "step": 1556
+    },
+    {
+      "epoch": 0.5199532476206379,
+      "grad_norm": 0.5565828724070242,
+      "learning_rate": 9.837378474295553e-06,
+      "loss": 0.2332,
+      "step": 1557
+    },
+    {
+      "epoch": 0.5202871931875105,
+      "grad_norm": 0.5064088876577069,
+      "learning_rate": 9.836886515586045e-06,
+      "loss": 0.217,
+      "step": 1558
+    },
+    {
+      "epoch": 0.520621138754383,
+      "grad_norm": 0.536537012097158,
+      "learning_rate": 9.83639382620788e-06,
+      "loss": 0.2276,
+      "step": 1559
+    },
+    {
+      "epoch": 0.5209550843212556,
+      "grad_norm": 0.518315034919932,
+      "learning_rate": 9.835900406235479e-06,
+      "loss": 0.2247,
+      "step": 1560
+    },
+    {
+      "epoch": 0.5212890298881282,
+      "grad_norm": 0.5401437965180717,
+      "learning_rate": 9.835406255743381e-06,
+      "loss": 0.229,
+      "step": 1561
+    },
+    {
+      "epoch": 0.5216229754550008,
+      "grad_norm": 0.5105619541123207,
+      "learning_rate": 9.834911374806231e-06,
+      "loss": 0.2335,
+      "step": 1562
+    },
+    {
+      "epoch": 0.5219569210218734,
+      "grad_norm": 0.519164367586733,
+      "learning_rate": 9.83441576349879e-06,
+      "loss": 0.2251,
+      "step": 1563
+    },
+    {
+      "epoch": 0.522290866588746,
+      "grad_norm": 0.569853562629724,
+      "learning_rate": 9.833919421895926e-06,
+      "loss": 0.2537,
+      "step": 1564
+    },
+    {
+      "epoch": 0.5226248121556186,
+      "grad_norm": 0.516625707086611,
+      "learning_rate": 9.833422350072615e-06,
+      "loss": 0.2161,
+      "step": 1565
+    },
+    {
+      "epoch": 0.5229587577224912,
+      "grad_norm": 0.6535880193463475,
+      "learning_rate": 9.832924548103945e-06,
+      "loss": 0.209,
+      "step": 1566
+    },
+    {
+      "epoch": 0.5232927032893638,
+      "grad_norm": 0.4916515442829188,
+      "learning_rate": 9.832426016065117e-06,
+      "loss": 0.2185,
+      "step": 1567
+    },
+    {
+      "epoch": 0.5236266488562364,
+      "grad_norm": 0.5814664612841597,
+      "learning_rate": 9.83192675403144e-06,
+      "loss": 0.2266,
+      "step": 1568
+    },
+    {
+      "epoch": 0.523960594423109,
+      "grad_norm": 0.5884057167346974,
+      "learning_rate": 9.831426762078331e-06,
+      "loss": 0.2244,
+      "step": 1569
+    },
+    {
+      "epoch": 0.5242945399899817,
+      "grad_norm": 0.6183375261868602,
+      "learning_rate": 9.830926040281321e-06,
+      "loss": 0.2445,
+      "step": 1570
+    },
+    {
+      "epoch": 0.5246284855568543,
+      "grad_norm": 0.5200985380140108,
+      "learning_rate": 9.830424588716053e-06,
+      "loss": 0.2248,
+      "step": 1571
+    },
+    {
+      "epoch": 0.5249624311237269,
+      "grad_norm": 0.5420915104887184,
+      "learning_rate": 9.829922407458273e-06,
+      "loss": 0.2327,
+      "step": 1572
+    },
+    {
+      "epoch": 0.5252963766905995,
+      "grad_norm": 0.5270598323458139,
+      "learning_rate": 9.829419496583843e-06,
+      "loss": 0.2256,
+      "step": 1573
+    },
+    {
+      "epoch": 0.5256303222574721,
+      "grad_norm": 0.5239074446729693,
+      "learning_rate": 9.828915856168734e-06,
+      "loss": 0.2284,
+      "step": 1574
+    },
+    {
+      "epoch": 0.5259642678243446,
+      "grad_norm": 0.5489528638296536,
+      "learning_rate": 9.828411486289026e-06,
+      "loss": 0.2346,
+      "step": 1575
+    },
+    {
+      "epoch": 0.5262982133912172,
+      "grad_norm": 0.5306622558740911,
+      "learning_rate": 9.82790638702091e-06,
+      "loss": 0.2334,
+      "step": 1576
+    },
+    {
+      "epoch": 0.5266321589580898,
+      "grad_norm": 0.5020797546266,
+      "learning_rate": 9.827400558440687e-06,
+      "loss": 0.2193,
+      "step": 1577
+    },
+    {
+      "epoch": 0.5269661045249624,
+      "grad_norm": 0.4958858002471795,
+      "learning_rate": 9.826894000624769e-06,
+      "loss": 0.2214,
+      "step": 1578
+    },
+    {
+      "epoch": 0.527300050091835,
+      "grad_norm": 0.5495730199464439,
+      "learning_rate": 9.826386713649678e-06,
+      "loss": 0.2391,
+      "step": 1579
+    },
+    {
+      "epoch": 0.5276339956587076,
+      "grad_norm": 0.5337213938676789,
+      "learning_rate": 9.825878697592046e-06,
+      "loss": 0.241,
+      "step": 1580
+    },
+    {
+      "epoch": 0.5279679412255802,
+      "grad_norm": 0.5243139931047547,
+      "learning_rate": 9.825369952528611e-06,
+      "loss": 0.2347,
+      "step": 1581
+    },
+    {
+      "epoch": 0.5283018867924528,
+      "grad_norm": 0.46951090232168313,
+      "learning_rate": 9.824860478536231e-06,
+      "loss": 0.2174,
+      "step": 1582
+    },
+    {
+      "epoch": 0.5286358323593254,
+      "grad_norm": 0.5591143330075115,
+      "learning_rate": 9.824350275691864e-06,
+      "loss": 0.2315,
+      "step": 1583
+    },
+    {
+      "epoch": 0.528969777926198,
+      "grad_norm": 0.5741634648736227,
+      "learning_rate": 9.823839344072582e-06,
+      "loss": 0.2392,
+      "step": 1584
+    },
+    {
+      "epoch": 0.5293037234930706,
+      "grad_norm": 0.5477207037243178,
+      "learning_rate": 9.823327683755566e-06,
+      "loss": 0.2479,
+      "step": 1585
+    },
+    {
+      "epoch": 0.5296376690599433,
+      "grad_norm": 0.5814226357210571,
+      "learning_rate": 9.822815294818113e-06,
+      "loss": 0.2503,
+      "step": 1586
+    },
+    {
+      "epoch": 0.5299716146268159,
+      "grad_norm": 0.5959177044049923,
+      "learning_rate": 9.822302177337624e-06,
+      "loss": 0.2414,
+      "step": 1587
+    },
+    {
+      "epoch": 0.5303055601936885,
+      "grad_norm": 0.5034065372193987,
+      "learning_rate": 9.821788331391609e-06,
+      "loss": 0.226,
+      "step": 1588
+    },
+    {
+      "epoch": 0.5306395057605611,
+      "grad_norm": 0.5444917509379016,
+      "learning_rate": 9.821273757057692e-06,
+      "loss": 0.2208,
+      "step": 1589
+    },
+    {
+      "epoch": 0.5309734513274337,
+      "grad_norm": 0.6340433755540438,
+      "learning_rate": 9.820758454413606e-06,
+      "loss": 0.2341,
+      "step": 1590
+    },
+    {
+      "epoch": 0.5313073968943063,
+      "grad_norm": 0.5068735932793473,
+      "learning_rate": 9.820242423537192e-06,
+      "loss": 0.2279,
+      "step": 1591
+    },
+    {
+      "epoch": 0.5316413424611788,
+      "grad_norm": 0.5514835902750657,
+      "learning_rate": 9.819725664506404e-06,
+      "loss": 0.2397,
+      "step": 1592
+    },
+    {
+      "epoch": 0.5319752880280514,
+      "grad_norm": 0.540033557370152,
+      "learning_rate": 9.819208177399303e-06,
+      "loss": 0.2239,
+      "step": 1593
+    },
+    {
+      "epoch": 0.532309233594924,
+      "grad_norm": 0.592977095077727,
+      "learning_rate": 9.818689962294063e-06,
+      "loss": 0.2339,
+      "step": 1594
+    },
+    {
+      "epoch": 0.5326431791617966,
+      "grad_norm": 0.5871207671435377,
+      "learning_rate": 9.818171019268965e-06,
+      "loss": 0.2293,
+      "step": 1595
+    },
+    {
+      "epoch": 0.5329771247286692,
+      "grad_norm": 0.5947688900274528,
+      "learning_rate": 9.817651348402403e-06,
+      "loss": 0.2514,
+      "step": 1596
+    },
+    {
+      "epoch": 0.5333110702955418,
+      "grad_norm": 0.5812926758957463,
+      "learning_rate": 9.81713094977288e-06,
+      "loss": 0.2248,
+      "step": 1597
+    },
+    {
+      "epoch": 0.5336450158624144,
+      "grad_norm": 0.513970380059725,
+      "learning_rate": 9.816609823459007e-06,
+      "loss": 0.2375,
+      "step": 1598
+    },
+    {
+      "epoch": 0.533978961429287,
+      "grad_norm": 0.5034512417703262,
+      "learning_rate": 9.816087969539506e-06,
+      "loss": 0.2287,
+      "step": 1599
+    },
+    {
+      "epoch": 0.5343129069961596,
+      "grad_norm": 0.5550546878199349,
+      "learning_rate": 9.815565388093209e-06,
+      "loss": 0.2396,
+      "step": 1600
+    },
+    {
+      "epoch": 0.5346468525630322,
+      "grad_norm": 0.5324993847600048,
+      "learning_rate": 9.81504207919906e-06,
+      "loss": 0.2223,
+      "step": 1601
+    },
+    {
+      "epoch": 0.5349807981299048,
+      "grad_norm": 0.526275829556474,
+      "learning_rate": 9.814518042936107e-06,
+      "loss": 0.2431,
+      "step": 1602
+    },
+    {
+      "epoch": 0.5353147436967775,
+      "grad_norm": 0.5494997861868455,
+      "learning_rate": 9.813993279383518e-06,
+      "loss": 0.2323,
+      "step": 1603
+    },
+    {
+      "epoch": 0.5356486892636501,
+      "grad_norm": 0.5171193572022826,
+      "learning_rate": 9.813467788620559e-06,
+      "loss": 0.2416,
+      "step": 1604
+    },
+    {
+      "epoch": 0.5359826348305227,
+      "grad_norm": 0.5027893595121139,
+      "learning_rate": 9.812941570726615e-06,
+      "loss": 0.218,
+      "step": 1605
+    },
+    {
+      "epoch": 0.5363165803973953,
+      "grad_norm": 0.5506740906758446,
+      "learning_rate": 9.812414625781175e-06,
+      "loss": 0.2296,
+      "step": 1606
+    },
+    {
+      "epoch": 0.5366505259642679,
+      "grad_norm": 0.5183511110941164,
+      "learning_rate": 9.811886953863841e-06,
+      "loss": 0.2082,
+      "step": 1607
+    },
+    {
+      "epoch": 0.5369844715311404,
+      "grad_norm": 0.6302353066399081,
+      "learning_rate": 9.811358555054326e-06,
+      "loss": 0.2228,
+      "step": 1608
+    },
+    {
+      "epoch": 0.537318417098013,
+      "grad_norm": 0.5341795817076737,
+      "learning_rate": 9.810829429432449e-06,
+      "loss": 0.2293,
+      "step": 1609
+    },
+    {
+      "epoch": 0.5376523626648856,
+      "grad_norm": 0.5763059453482908,
+      "learning_rate": 9.81029957707814e-06,
+      "loss": 0.2558,
+      "step": 1610
+    },
+    {
+      "epoch": 0.5379863082317582,
+      "grad_norm": 0.6041426482444534,
+      "learning_rate": 9.809768998071442e-06,
+      "loss": 0.2293,
+      "step": 1611
+    },
+    {
+      "epoch": 0.5383202537986308,
+      "grad_norm": 0.5132151326478286,
+      "learning_rate": 9.809237692492503e-06,
+      "loss": 0.2185,
+      "step": 1612
+    },
+    {
+      "epoch": 0.5386541993655034,
+      "grad_norm": 0.5391826703922983,
+      "learning_rate": 9.808705660421582e-06,
+      "loss": 0.2323,
+      "step": 1613
+    },
+    {
+      "epoch": 0.538988144932376,
+      "grad_norm": 0.6461704927245718,
+      "learning_rate": 9.808172901939053e-06,
+      "loss": 0.2125,
+      "step": 1614
+    },
+    {
+      "epoch": 0.5393220904992486,
+      "grad_norm": 0.5270375186381666,
+      "learning_rate": 9.807639417125392e-06,
+      "loss": 0.2176,
+      "step": 1615
+    },
+    {
+      "epoch": 0.5396560360661212,
+      "grad_norm": 0.5445635894652113,
+      "learning_rate": 9.807105206061186e-06,
+      "loss": 0.2274,
+      "step": 1616
+    },
+    {
+      "epoch": 0.5399899816329938,
+      "grad_norm": 0.5317828887600755,
+      "learning_rate": 9.80657026882714e-06,
+      "loss": 0.2369,
+      "step": 1617
+    },
+    {
+      "epoch": 0.5403239271998664,
+      "grad_norm": 0.5686035453100612,
+      "learning_rate": 9.80603460550406e-06,
+      "loss": 0.2321,
+      "step": 1618
+    },
+    {
+      "epoch": 0.540657872766739,
+      "grad_norm": 0.527823856857519,
+      "learning_rate": 9.805498216172861e-06,
+      "loss": 0.2326,
+      "step": 1619
+    },
+    {
+      "epoch": 0.5409918183336117,
+      "grad_norm": 0.6946489211645083,
+      "learning_rate": 9.804961100914575e-06,
+      "loss": 0.2275,
+      "step": 1620
+    },
+    {
+      "epoch": 0.5413257639004843,
+      "grad_norm": 0.5021269916482256,
+      "learning_rate": 9.804423259810338e-06,
+      "loss": 0.217,
+      "step": 1621
+    },
+    {
+      "epoch": 0.5416597094673569,
+      "grad_norm": 0.5280955200768908,
+      "learning_rate": 9.803884692941397e-06,
+      "loss": 0.232,
+      "step": 1622
+    },
+    {
+      "epoch": 0.5419936550342295,
+      "grad_norm": 0.524502353180037,
+      "learning_rate": 9.803345400389111e-06,
+      "loss": 0.2295,
+      "step": 1623
+    },
+    {
+      "epoch": 0.542327600601102,
+      "grad_norm": 0.5588116006844829,
+      "learning_rate": 9.802805382234941e-06,
+      "loss": 0.2533,
+      "step": 1624
+    },
+    {
+      "epoch": 0.5426615461679746,
+      "grad_norm": 0.500852907399858,
+      "learning_rate": 9.80226463856047e-06,
+      "loss": 0.2134,
+      "step": 1625
+    },
+    {
+      "epoch": 0.5429954917348472,
+      "grad_norm": 0.6862541833794148,
+      "learning_rate": 9.801723169447378e-06,
+      "loss": 0.2506,
+      "step": 1626
+    },
+    {
+      "epoch": 0.5433294373017198,
+      "grad_norm": 0.5040386542518999,
+      "learning_rate": 9.801180974977466e-06,
+      "loss": 0.227,
+      "step": 1627
+    },
+    {
+      "epoch": 0.5436633828685924,
+      "grad_norm": 0.5474990228218318,
+      "learning_rate": 9.800638055232635e-06,
+      "loss": 0.2166,
+      "step": 1628
+    },
+    {
+      "epoch": 0.543997328435465,
+      "grad_norm": 0.6388226847856388,
+      "learning_rate": 9.800094410294897e-06,
+      "loss": 0.2487,
+      "step": 1629
+    },
+    {
+      "epoch": 0.5443312740023376,
+      "grad_norm": 0.4765248139449764,
+      "learning_rate": 9.799550040246381e-06,
+      "loss": 0.2129,
+      "step": 1630
+    },
+    {
+      "epoch": 0.5446652195692102,
+      "grad_norm": 0.5019451317873559,
+      "learning_rate": 9.799004945169319e-06,
+      "loss": 0.2038,
+      "step": 1631
+    },
+    {
+      "epoch": 0.5449991651360828,
+      "grad_norm": 0.48693504267132603,
+      "learning_rate": 9.798459125146054e-06,
+      "loss": 0.2379,
+      "step": 1632
+    },
+    {
+      "epoch": 0.5453331107029554,
+      "grad_norm": 0.5011793654105599,
+      "learning_rate": 9.797912580259037e-06,
+      "loss": 0.2278,
+      "step": 1633
+    },
+    {
+      "epoch": 0.545667056269828,
+      "grad_norm": 0.5627274706249237,
+      "learning_rate": 9.797365310590832e-06,
+      "loss": 0.2399,
+      "step": 1634
+    },
+    {
+      "epoch": 0.5460010018367006,
+      "grad_norm": 0.5317296411359453,
+      "learning_rate": 9.796817316224107e-06,
+      "loss": 0.2313,
+      "step": 1635
+    },
+    {
+      "epoch": 0.5463349474035732,
+      "grad_norm": 0.5256731560457311,
+      "learning_rate": 9.79626859724165e-06,
+      "loss": 0.2274,
+      "step": 1636
+    },
+    {
+      "epoch": 0.5466688929704459,
+      "grad_norm": 0.5295939011926792,
+      "learning_rate": 9.795719153726345e-06,
+      "loss": 0.2306,
+      "step": 1637
+    },
+    {
+      "epoch": 0.5470028385373185,
+      "grad_norm": 0.49722225435126244,
+      "learning_rate": 9.795168985761192e-06,
+      "loss": 0.2424,
+      "step": 1638
+    },
+    {
+      "epoch": 0.5473367841041911,
+      "grad_norm": 0.4782717883596217,
+      "learning_rate": 9.794618093429305e-06,
+      "loss": 0.2337,
+      "step": 1639
+    },
+    {
+      "epoch": 0.5476707296710637,
+      "grad_norm": 0.48131945296813267,
+      "learning_rate": 9.794066476813901e-06,
+      "loss": 0.2223,
+      "step": 1640
+    },
+    {
+      "epoch": 0.5480046752379362,
+      "grad_norm": 0.5320451343577421,
+      "learning_rate": 9.793514135998306e-06,
+      "loss": 0.2194,
+      "step": 1641
+    },
+    {
+      "epoch": 0.5483386208048088,
+      "grad_norm": 0.5662644769319405,
+      "learning_rate": 9.792961071065958e-06,
+      "loss": 0.2404,
+      "step": 1642
+    },
+    {
+      "epoch": 0.5486725663716814,
+      "grad_norm": 0.4708050063434642,
+      "learning_rate": 9.792407282100407e-06,
+      "loss": 0.2125,
+      "step": 1643
+    },
+    {
+      "epoch": 0.549006511938554,
+      "grad_norm": 0.5331092111658083,
+      "learning_rate": 9.791852769185306e-06,
+      "loss": 0.2317,
+      "step": 1644
+    },
+    {
+      "epoch": 0.5493404575054266,
+      "grad_norm": 0.46285348701283613,
+      "learning_rate": 9.791297532404422e-06,
+      "loss": 0.2188,
+      "step": 1645
+    },
+    {
+      "epoch": 0.5496744030722992,
+      "grad_norm": 0.6966380998201791,
+      "learning_rate": 9.790741571841629e-06,
+      "loss": 0.2434,
+      "step": 1646
+    },
+    {
+      "epoch": 0.5500083486391718,
+      "grad_norm": 0.6705149653085449,
+      "learning_rate": 9.790184887580914e-06,
+      "loss": 0.2545,
+      "step": 1647
+    },
+    {
+      "epoch": 0.5503422942060444,
+      "grad_norm": 0.5488413845730032,
+      "learning_rate": 9.78962747970637e-06,
+      "loss": 0.2192,
+      "step": 1648
+    },
+    {
+      "epoch": 0.550676239772917,
+      "grad_norm": 0.508399250930862,
+      "learning_rate": 9.789069348302197e-06,
+      "loss": 0.2147,
+      "step": 1649
+    },
+    {
+      "epoch": 0.5510101853397896,
+      "grad_norm": 0.5308431015969063,
+      "learning_rate": 9.78851049345271e-06,
+      "loss": 0.2451,
+      "step": 1650
+    },
+    {
+      "epoch": 0.5513441309066622,
+      "grad_norm": 0.5254441033154762,
+      "learning_rate": 9.78795091524233e-06,
+      "loss": 0.2307,
+      "step": 1651
+    },
+    {
+      "epoch": 0.5516780764735348,
+      "grad_norm": 0.49287285493077243,
+      "learning_rate": 9.78739061375559e-06,
+      "loss": 0.2279,
+      "step": 1652
+    },
+    {
+      "epoch": 0.5520120220404074,
+      "grad_norm": 0.6250540163648396,
+      "learning_rate": 9.786829589077125e-06,
+      "loss": 0.2568,
+      "step": 1653
+    },
+    {
+      "epoch": 0.55234596760728,
+      "grad_norm": 0.5757068773628706,
+      "learning_rate": 9.78626784129169e-06,
+      "loss": 0.2388,
+      "step": 1654
+    },
+    {
+      "epoch": 0.5526799131741527,
+      "grad_norm": 0.5131405555742891,
+      "learning_rate": 9.78570537048414e-06,
+      "loss": 0.2195,
+      "step": 1655
+    },
+    {
+      "epoch": 0.5530138587410253,
+      "grad_norm": 0.5273437954722322,
+      "learning_rate": 9.785142176739444e-06,
+      "loss": 0.234,
+      "step": 1656
+    },
+    {
+      "epoch": 0.5533478043078978,
+      "grad_norm": 0.5715667449885279,
+      "learning_rate": 9.784578260142679e-06,
+      "loss": 0.2364,
+      "step": 1657
+    },
+    {
+      "epoch": 0.5536817498747704,
+      "grad_norm": 0.4662686706134208,
+      "learning_rate": 9.784013620779031e-06,
+      "loss": 0.2174,
+      "step": 1658
+    },
+    {
+      "epoch": 0.554015695441643,
+      "grad_norm": 0.5573487695151922,
+      "learning_rate": 9.783448258733795e-06,
+      "loss": 0.2393,
+      "step": 1659
+    },
+    {
+      "epoch": 0.5543496410085156,
+      "grad_norm": 0.6004739211376822,
+      "learning_rate": 9.782882174092377e-06,
+      "loss": 0.2366,
+      "step": 1660
+    },
+    {
+      "epoch": 0.5546835865753882,
+      "grad_norm": 0.5130868339214583,
+      "learning_rate": 9.78231536694029e-06,
+      "loss": 0.2317,
+      "step": 1661
+    },
+    {
+      "epoch": 0.5550175321422608,
+      "grad_norm": 0.5081615667438504,
+      "learning_rate": 9.781747837363158e-06,
+      "loss": 0.2211,
+      "step": 1662
+    },
+    {
+      "epoch": 0.5553514777091334,
+      "grad_norm": 0.5243058551452536,
+      "learning_rate": 9.781179585446711e-06,
+      "loss": 0.2321,
+      "step": 1663
+    },
+    {
+      "epoch": 0.555685423276006,
+      "grad_norm": 0.5552330996626126,
+      "learning_rate": 9.780610611276791e-06,
+      "loss": 0.2263,
+      "step": 1664
+    },
+    {
+      "epoch": 0.5560193688428786,
+      "grad_norm": 0.5084354233773488,
+      "learning_rate": 9.780040914939349e-06,
+      "loss": 0.2209,
+      "step": 1665
+    },
+    {
+      "epoch": 0.5563533144097512,
+      "grad_norm": 0.8052644359842761,
+      "learning_rate": 9.779470496520442e-06,
+      "loss": 0.2786,
+      "step": 1666
+    },
+    {
+      "epoch": 0.5566872599766238,
+      "grad_norm": 0.5320336558276018,
+      "learning_rate": 9.77889935610624e-06,
+      "loss": 0.2182,
+      "step": 1667
+    },
+    {
+      "epoch": 0.5570212055434964,
+      "grad_norm": 0.5365736296465224,
+      "learning_rate": 9.778327493783022e-06,
+      "loss": 0.2251,
+      "step": 1668
+    },
+    {
+      "epoch": 0.557355151110369,
+      "grad_norm": 0.4630060106256875,
+      "learning_rate": 9.777754909637173e-06,
+      "loss": 0.2072,
+      "step": 1669
+    },
+    {
+      "epoch": 0.5576890966772416,
+      "grad_norm": 0.48881473250954816,
+      "learning_rate": 9.777181603755188e-06,
+      "loss": 0.2251,
+      "step": 1670
+    },
+    {
+      "epoch": 0.5580230422441143,
+      "grad_norm": 0.5416086012245569,
+      "learning_rate": 9.776607576223673e-06,
+      "loss": 0.2241,
+      "step": 1671
+    },
+    {
+      "epoch": 0.5583569878109869,
+      "grad_norm": 0.5213227615067991,
+      "learning_rate": 9.776032827129338e-06,
+      "loss": 0.221,
+      "step": 1672
+    },
+    {
+      "epoch": 0.5586909333778594,
+      "grad_norm": 0.5756279722517523,
+      "learning_rate": 9.775457356559013e-06,
+      "loss": 0.2274,
+      "step": 1673
+    },
+    {
+      "epoch": 0.559024878944732,
+      "grad_norm": 0.7857019129301895,
+      "learning_rate": 9.774881164599621e-06,
+      "loss": 0.2443,
+      "step": 1674
+    },
+    {
+      "epoch": 0.5593588245116046,
+      "grad_norm": 0.5730959574701345,
+      "learning_rate": 9.77430425133821e-06,
+      "loss": 0.2275,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5596927700784772,
+      "grad_norm": 0.5424650987984629,
+      "learning_rate": 9.773726616861926e-06,
+      "loss": 0.2386,
+      "step": 1676
+    },
+    {
+      "epoch": 0.5600267156453498,
+      "grad_norm": 0.527455714283128,
+      "learning_rate": 9.773148261258025e-06,
+      "loss": 0.2316,
+      "step": 1677
+    },
+    {
+      "epoch": 0.5603606612122224,
+      "grad_norm": 0.6038936532104409,
+      "learning_rate": 9.772569184613879e-06,
+      "loss": 0.2251,
+      "step": 1678
+    },
+    {
+      "epoch": 0.560694606779095,
+      "grad_norm": 0.500342452052048,
+      "learning_rate": 9.771989387016962e-06,
+      "loss": 0.2301,
+      "step": 1679
+    },
+    {
+      "epoch": 0.5610285523459676,
+      "grad_norm": 0.5596628524290133,
+      "learning_rate": 9.77140886855486e-06,
+      "loss": 0.2347,
+      "step": 1680
+    },
+    {
+      "epoch": 0.5613624979128402,
+      "grad_norm": 0.513227216859956,
+      "learning_rate": 9.770827629315266e-06,
+      "loss": 0.2316,
+      "step": 1681
+    },
+    {
+      "epoch": 0.5616964434797128,
+      "grad_norm": 0.5571244273244604,
+      "learning_rate": 9.770245669385984e-06,
+      "loss": 0.2246,
+      "step": 1682
+    },
+    {
+      "epoch": 0.5620303890465854,
+      "grad_norm": 0.5315880903884372,
+      "learning_rate": 9.76966298885493e-06,
+      "loss": 0.2215,
+      "step": 1683
+    },
+    {
+      "epoch": 0.562364334613458,
+      "grad_norm": 0.49598745769160424,
+      "learning_rate": 9.769079587810115e-06,
+      "loss": 0.2309,
+      "step": 1684
+    },
+    {
+      "epoch": 0.5626982801803306,
+      "grad_norm": 0.5013001295478747,
+      "learning_rate": 9.768495466339675e-06,
+      "loss": 0.2294,
+      "step": 1685
+    },
+    {
+      "epoch": 0.5630322257472032,
+      "grad_norm": 0.5009756499143495,
+      "learning_rate": 9.767910624531852e-06,
+      "loss": 0.2178,
+      "step": 1686
+    },
+    {
+      "epoch": 0.5633661713140758,
+      "grad_norm": 0.5041983827071985,
+      "learning_rate": 9.767325062474984e-06,
+      "loss": 0.2264,
+      "step": 1687
+    },
+    {
+      "epoch": 0.5637001168809485,
+      "grad_norm": 0.5388073393930474,
+      "learning_rate": 9.766738780257535e-06,
+      "loss": 0.2443,
+      "step": 1688
+    },
+    {
+      "epoch": 0.564034062447821,
+      "grad_norm": 0.4849109322874211,
+      "learning_rate": 9.766151777968063e-06,
+      "loss": 0.2238,
+      "step": 1689
+    },
+    {
+      "epoch": 0.5643680080146936,
+      "grad_norm": 0.5165836599020902,
+      "learning_rate": 9.765564055695249e-06,
+      "loss": 0.2492,
+      "step": 1690
+    },
+    {
+      "epoch": 0.5647019535815662,
+      "grad_norm": 0.5901306226307081,
+      "learning_rate": 9.76497561352787e-06,
+      "loss": 0.2369,
+      "step": 1691
+    },
+    {
+      "epoch": 0.5650358991484388,
+      "grad_norm": 0.529734231185276,
+      "learning_rate": 9.764386451554819e-06,
+      "loss": 0.2289,
+      "step": 1692
+    },
+    {
+      "epoch": 0.5653698447153114,
+      "grad_norm": 0.5307573390534124,
+      "learning_rate": 9.763796569865095e-06,
+      "loss": 0.2312,
+      "step": 1693
+    },
+    {
+      "epoch": 0.565703790282184,
+      "grad_norm": 0.5486503617208062,
+      "learning_rate": 9.763205968547808e-06,
+      "loss": 0.2162,
+      "step": 1694
+    },
+    {
+      "epoch": 0.5660377358490566,
+      "grad_norm": 0.5649817924362338,
+      "learning_rate": 9.762614647692175e-06,
+      "loss": 0.2191,
+      "step": 1695
+    },
+    {
+      "epoch": 0.5663716814159292,
+      "grad_norm": 0.533218563853688,
+      "learning_rate": 9.762022607387522e-06,
+      "loss": 0.2277,
+      "step": 1696
+    },
+    {
+      "epoch": 0.5667056269828018,
+      "grad_norm": 0.5396005837210106,
+      "learning_rate": 9.761429847723281e-06,
+      "loss": 0.251,
+      "step": 1697
+    },
+    {
+      "epoch": 0.5670395725496744,
+      "grad_norm": 0.6021708734947762,
+      "learning_rate": 9.760836368788999e-06,
+      "loss": 0.2298,
+      "step": 1698
+    },
+    {
+      "epoch": 0.567373518116547,
+      "grad_norm": 0.5571831043526276,
+      "learning_rate": 9.760242170674325e-06,
+      "loss": 0.221,
+      "step": 1699
+    },
+    {
+      "epoch": 0.5677074636834196,
+      "grad_norm": 0.6496040372874637,
+      "learning_rate": 9.759647253469023e-06,
+      "loss": 0.2518,
+      "step": 1700
+    },
+    {
+      "epoch": 0.5680414092502922,
+      "grad_norm": 0.5260316723253985,
+      "learning_rate": 9.75905161726296e-06,
+      "loss": 0.2489,
+      "step": 1701
+    },
+    {
+      "epoch": 0.5683753548171648,
+      "grad_norm": 0.522847565423199,
+      "learning_rate": 9.758455262146114e-06,
+      "loss": 0.2379,
+      "step": 1702
+    },
+    {
+      "epoch": 0.5687093003840374,
+      "grad_norm": 0.5215908948559349,
+      "learning_rate": 9.757858188208571e-06,
+      "loss": 0.2267,
+      "step": 1703
+    },
+    {
+      "epoch": 0.56904324595091,
+      "grad_norm": 0.4937749913698861,
+      "learning_rate": 9.757260395540527e-06,
+      "loss": 0.2167,
+      "step": 1704
+    },
+    {
+      "epoch": 0.5693771915177827,
+      "grad_norm": 0.5231134607360919,
+      "learning_rate": 9.756661884232286e-06,
+      "loss": 0.2288,
+      "step": 1705
+    },
+    {
+      "epoch": 0.5697111370846552,
+      "grad_norm": 0.48578184135466346,
+      "learning_rate": 9.756062654374259e-06,
+      "loss": 0.2254,
+      "step": 1706
+    },
+    {
+      "epoch": 0.5700450826515278,
+      "grad_norm": 0.4979412299848651,
+      "learning_rate": 9.755462706056966e-06,
+      "loss": 0.2266,
+      "step": 1707
+    },
+    {
+      "epoch": 0.5703790282184004,
+      "grad_norm": 0.5441091075782964,
+      "learning_rate": 9.75486203937104e-06,
+      "loss": 0.2193,
+      "step": 1708
+    },
+    {
+      "epoch": 0.570712973785273,
+      "grad_norm": 0.5216793063707312,
+      "learning_rate": 9.754260654407214e-06,
+      "loss": 0.2231,
+      "step": 1709
+    },
+    {
+      "epoch": 0.5710469193521456,
+      "grad_norm": 0.5004552915915513,
+      "learning_rate": 9.753658551256338e-06,
+      "loss": 0.2399,
+      "step": 1710
+    },
+    {
+      "epoch": 0.5713808649190182,
+      "grad_norm": 0.4875294674870155,
+      "learning_rate": 9.753055730009364e-06,
+      "loss": 0.2258,
+      "step": 1711
+    },
+    {
+      "epoch": 0.5717148104858908,
+      "grad_norm": 0.6035412634336383,
+      "learning_rate": 9.752452190757358e-06,
+      "loss": 0.2284,
+      "step": 1712
+    },
+    {
+      "epoch": 0.5720487560527634,
+      "grad_norm": 0.4665053445860015,
+      "learning_rate": 9.751847933591489e-06,
+      "loss": 0.2112,
+      "step": 1713
+    },
+    {
+      "epoch": 0.572382701619636,
+      "grad_norm": 0.4892405263993013,
+      "learning_rate": 9.75124295860304e-06,
+      "loss": 0.228,
+      "step": 1714
+    },
+    {
+      "epoch": 0.5727166471865086,
+      "grad_norm": 0.47689612627770583,
+      "learning_rate": 9.750637265883395e-06,
+      "loss": 0.2217,
+      "step": 1715
+    },
+    {
+      "epoch": 0.5730505927533812,
+      "grad_norm": 0.5851060256324795,
+      "learning_rate": 9.750030855524058e-06,
+      "loss": 0.2308,
+      "step": 1716
+    },
+    {
+      "epoch": 0.5733845383202538,
+      "grad_norm": 0.4739718971007316,
+      "learning_rate": 9.749423727616628e-06,
+      "loss": 0.2276,
+      "step": 1717
+    },
+    {
+      "epoch": 0.5737184838871264,
+      "grad_norm": 0.5130343462285478,
+      "learning_rate": 9.748815882252823e-06,
+      "loss": 0.2379,
+      "step": 1718
+    },
+    {
+      "epoch": 0.574052429453999,
+      "grad_norm": 0.5085916890670312,
+      "learning_rate": 9.748207319524462e-06,
+      "loss": 0.2238,
+      "step": 1719
+    },
+    {
+      "epoch": 0.5743863750208716,
+      "grad_norm": 0.5028286185116214,
+      "learning_rate": 9.747598039523476e-06,
+      "loss": 0.2195,
+      "step": 1720
+    },
+    {
+      "epoch": 0.5747203205877442,
+      "grad_norm": 0.5672833426547657,
+      "learning_rate": 9.746988042341907e-06,
+      "loss": 0.2429,
+      "step": 1721
+    },
+    {
+      "epoch": 0.5750542661546167,
+      "grad_norm": 0.5485257563867748,
+      "learning_rate": 9.746377328071899e-06,
+      "loss": 0.2126,
+      "step": 1722
+    },
+    {
+      "epoch": 0.5753882117214894,
+      "grad_norm": 0.5472794541807945,
+      "learning_rate": 9.74576589680571e-06,
+      "loss": 0.2257,
+      "step": 1723
+    },
+    {
+      "epoch": 0.575722157288362,
+      "grad_norm": 0.6005489214038497,
+      "learning_rate": 9.745153748635702e-06,
+      "loss": 0.2297,
+      "step": 1724
+    },
+    {
+      "epoch": 0.5760561028552346,
+      "grad_norm": 0.5331745130852713,
+      "learning_rate": 9.744540883654348e-06,
+      "loss": 0.2299,
+      "step": 1725
+    },
+    {
+      "epoch": 0.5763900484221072,
+      "grad_norm": 0.5467918769216435,
+      "learning_rate": 9.743927301954229e-06,
+      "loss": 0.2434,
+      "step": 1726
+    },
+    {
+      "epoch": 0.5767239939889798,
+      "grad_norm": 0.47631200906324167,
+      "learning_rate": 9.743313003628033e-06,
+      "loss": 0.2294,
+      "step": 1727
+    },
+    {
+      "epoch": 0.5770579395558524,
+      "grad_norm": 1.028042272396889,
+      "learning_rate": 9.742697988768557e-06,
+      "loss": 0.2286,
+      "step": 1728
+    },
+    {
+      "epoch": 0.577391885122725,
+      "grad_norm": 0.5723656104484236,
+      "learning_rate": 9.742082257468705e-06,
+      "loss": 0.2328,
+      "step": 1729
+    },
+    {
+      "epoch": 0.5777258306895976,
+      "grad_norm": 0.5658357107734303,
+      "learning_rate": 9.741465809821493e-06,
+      "loss": 0.2206,
+      "step": 1730
+    },
+    {
+      "epoch": 0.5780597762564702,
+      "grad_norm": 0.5303347638147783,
+      "learning_rate": 9.74084864592004e-06,
+      "loss": 0.2382,
+      "step": 1731
+    },
+    {
+      "epoch": 0.5783937218233428,
+      "grad_norm": 0.5591531368062704,
+      "learning_rate": 9.74023076585758e-06,
+      "loss": 0.2216,
+      "step": 1732
+    },
+    {
+      "epoch": 0.5787276673902154,
+      "grad_norm": 0.5152285449211208,
+      "learning_rate": 9.739612169727446e-06,
+      "loss": 0.2016,
+      "step": 1733
+    },
+    {
+      "epoch": 0.579061612957088,
+      "grad_norm": 0.5240518968639969,
+      "learning_rate": 9.73899285762309e-06,
+      "loss": 0.236,
+      "step": 1734
+    },
+    {
+      "epoch": 0.5793955585239606,
+      "grad_norm": 0.5941452749632281,
+      "learning_rate": 9.738372829638058e-06,
+      "loss": 0.2208,
+      "step": 1735
+    },
+    {
+      "epoch": 0.5797295040908332,
+      "grad_norm": 0.5383582980347633,
+      "learning_rate": 9.73775208586602e-06,
+      "loss": 0.2231,
+      "step": 1736
+    },
+    {
+      "epoch": 0.5800634496577058,
+      "grad_norm": 0.5097873416009885,
+      "learning_rate": 9.737130626400745e-06,
+      "loss": 0.2214,
+      "step": 1737
+    },
+    {
+      "epoch": 0.5803973952245783,
+      "grad_norm": 0.6409001150404144,
+      "learning_rate": 9.736508451336111e-06,
+      "loss": 0.2565,
+      "step": 1738
+    },
+    {
+      "epoch": 0.580731340791451,
+      "grad_norm": 0.5171820969817694,
+      "learning_rate": 9.735885560766104e-06,
+      "loss": 0.2247,
+      "step": 1739
+    },
+    {
+      "epoch": 0.5810652863583236,
+      "grad_norm": 0.565262498850217,
+      "learning_rate": 9.73526195478482e-06,
+      "loss": 0.2525,
+      "step": 1740
+    },
+    {
+      "epoch": 0.5813992319251962,
+      "grad_norm": 0.5191663542448296,
+      "learning_rate": 9.73463763348646e-06,
+      "loss": 0.2269,
+      "step": 1741
+    },
+    {
+      "epoch": 0.5817331774920688,
+      "grad_norm": 0.5008213494455472,
+      "learning_rate": 9.734012596965341e-06,
+      "loss": 0.2292,
+      "step": 1742
+    },
+    {
+      "epoch": 0.5820671230589414,
+      "grad_norm": 0.4985166781819941,
+      "learning_rate": 9.733386845315875e-06,
+      "loss": 0.2251,
+      "step": 1743
+    },
+    {
+      "epoch": 0.582401068625814,
+      "grad_norm": 0.7068726658287037,
+      "learning_rate": 9.732760378632592e-06,
+      "loss": 0.2575,
+      "step": 1744
+    },
+    {
+      "epoch": 0.5827350141926866,
+      "grad_norm": 0.532950090400461,
+      "learning_rate": 9.73213319701013e-06,
+      "loss": 0.2203,
+      "step": 1745
+    },
+    {
+      "epoch": 0.5830689597595592,
+      "grad_norm": 0.4870910187814131,
+      "learning_rate": 9.731505300543228e-06,
+      "loss": 0.2235,
+      "step": 1746
+    },
+    {
+      "epoch": 0.5834029053264318,
+      "grad_norm": 0.5183483007891654,
+      "learning_rate": 9.730876689326739e-06,
+      "loss": 0.2391,
+      "step": 1747
+    },
+    {
+      "epoch": 0.5837368508933044,
+      "grad_norm": 0.5822886754220148,
+      "learning_rate": 9.730247363455621e-06,
+      "loss": 0.246,
+      "step": 1748
+    },
+    {
+      "epoch": 0.584070796460177,
+      "grad_norm": 0.5303604700191408,
+      "learning_rate": 9.729617323024943e-06,
+      "loss": 0.2161,
+      "step": 1749
+    },
+    {
+      "epoch": 0.5844047420270496,
+      "grad_norm": 0.5402294666419608,
+      "learning_rate": 9.728986568129876e-06,
+      "loss": 0.2165,
+      "step": 1750
+    },
+    {
+      "epoch": 0.5847386875939222,
+      "grad_norm": 0.5229291505268835,
+      "learning_rate": 9.72835509886571e-06,
+      "loss": 0.2487,
+      "step": 1751
+    },
+    {
+      "epoch": 0.5850726331607948,
+      "grad_norm": 1.005248107058838,
+      "learning_rate": 9.727722915327828e-06,
+      "loss": 0.2344,
+      "step": 1752
+    },
+    {
+      "epoch": 0.5854065787276674,
+      "grad_norm": 0.4984992426091078,
+      "learning_rate": 9.727090017611736e-06,
+      "loss": 0.196,
+      "step": 1753
+    },
+    {
+      "epoch": 0.58574052429454,
+      "grad_norm": 0.5333276983734322,
+      "learning_rate": 9.726456405813033e-06,
+      "loss": 0.2264,
+      "step": 1754
+    },
+    {
+      "epoch": 0.5860744698614125,
+      "grad_norm": 0.45641453158909184,
+      "learning_rate": 9.725822080027442e-06,
+      "loss": 0.2239,
+      "step": 1755
+    },
+    {
+      "epoch": 0.5864084154282851,
+      "grad_norm": 0.48841035707321817,
+      "learning_rate": 9.725187040350778e-06,
+      "loss": 0.222,
+      "step": 1756
+    },
+    {
+      "epoch": 0.5867423609951578,
+      "grad_norm": 0.4741064605363138,
+      "learning_rate": 9.724551286878976e-06,
+      "loss": 0.2108,
+      "step": 1757
+    },
+    {
+      "epoch": 0.5870763065620304,
+      "grad_norm": 0.502475937380326,
+      "learning_rate": 9.723914819708073e-06,
+      "loss": 0.2253,
+      "step": 1758
+    },
+    {
+      "epoch": 0.587410252128903,
+      "grad_norm": 0.5194244729847486,
+      "learning_rate": 9.723277638934212e-06,
+      "loss": 0.2341,
+      "step": 1759
+    },
+    {
+      "epoch": 0.5877441976957756,
+      "grad_norm": 0.4916149636167516,
+      "learning_rate": 9.72263974465365e-06,
+      "loss": 0.2306,
+      "step": 1760
+    },
+    {
+      "epoch": 0.5880781432626482,
+      "grad_norm": 0.5023458268143842,
+      "learning_rate": 9.722001136962746e-06,
+      "loss": 0.2227,
+      "step": 1761
+    },
+    {
+      "epoch": 0.5884120888295208,
+      "grad_norm": 0.5608383767110363,
+      "learning_rate": 9.721361815957973e-06,
+      "loss": 0.2491,
+      "step": 1762
+    },
+    {
+      "epoch": 0.5887460343963934,
+      "grad_norm": 0.5144301238589608,
+      "learning_rate": 9.720721781735905e-06,
+      "loss": 0.2222,
+      "step": 1763
+    },
+    {
+      "epoch": 0.589079979963266,
+      "grad_norm": 0.5018433935726713,
+      "learning_rate": 9.720081034393226e-06,
+      "loss": 0.2258,
+      "step": 1764
+    },
+    {
+      "epoch": 0.5894139255301386,
+      "grad_norm": 0.49790563062526416,
+      "learning_rate": 9.71943957402673e-06,
+      "loss": 0.2285,
+      "step": 1765
+    },
+    {
+      "epoch": 0.5897478710970112,
+      "grad_norm": 0.4748861509241753,
+      "learning_rate": 9.718797400733314e-06,
+      "loss": 0.218,
+      "step": 1766
+    },
+    {
+      "epoch": 0.5900818166638838,
+      "grad_norm": 0.5390774391581642,
+      "learning_rate": 9.718154514609992e-06,
+      "loss": 0.2286,
+      "step": 1767
+    },
+    {
+      "epoch": 0.5904157622307564,
+      "grad_norm": 0.5536961578875585,
+      "learning_rate": 9.717510915753876e-06,
+      "loss": 0.2402,
+      "step": 1768
+    },
+    {
+      "epoch": 0.590749707797629,
+      "grad_norm": 0.6695127415098429,
+      "learning_rate": 9.716866604262189e-06,
+      "loss": 0.2235,
+      "step": 1769
+    },
+    {
+      "epoch": 0.5910836533645016,
+      "grad_norm": 0.5069772545716459,
+      "learning_rate": 9.716221580232261e-06,
+      "loss": 0.2201,
+      "step": 1770
+    },
+    {
+      "epoch": 0.5914175989313741,
+      "grad_norm": 0.8037833620387766,
+      "learning_rate": 9.715575843761534e-06,
+      "loss": 0.231,
+      "step": 1771
+    },
+    {
+      "epoch": 0.5917515444982467,
+      "grad_norm": 0.5068062069274792,
+      "learning_rate": 9.714929394947548e-06,
+      "loss": 0.2211,
+      "step": 1772
+    },
+    {
+      "epoch": 0.5920854900651193,
+      "grad_norm": 0.5549550762490882,
+      "learning_rate": 9.714282233887962e-06,
+      "loss": 0.2393,
+      "step": 1773
+    },
+    {
+      "epoch": 0.592419435631992,
+      "grad_norm": 0.5395976419004379,
+      "learning_rate": 9.713634360680537e-06,
+      "loss": 0.2252,
+      "step": 1774
+    },
+    {
+      "epoch": 0.5927533811988646,
+      "grad_norm": 0.4892732346046282,
+      "learning_rate": 9.712985775423141e-06,
+      "loss": 0.2193,
+      "step": 1775
+    },
+    {
+      "epoch": 0.5930873267657372,
+      "grad_norm": 0.4711153870554155,
+      "learning_rate": 9.712336478213747e-06,
+      "loss": 0.2315,
+      "step": 1776
+    },
+    {
+      "epoch": 0.5934212723326098,
+      "grad_norm": 0.4762551984595192,
+      "learning_rate": 9.711686469150444e-06,
+      "loss": 0.2355,
+      "step": 1777
+    },
+    {
+      "epoch": 0.5937552178994824,
+      "grad_norm": 0.539286037756309,
+      "learning_rate": 9.711035748331421e-06,
+      "loss": 0.2305,
+      "step": 1778
+    },
+    {
+      "epoch": 0.594089163466355,
+      "grad_norm": 0.44306089667586035,
+      "learning_rate": 9.710384315854977e-06,
+      "loss": 0.2119,
+      "step": 1779
+    },
+    {
+      "epoch": 0.5944231090332276,
+      "grad_norm": 0.54924454544785,
+      "learning_rate": 9.70973217181952e-06,
+      "loss": 0.2452,
+      "step": 1780
+    },
+    {
+      "epoch": 0.5947570546001002,
+      "grad_norm": 0.6928470758099178,
+      "learning_rate": 9.709079316323564e-06,
+      "loss": 0.2344,
+      "step": 1781
+    },
+    {
+      "epoch": 0.5950910001669728,
+      "grad_norm": 0.482232968127417,
+      "learning_rate": 9.70842574946573e-06,
+      "loss": 0.2207,
+      "step": 1782
+    },
+    {
+      "epoch": 0.5954249457338454,
+      "grad_norm": 0.49136177006387877,
+      "learning_rate": 9.707771471344744e-06,
+      "loss": 0.2261,
+      "step": 1783
+    },
+    {
+      "epoch": 0.595758891300718,
+      "grad_norm": 0.46768352694125126,
+      "learning_rate": 9.707116482059447e-06,
+      "loss": 0.2336,
+      "step": 1784
+    },
+    {
+      "epoch": 0.5960928368675906,
+      "grad_norm": 0.5029578869306203,
+      "learning_rate": 9.70646078170878e-06,
+      "loss": 0.2299,
+      "step": 1785
+    },
+    {
+      "epoch": 0.5964267824344632,
+      "grad_norm": 0.6031061149189456,
+      "learning_rate": 9.705804370391794e-06,
+      "loss": 0.2338,
+      "step": 1786
+    },
+    {
+      "epoch": 0.5967607280013357,
+      "grad_norm": 0.5619990384851195,
+      "learning_rate": 9.705147248207652e-06,
+      "loss": 0.2458,
+      "step": 1787
+    },
+    {
+      "epoch": 0.5970946735682083,
+      "grad_norm": 0.481566798104976,
+      "learning_rate": 9.704489415255614e-06,
+      "loss": 0.2272,
+      "step": 1788
+    },
+    {
+      "epoch": 0.5974286191350809,
+      "grad_norm": 0.5094230596305139,
+      "learning_rate": 9.703830871635057e-06,
+      "loss": 0.2366,
+      "step": 1789
+    },
+    {
+      "epoch": 0.5977625647019535,
+      "grad_norm": 0.5032186806617618,
+      "learning_rate": 9.703171617445461e-06,
+      "loss": 0.2378,
+      "step": 1790
+    },
+    {
+      "epoch": 0.5980965102688262,
+      "grad_norm": 0.5174151813237029,
+      "learning_rate": 9.702511652786414e-06,
+      "loss": 0.2194,
+      "step": 1791
+    },
+    {
+      "epoch": 0.5984304558356988,
+      "grad_norm": 0.5280668787057868,
+      "learning_rate": 9.701850977757611e-06,
+      "loss": 0.2236,
+      "step": 1792
+    },
+    {
+      "epoch": 0.5987644014025714,
+      "grad_norm": 0.5381616037918245,
+      "learning_rate": 9.701189592458858e-06,
+      "loss": 0.2243,
+      "step": 1793
+    },
+    {
+      "epoch": 0.599098346969444,
+      "grad_norm": 0.5404391079985676,
+      "learning_rate": 9.70052749699006e-06,
+      "loss": 0.2281,
+      "step": 1794
+    },
+    {
+      "epoch": 0.5994322925363166,
+      "grad_norm": 0.5628113540045915,
+      "learning_rate": 9.699864691451236e-06,
+      "loss": 0.23,
+      "step": 1795
+    },
+    {
+      "epoch": 0.5997662381031892,
+      "grad_norm": 0.5329453587759123,
+      "learning_rate": 9.699201175942514e-06,
+      "loss": 0.215,
+      "step": 1796
+    },
+    {
+      "epoch": 0.6001001836700618,
+      "grad_norm": 0.4718836619269112,
+      "learning_rate": 9.698536950564121e-06,
+      "loss": 0.2123,
+      "step": 1797
+    },
+    {
+      "epoch": 0.6004341292369344,
+      "grad_norm": 0.5539291679398612,
+      "learning_rate": 9.6978720154164e-06,
+      "loss": 0.2108,
+      "step": 1798
+    },
+    {
+      "epoch": 0.600768074803807,
+      "grad_norm": 0.5144961691465181,
+      "learning_rate": 9.697206370599793e-06,
+      "loss": 0.2322,
+      "step": 1799
+    },
+    {
+      "epoch": 0.6011020203706796,
+      "grad_norm": 0.5731553026791534,
+      "learning_rate": 9.696540016214857e-06,
+      "loss": 0.2286,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6014359659375522,
+      "grad_norm": 0.5188370628082019,
+      "learning_rate": 9.695872952362253e-06,
+      "loss": 0.2284,
+      "step": 1801
+    },
+    {
+      "epoch": 0.6017699115044248,
+      "grad_norm": 0.5671772412519566,
+      "learning_rate": 9.695205179142746e-06,
+      "loss": 0.2338,
+      "step": 1802
+    },
+    {
+      "epoch": 0.6021038570712974,
+      "grad_norm": 0.5086216193765247,
+      "learning_rate": 9.694536696657213e-06,
+      "loss": 0.2197,
+      "step": 1803
+    },
+    {
+      "epoch": 0.6024378026381699,
+      "grad_norm": 0.5238150472595622,
+      "learning_rate": 9.693867505006634e-06,
+      "loss": 0.2209,
+      "step": 1804
+    },
+    {
+      "epoch": 0.6027717482050425,
+      "grad_norm": 0.49537223803612296,
+      "learning_rate": 9.693197604292101e-06,
+      "loss": 0.2314,
+      "step": 1805
+    },
+    {
+      "epoch": 0.6031056937719151,
+      "grad_norm": 0.537065895743497,
+      "learning_rate": 9.69252699461481e-06,
+      "loss": 0.2411,
+      "step": 1806
+    },
+    {
+      "epoch": 0.6034396393387877,
+      "grad_norm": 0.5685097570680073,
+      "learning_rate": 9.691855676076064e-06,
+      "loss": 0.24,
+      "step": 1807
+    },
+    {
+      "epoch": 0.6037735849056604,
+      "grad_norm": 0.48659737235918643,
+      "learning_rate": 9.691183648777271e-06,
+      "loss": 0.217,
+      "step": 1808
+    },
+    {
+      "epoch": 0.604107530472533,
+      "grad_norm": 0.48629641393278755,
+      "learning_rate": 9.690510912819952e-06,
+      "loss": 0.2199,
+      "step": 1809
+    },
+    {
+      "epoch": 0.6044414760394056,
+      "grad_norm": 0.547728598895119,
+      "learning_rate": 9.689837468305732e-06,
+      "loss": 0.2299,
+      "step": 1810
+    },
+    {
+      "epoch": 0.6047754216062782,
+      "grad_norm": 0.5394439412865157,
+      "learning_rate": 9.689163315336339e-06,
+      "loss": 0.243,
+      "step": 1811
+    },
+    {
+      "epoch": 0.6051093671731508,
+      "grad_norm": 0.5103261884061013,
+      "learning_rate": 9.688488454013616e-06,
+      "loss": 0.2388,
+      "step": 1812
+    },
+    {
+      "epoch": 0.6054433127400234,
+      "grad_norm": 0.5099890466296649,
+      "learning_rate": 9.687812884439506e-06,
+      "loss": 0.2344,
+      "step": 1813
+    },
+    {
+      "epoch": 0.605777258306896,
+      "grad_norm": 0.46788777418989,
+      "learning_rate": 9.687136606716064e-06,
+      "loss": 0.2187,
+      "step": 1814
+    },
+    {
+      "epoch": 0.6061112038737686,
+      "grad_norm": 0.5249164624914014,
+      "learning_rate": 9.686459620945445e-06,
+      "loss": 0.2227,
+      "step": 1815
+    },
+    {
+      "epoch": 0.6064451494406412,
+      "grad_norm": 0.5331457391545483,
+      "learning_rate": 9.685781927229923e-06,
+      "loss": 0.2443,
+      "step": 1816
+    },
+    {
+      "epoch": 0.6067790950075138,
+      "grad_norm": 0.5886242608442575,
+      "learning_rate": 9.685103525671864e-06,
+      "loss": 0.2531,
+      "step": 1817
+    },
+    {
+      "epoch": 0.6071130405743864,
+      "grad_norm": 0.49860326082795825,
+      "learning_rate": 9.684424416373754e-06,
+      "loss": 0.2254,
+      "step": 1818
+    },
+    {
+      "epoch": 0.607446986141259,
+      "grad_norm": 0.5325971197360724,
+      "learning_rate": 9.683744599438178e-06,
+      "loss": 0.2294,
+      "step": 1819
+    },
+    {
+      "epoch": 0.6077809317081315,
+      "grad_norm": 0.5084348336824585,
+      "learning_rate": 9.683064074967832e-06,
+      "loss": 0.2375,
+      "step": 1820
+    },
+    {
+      "epoch": 0.6081148772750041,
+      "grad_norm": 0.46827119088454616,
+      "learning_rate": 9.682382843065516e-06,
+      "loss": 0.2146,
+      "step": 1821
+    },
+    {
+      "epoch": 0.6084488228418767,
+      "grad_norm": 0.5179404560268646,
+      "learning_rate": 9.681700903834137e-06,
+      "loss": 0.2324,
+      "step": 1822
+    },
+    {
+      "epoch": 0.6087827684087493,
+      "grad_norm": 0.5482754379272196,
+      "learning_rate": 9.681018257376713e-06,
+      "loss": 0.2266,
+      "step": 1823
+    },
+    {
+      "epoch": 0.609116713975622,
+      "grad_norm": 0.5402115399595885,
+      "learning_rate": 9.680334903796363e-06,
+      "loss": 0.2437,
+      "step": 1824
+    },
+    {
+      "epoch": 0.6094506595424946,
+      "grad_norm": 0.485938659979222,
+      "learning_rate": 9.679650843196318e-06,
+      "loss": 0.2379,
+      "step": 1825
+    },
+    {
+      "epoch": 0.6097846051093672,
+      "grad_norm": 0.47740857211098514,
+      "learning_rate": 9.678966075679909e-06,
+      "loss": 0.2294,
+      "step": 1826
+    },
+    {
+      "epoch": 0.6101185506762398,
+      "grad_norm": 0.5105546022851197,
+      "learning_rate": 9.678280601350584e-06,
+      "loss": 0.2342,
+      "step": 1827
+    },
+    {
+      "epoch": 0.6104524962431124,
+      "grad_norm": 0.5084617379607869,
+      "learning_rate": 9.67759442031189e-06,
+      "loss": 0.2333,
+      "step": 1828
+    },
+    {
+      "epoch": 0.610786441809985,
+      "grad_norm": 0.4877401739921855,
+      "learning_rate": 9.676907532667478e-06,
+      "loss": 0.2286,
+      "step": 1829
+    },
+    {
+      "epoch": 0.6111203873768576,
+      "grad_norm": 0.48925434897179987,
+      "learning_rate": 9.676219938521116e-06,
+      "loss": 0.2182,
+      "step": 1830
+    },
+    {
+      "epoch": 0.6114543329437302,
+      "grad_norm": 0.4795913747536622,
+      "learning_rate": 9.675531637976673e-06,
+      "loss": 0.2183,
+      "step": 1831
+    },
+    {
+      "epoch": 0.6117882785106028,
+      "grad_norm": 0.49260466676010023,
+      "learning_rate": 9.674842631138121e-06,
+      "loss": 0.2165,
+      "step": 1832
+    },
+    {
+      "epoch": 0.6121222240774754,
+      "grad_norm": 0.48704610155668626,
+      "learning_rate": 9.674152918109547e-06,
+      "loss": 0.2306,
+      "step": 1833
+    },
+    {
+      "epoch": 0.612456169644348,
+      "grad_norm": 0.5738129836410425,
+      "learning_rate": 9.673462498995138e-06,
+      "loss": 0.225,
+      "step": 1834
+    },
+    {
+      "epoch": 0.6127901152112206,
+      "grad_norm": 0.5935737806472424,
+      "learning_rate": 9.672771373899192e-06,
+      "loss": 0.2263,
+      "step": 1835
+    },
+    {
+      "epoch": 0.6131240607780931,
+      "grad_norm": 0.4805492485102887,
+      "learning_rate": 9.672079542926108e-06,
+      "loss": 0.2326,
+      "step": 1836
+    },
+    {
+      "epoch": 0.6134580063449657,
+      "grad_norm": 0.5717231219637876,
+      "learning_rate": 9.671387006180398e-06,
+      "loss": 0.2354,
+      "step": 1837
+    },
+    {
+      "epoch": 0.6137919519118383,
+      "grad_norm": 0.5224215947069372,
+      "learning_rate": 9.670693763766674e-06,
+      "loss": 0.2151,
+      "step": 1838
+    },
+    {
+      "epoch": 0.6141258974787109,
+      "grad_norm": 0.526357591657413,
+      "learning_rate": 9.669999815789664e-06,
+      "loss": 0.233,
+      "step": 1839
+    },
+    {
+      "epoch": 0.6144598430455835,
+      "grad_norm": 0.453213851588365,
+      "learning_rate": 9.669305162354194e-06,
+      "loss": 0.2039,
+      "step": 1840
+    },
+    {
+      "epoch": 0.6147937886124561,
+      "grad_norm": 0.5265771089532885,
+      "learning_rate": 9.6686098035652e-06,
+      "loss": 0.226,
+      "step": 1841
+    },
+    {
+      "epoch": 0.6151277341793288,
+      "grad_norm": 0.4822531246116318,
+      "learning_rate": 9.667913739527724e-06,
+      "loss": 0.2266,
+      "step": 1842
+    },
+    {
+      "epoch": 0.6154616797462014,
+      "grad_norm": 0.5092579744932829,
+      "learning_rate": 9.667216970346916e-06,
+      "loss": 0.2227,
+      "step": 1843
+    },
+    {
+      "epoch": 0.615795625313074,
+      "grad_norm": 0.5705960019212173,
+      "learning_rate": 9.666519496128027e-06,
+      "loss": 0.2179,
+      "step": 1844
+    },
+    {
+      "epoch": 0.6161295708799466,
+      "grad_norm": 0.5870450982878944,
+      "learning_rate": 9.665821316976423e-06,
+      "loss": 0.2309,
+      "step": 1845
+    },
+    {
+      "epoch": 0.6164635164468192,
+      "grad_norm": 0.48323049309301813,
+      "learning_rate": 9.665122432997571e-06,
+      "loss": 0.2228,
+      "step": 1846
+    },
+    {
+      "epoch": 0.6167974620136918,
+      "grad_norm": 0.43522897761120205,
+      "learning_rate": 9.664422844297045e-06,
+      "loss": 0.2168,
+      "step": 1847
+    },
+    {
+      "epoch": 0.6171314075805644,
+      "grad_norm": 0.4797401271036264,
+      "learning_rate": 9.663722550980528e-06,
+      "loss": 0.2166,
+      "step": 1848
+    },
+    {
+      "epoch": 0.617465353147437,
+      "grad_norm": 0.5404918091345311,
+      "learning_rate": 9.663021553153805e-06,
+      "loss": 0.2155,
+      "step": 1849
+    },
+    {
+      "epoch": 0.6177992987143096,
+      "grad_norm": 0.5019775079244504,
+      "learning_rate": 9.66231985092277e-06,
+      "loss": 0.205,
+      "step": 1850
+    },
+    {
+      "epoch": 0.6181332442811822,
+      "grad_norm": 0.5126171358341591,
+      "learning_rate": 9.661617444393427e-06,
+      "loss": 0.2342,
+      "step": 1851
+    },
+    {
+      "epoch": 0.6184671898480548,
+      "grad_norm": 0.5298697261126594,
+      "learning_rate": 9.660914333671878e-06,
+      "loss": 0.2189,
+      "step": 1852
+    },
+    {
+      "epoch": 0.6188011354149273,
+      "grad_norm": 0.6977397151332118,
+      "learning_rate": 9.66021051886434e-06,
+      "loss": 0.242,
+      "step": 1853
+    },
+    {
+      "epoch": 0.6191350809817999,
+      "grad_norm": 0.4420208100094779,
+      "learning_rate": 9.65950600007713e-06,
+      "loss": 0.1995,
+      "step": 1854
+    },
+    {
+      "epoch": 0.6194690265486725,
+      "grad_norm": 0.49759113853149267,
+      "learning_rate": 9.658800777416676e-06,
+      "loss": 0.2142,
+      "step": 1855
+    },
+    {
+      "epoch": 0.6198029721155451,
+      "grad_norm": 0.4644683178989199,
+      "learning_rate": 9.658094850989508e-06,
+      "loss": 0.206,
+      "step": 1856
+    },
+    {
+      "epoch": 0.6201369176824177,
+      "grad_norm": 0.7389862240060607,
+      "learning_rate": 9.657388220902265e-06,
+      "loss": 0.2231,
+      "step": 1857
+    },
+    {
+      "epoch": 0.6204708632492903,
+      "grad_norm": 0.4606947820212059,
+      "learning_rate": 9.656680887261693e-06,
+      "loss": 0.2115,
+      "step": 1858
+    },
+    {
+      "epoch": 0.620804808816163,
+      "grad_norm": 0.5020543120959755,
+      "learning_rate": 9.655972850174642e-06,
+      "loss": 0.2274,
+      "step": 1859
+    },
+    {
+      "epoch": 0.6211387543830356,
+      "grad_norm": 0.44475121899664427,
+      "learning_rate": 9.65526410974807e-06,
+      "loss": 0.2113,
+      "step": 1860
+    },
+    {
+      "epoch": 0.6214726999499082,
+      "grad_norm": 0.47941219176464955,
+      "learning_rate": 9.65455466608904e-06,
+      "loss": 0.2232,
+      "step": 1861
+    },
+    {
+      "epoch": 0.6218066455167808,
+      "grad_norm": 0.4662479753839037,
+      "learning_rate": 9.653844519304722e-06,
+      "loss": 0.2238,
+      "step": 1862
+    },
+    {
+      "epoch": 0.6221405910836534,
+      "grad_norm": 0.5892373131541172,
+      "learning_rate": 9.653133669502393e-06,
+      "loss": 0.2307,
+      "step": 1863
+    },
+    {
+      "epoch": 0.622474536650526,
+      "grad_norm": 0.46292878818156435,
+      "learning_rate": 9.652422116789432e-06,
+      "loss": 0.221,
+      "step": 1864
+    },
+    {
+      "epoch": 0.6228084822173986,
+      "grad_norm": 0.5683790013976829,
+      "learning_rate": 9.651709861273334e-06,
+      "loss": 0.2267,
+      "step": 1865
+    },
+    {
+      "epoch": 0.6231424277842712,
+      "grad_norm": 0.4718926218027168,
+      "learning_rate": 9.650996903061685e-06,
+      "loss": 0.2138,
+      "step": 1866
+    },
+    {
+      "epoch": 0.6234763733511438,
+      "grad_norm": 0.496502433614263,
+      "learning_rate": 9.650283242262192e-06,
+      "loss": 0.2246,
+      "step": 1867
+    },
+    {
+      "epoch": 0.6238103189180164,
+      "grad_norm": 0.4876162037690022,
+      "learning_rate": 9.64956887898266e-06,
+      "loss": 0.2303,
+      "step": 1868
+    },
+    {
+      "epoch": 0.6241442644848889,
+      "grad_norm": 0.5145615842272623,
+      "learning_rate": 9.648853813331e-06,
+      "loss": 0.2278,
+      "step": 1869
+    },
+    {
+      "epoch": 0.6244782100517615,
+      "grad_norm": 0.527130326304707,
+      "learning_rate": 9.648138045415236e-06,
+      "loss": 0.2301,
+      "step": 1870
+    },
+    {
+      "epoch": 0.6248121556186341,
+      "grad_norm": 0.49041057430030827,
+      "learning_rate": 9.647421575343488e-06,
+      "loss": 0.2271,
+      "step": 1871
+    },
+    {
+      "epoch": 0.6251461011855067,
+      "grad_norm": 0.4861327887722697,
+      "learning_rate": 9.646704403223991e-06,
+      "loss": 0.2153,
+      "step": 1872
+    },
+    {
+      "epoch": 0.6254800467523793,
+      "grad_norm": 0.4859486191134469,
+      "learning_rate": 9.64598652916508e-06,
+      "loss": 0.2159,
+      "step": 1873
+    },
+    {
+      "epoch": 0.6258139923192519,
+      "grad_norm": 0.46779863060566784,
+      "learning_rate": 9.6452679532752e-06,
+      "loss": 0.2145,
+      "step": 1874
+    },
+    {
+      "epoch": 0.6261479378861246,
+      "grad_norm": 0.5215440387248278,
+      "learning_rate": 9.644548675662897e-06,
+      "loss": 0.2309,
+      "step": 1875
+    },
+    {
+      "epoch": 0.6264818834529972,
+      "grad_norm": 0.5113295193583,
+      "learning_rate": 9.64382869643683e-06,
+      "loss": 0.2393,
+      "step": 1876
+    },
+    {
+      "epoch": 0.6268158290198698,
+      "grad_norm": 0.6629216513399953,
+      "learning_rate": 9.64310801570576e-06,
+      "loss": 0.243,
+      "step": 1877
+    },
+    {
+      "epoch": 0.6271497745867424,
+      "grad_norm": 0.5805733643515224,
+      "learning_rate": 9.642386633578553e-06,
+      "loss": 0.236,
+      "step": 1878
+    },
+    {
+      "epoch": 0.627483720153615,
+      "grad_norm": 0.5211638075348191,
+      "learning_rate": 9.641664550164182e-06,
+      "loss": 0.2262,
+      "step": 1879
+    },
+    {
+      "epoch": 0.6278176657204876,
+      "grad_norm": 0.44593296545663935,
+      "learning_rate": 9.640941765571727e-06,
+      "loss": 0.231,
+      "step": 1880
+    },
+    {
+      "epoch": 0.6281516112873602,
+      "grad_norm": 0.47615138615510033,
+      "learning_rate": 9.640218279910374e-06,
+      "loss": 0.2146,
+      "step": 1881
+    },
+    {
+      "epoch": 0.6284855568542328,
+      "grad_norm": 0.5022339248493972,
+      "learning_rate": 9.639494093289412e-06,
+      "loss": 0.2291,
+      "step": 1882
+    },
+    {
+      "epoch": 0.6288195024211054,
+      "grad_norm": 0.46430936600624695,
+      "learning_rate": 9.638769205818239e-06,
+      "loss": 0.2071,
+      "step": 1883
+    },
+    {
+      "epoch": 0.629153447987978,
+      "grad_norm": 0.5226803610865401,
+      "learning_rate": 9.638043617606358e-06,
+      "loss": 0.2414,
+      "step": 1884
+    },
+    {
+      "epoch": 0.6294873935548505,
+      "grad_norm": 0.48540480199653707,
+      "learning_rate": 9.637317328763378e-06,
+      "loss": 0.2233,
+      "step": 1885
+    },
+    {
+      "epoch": 0.6298213391217231,
+      "grad_norm": 0.44530932798385103,
+      "learning_rate": 9.636590339399012e-06,
+      "loss": 0.2044,
+      "step": 1886
+    },
+    {
+      "epoch": 0.6301552846885957,
+      "grad_norm": 0.46347505096860814,
+      "learning_rate": 9.63586264962308e-06,
+      "loss": 0.226,
+      "step": 1887
+    },
+    {
+      "epoch": 0.6304892302554683,
+      "grad_norm": 0.5058077241719129,
+      "learning_rate": 9.635134259545511e-06,
+      "loss": 0.2334,
+      "step": 1888
+    },
+    {
+      "epoch": 0.6308231758223409,
+      "grad_norm": 0.5197637248708888,
+      "learning_rate": 9.634405169276335e-06,
+      "loss": 0.2325,
+      "step": 1889
+    },
+    {
+      "epoch": 0.6311571213892135,
+      "grad_norm": 0.5179888030499751,
+      "learning_rate": 9.63367537892569e-06,
+      "loss": 0.2226,
+      "step": 1890
+    },
+    {
+      "epoch": 0.6314910669560861,
+      "grad_norm": 0.6237768819180914,
+      "learning_rate": 9.63294488860382e-06,
+      "loss": 0.2429,
+      "step": 1891
+    },
+    {
+      "epoch": 0.6318250125229588,
+      "grad_norm": 0.4783353974434491,
+      "learning_rate": 9.63221369842107e-06,
+      "loss": 0.2161,
+      "step": 1892
+    },
+    {
+      "epoch": 0.6321589580898314,
+      "grad_norm": 0.4625966356502894,
+      "learning_rate": 9.631481808487902e-06,
+      "loss": 0.219,
+      "step": 1893
+    },
+    {
+      "epoch": 0.632492903656704,
+      "grad_norm": 0.5729936471207139,
+      "learning_rate": 9.63074921891487e-06,
+      "loss": 0.2275,
+      "step": 1894
+    },
+    {
+      "epoch": 0.6328268492235766,
+      "grad_norm": 0.4584520008958071,
+      "learning_rate": 9.630015929812646e-06,
+      "loss": 0.2271,
+      "step": 1895
+    },
+    {
+      "epoch": 0.6331607947904492,
+      "grad_norm": 0.5542075963679276,
+      "learning_rate": 9.629281941291998e-06,
+      "loss": 0.2412,
+      "step": 1896
+    },
+    {
+      "epoch": 0.6334947403573218,
+      "grad_norm": 0.47050880416056196,
+      "learning_rate": 9.628547253463804e-06,
+      "loss": 0.2254,
+      "step": 1897
+    },
+    {
+      "epoch": 0.6338286859241944,
+      "grad_norm": 0.5287224363148264,
+      "learning_rate": 9.627811866439048e-06,
+      "loss": 0.226,
+      "step": 1898
+    },
+    {
+      "epoch": 0.634162631491067,
+      "grad_norm": 0.5349630430590014,
+      "learning_rate": 9.627075780328818e-06,
+      "loss": 0.2457,
+      "step": 1899
+    },
+    {
+      "epoch": 0.6344965770579396,
+      "grad_norm": 0.48848827404295203,
+      "learning_rate": 9.626338995244313e-06,
+      "loss": 0.2239,
+      "step": 1900
+    },
+    {
+      "epoch": 0.6348305226248122,
+      "grad_norm": 0.5392885299832741,
+      "learning_rate": 9.625601511296826e-06,
+      "loss": 0.2292,
+      "step": 1901
+    },
+    {
+      "epoch": 0.6351644681916847,
+      "grad_norm": 0.51626967141371,
+      "learning_rate": 9.624863328597767e-06,
+      "loss": 0.2411,
+      "step": 1902
+    },
+    {
+      "epoch": 0.6354984137585573,
+      "grad_norm": 0.48472493185557913,
+      "learning_rate": 9.624124447258647e-06,
+      "loss": 0.2177,
+      "step": 1903
+    },
+    {
+      "epoch": 0.6358323593254299,
+      "grad_norm": 0.6068590044334997,
+      "learning_rate": 9.62338486739108e-06,
+      "loss": 0.2214,
+      "step": 1904
+    },
+    {
+      "epoch": 0.6361663048923025,
+      "grad_norm": 0.5569455920656661,
+      "learning_rate": 9.62264458910679e-06,
+      "loss": 0.2227,
+      "step": 1905
+    },
+    {
+      "epoch": 0.6365002504591751,
+      "grad_norm": 0.4943658711020671,
+      "learning_rate": 9.621903612517608e-06,
+      "loss": 0.22,
+      "step": 1906
+    },
+    {
+      "epoch": 0.6368341960260477,
+      "grad_norm": 0.6738439864635068,
+      "learning_rate": 9.621161937735463e-06,
+      "loss": 0.2215,
+      "step": 1907
+    },
+    {
+      "epoch": 0.6371681415929203,
+      "grad_norm": 0.5125207288216446,
+      "learning_rate": 9.620419564872394e-06,
+      "loss": 0.2311,
+      "step": 1908
+    },
+    {
+      "epoch": 0.637502087159793,
+      "grad_norm": 0.5477693989903016,
+      "learning_rate": 9.619676494040547e-06,
+      "loss": 0.2447,
+      "step": 1909
+    },
+    {
+      "epoch": 0.6378360327266656,
+      "grad_norm": 0.618827790959411,
+      "learning_rate": 9.61893272535217e-06,
+      "loss": 0.2269,
+      "step": 1910
+    },
+    {
+      "epoch": 0.6381699782935382,
+      "grad_norm": 0.5831020385440705,
+      "learning_rate": 9.618188258919618e-06,
+      "loss": 0.2504,
+      "step": 1911
+    },
+    {
+      "epoch": 0.6385039238604108,
+      "grad_norm": 0.4575021594118544,
+      "learning_rate": 9.617443094855354e-06,
+      "loss": 0.2212,
+      "step": 1912
+    },
+    {
+      "epoch": 0.6388378694272834,
+      "grad_norm": 0.474232930993678,
+      "learning_rate": 9.61669723327194e-06,
+      "loss": 0.2352,
+      "step": 1913
+    },
+    {
+      "epoch": 0.639171814994156,
+      "grad_norm": 0.5374800460315998,
+      "learning_rate": 9.615950674282049e-06,
+      "loss": 0.2396,
+      "step": 1914
+    },
+    {
+      "epoch": 0.6395057605610286,
+      "grad_norm": 0.5287587226953798,
+      "learning_rate": 9.61520341799846e-06,
+      "loss": 0.2302,
+      "step": 1915
+    },
+    {
+      "epoch": 0.6398397061279012,
+      "grad_norm": 0.5237854216155173,
+      "learning_rate": 9.614455464534049e-06,
+      "loss": 0.2453,
+      "step": 1916
+    },
+    {
+      "epoch": 0.6401736516947738,
+      "grad_norm": 0.44508260080269146,
+      "learning_rate": 9.613706814001809e-06,
+      "loss": 0.2066,
+      "step": 1917
+    },
+    {
+      "epoch": 0.6405075972616463,
+      "grad_norm": 0.5211596090749453,
+      "learning_rate": 9.612957466514829e-06,
+      "loss": 0.2279,
+      "step": 1918
+    },
+    {
+      "epoch": 0.6408415428285189,
+      "grad_norm": 0.4862656406723413,
+      "learning_rate": 9.61220742218631e-06,
+      "loss": 0.2304,
+      "step": 1919
+    },
+    {
+      "epoch": 0.6411754883953915,
+      "grad_norm": 0.4472518622732128,
+      "learning_rate": 9.61145668112955e-06,
+      "loss": 0.2296,
+      "step": 1920
+    },
+    {
+      "epoch": 0.6415094339622641,
+      "grad_norm": 0.5348057312329286,
+      "learning_rate": 9.610705243457962e-06,
+      "loss": 0.2421,
+      "step": 1921
+    },
+    {
+      "epoch": 0.6418433795291367,
+      "grad_norm": 0.539916851096655,
+      "learning_rate": 9.609953109285057e-06,
+      "loss": 0.2225,
+      "step": 1922
+    },
+    {
+      "epoch": 0.6421773250960093,
+      "grad_norm": 0.486698830537002,
+      "learning_rate": 9.609200278724456e-06,
+      "loss": 0.2303,
+      "step": 1923
+    },
+    {
+      "epoch": 0.6425112706628819,
+      "grad_norm": 0.4853767234392697,
+      "learning_rate": 9.60844675188988e-06,
+      "loss": 0.2358,
+      "step": 1924
+    },
+    {
+      "epoch": 0.6428452162297545,
+      "grad_norm": 0.5783948282032476,
+      "learning_rate": 9.60769252889516e-06,
+      "loss": 0.2549,
+      "step": 1925
+    },
+    {
+      "epoch": 0.6431791617966272,
+      "grad_norm": 0.4677106105333194,
+      "learning_rate": 9.606937609854227e-06,
+      "loss": 0.2311,
+      "step": 1926
+    },
+    {
+      "epoch": 0.6435131073634998,
+      "grad_norm": 0.7242966644398318,
+      "learning_rate": 9.606181994881124e-06,
+      "loss": 0.2273,
+      "step": 1927
+    },
+    {
+      "epoch": 0.6438470529303724,
+      "grad_norm": 0.5186721278224474,
+      "learning_rate": 9.605425684089998e-06,
+      "loss": 0.246,
+      "step": 1928
+    },
+    {
+      "epoch": 0.644180998497245,
+      "grad_norm": 0.6545504763262779,
+      "learning_rate": 9.604668677595093e-06,
+      "loss": 0.2339,
+      "step": 1929
+    },
+    {
+      "epoch": 0.6445149440641176,
+      "grad_norm": 0.4496430333660451,
+      "learning_rate": 9.603910975510764e-06,
+      "loss": 0.2205,
+      "step": 1930
+    },
+    {
+      "epoch": 0.6448488896309902,
+      "grad_norm": 0.5824478060538069,
+      "learning_rate": 9.603152577951476e-06,
+      "loss": 0.2292,
+      "step": 1931
+    },
+    {
+      "epoch": 0.6451828351978628,
+      "grad_norm": 0.4683634543655863,
+      "learning_rate": 9.60239348503179e-06,
+      "loss": 0.2103,
+      "step": 1932
+    },
+    {
+      "epoch": 0.6455167807647354,
+      "grad_norm": 0.4654095627434126,
+      "learning_rate": 9.601633696866376e-06,
+      "loss": 0.2247,
+      "step": 1933
+    },
+    {
+      "epoch": 0.6458507263316079,
+      "grad_norm": 0.503168780861353,
+      "learning_rate": 9.60087321357001e-06,
+      "loss": 0.2255,
+      "step": 1934
+    },
+    {
+      "epoch": 0.6461846718984805,
+      "grad_norm": 0.5396694281989363,
+      "learning_rate": 9.600112035257571e-06,
+      "loss": 0.2222,
+      "step": 1935
+    },
+    {
+      "epoch": 0.6465186174653531,
+      "grad_norm": 0.4407844703562947,
+      "learning_rate": 9.599350162044045e-06,
+      "loss": 0.2215,
+      "step": 1936
+    },
+    {
+      "epoch": 0.6468525630322257,
+      "grad_norm": 0.4504311612267907,
+      "learning_rate": 9.598587594044522e-06,
+      "loss": 0.2241,
+      "step": 1937
+    },
+    {
+      "epoch": 0.6471865085990983,
+      "grad_norm": 0.5078973609706072,
+      "learning_rate": 9.597824331374196e-06,
+      "loss": 0.2147,
+      "step": 1938
+    },
+    {
+      "epoch": 0.6475204541659709,
+      "grad_norm": 0.5105786340988594,
+      "learning_rate": 9.597060374148365e-06,
+      "loss": 0.222,
+      "step": 1939
+    },
+    {
+      "epoch": 0.6478543997328435,
+      "grad_norm": 0.4459757628273001,
+      "learning_rate": 9.596295722482439e-06,
+      "loss": 0.2175,
+      "step": 1940
+    },
+    {
+      "epoch": 0.6481883452997161,
+      "grad_norm": 0.5146618161191947,
+      "learning_rate": 9.595530376491924e-06,
+      "loss": 0.2367,
+      "step": 1941
+    },
+    {
+      "epoch": 0.6485222908665887,
+      "grad_norm": 0.4924035886938882,
+      "learning_rate": 9.594764336292432e-06,
+      "loss": 0.227,
+      "step": 1942
+    },
+    {
+      "epoch": 0.6488562364334614,
+      "grad_norm": 0.4836297523018866,
+      "learning_rate": 9.593997601999689e-06,
+      "loss": 0.2184,
+      "step": 1943
+    },
+    {
+      "epoch": 0.649190182000334,
+      "grad_norm": 0.4954472502507133,
+      "learning_rate": 9.593230173729514e-06,
+      "loss": 0.2289,
+      "step": 1944
+    },
+    {
+      "epoch": 0.6495241275672066,
+      "grad_norm": 0.46892208039162586,
+      "learning_rate": 9.592462051597838e-06,
+      "loss": 0.2333,
+      "step": 1945
+    },
+    {
+      "epoch": 0.6498580731340792,
+      "grad_norm": 0.4707733396458665,
+      "learning_rate": 9.591693235720695e-06,
+      "loss": 0.2179,
+      "step": 1946
+    },
+    {
+      "epoch": 0.6501920187009518,
+      "grad_norm": 0.5021563108304271,
+      "learning_rate": 9.590923726214224e-06,
+      "loss": 0.2312,
+      "step": 1947
+    },
+    {
+      "epoch": 0.6505259642678244,
+      "grad_norm": 0.5216409946552976,
+      "learning_rate": 9.590153523194665e-06,
+      "loss": 0.236,
+      "step": 1948
+    },
+    {
+      "epoch": 0.650859909834697,
+      "grad_norm": 0.445367026551766,
+      "learning_rate": 9.589382626778371e-06,
+      "loss": 0.2039,
+      "step": 1949
+    },
+    {
+      "epoch": 0.6511938554015696,
+      "grad_norm": 0.4719313967503157,
+      "learning_rate": 9.588611037081793e-06,
+      "loss": 0.2291,
+      "step": 1950
+    },
+    {
+      "epoch": 0.6515278009684421,
+      "grad_norm": 0.5041618490835817,
+      "learning_rate": 9.587838754221488e-06,
+      "loss": 0.2186,
+      "step": 1951
+    },
+    {
+      "epoch": 0.6518617465353147,
+      "grad_norm": 0.5656846698351666,
+      "learning_rate": 9.587065778314119e-06,
+      "loss": 0.2292,
+      "step": 1952
+    },
+    {
+      "epoch": 0.6521956921021873,
+      "grad_norm": 0.48863066421780066,
+      "learning_rate": 9.586292109476454e-06,
+      "loss": 0.2107,
+      "step": 1953
+    },
+    {
+      "epoch": 0.6525296376690599,
+      "grad_norm": 0.7874907012437945,
+      "learning_rate": 9.585517747825363e-06,
+      "loss": 0.2378,
+      "step": 1954
+    },
+    {
+      "epoch": 0.6528635832359325,
+      "grad_norm": 0.5095487818003742,
+      "learning_rate": 9.584742693477825e-06,
+      "loss": 0.2279,
+      "step": 1955
+    },
+    {
+      "epoch": 0.6531975288028051,
+      "grad_norm": 0.5643641986886433,
+      "learning_rate": 9.58396694655092e-06,
+      "loss": 0.2581,
+      "step": 1956
+    },
+    {
+      "epoch": 0.6535314743696777,
+      "grad_norm": 0.4898371660965593,
+      "learning_rate": 9.583190507161832e-06,
+      "loss": 0.2203,
+      "step": 1957
+    },
+    {
+      "epoch": 0.6538654199365503,
+      "grad_norm": 0.5921772486152574,
+      "learning_rate": 9.582413375427852e-06,
+      "loss": 0.2268,
+      "step": 1958
+    },
+    {
+      "epoch": 0.654199365503423,
+      "grad_norm": 0.45116730167892133,
+      "learning_rate": 9.581635551466376e-06,
+      "loss": 0.2195,
+      "step": 1959
+    },
+    {
+      "epoch": 0.6545333110702956,
+      "grad_norm": 0.47843623987223366,
+      "learning_rate": 9.580857035394904e-06,
+      "loss": 0.225,
+      "step": 1960
+    },
+    {
+      "epoch": 0.6548672566371682,
+      "grad_norm": 0.5197754147716475,
+      "learning_rate": 9.580077827331038e-06,
+      "loss": 0.2368,
+      "step": 1961
+    },
+    {
+      "epoch": 0.6552012022040408,
+      "grad_norm": 0.4170784963842771,
+      "learning_rate": 9.579297927392488e-06,
+      "loss": 0.2065,
+      "step": 1962
+    },
+    {
+      "epoch": 0.6555351477709134,
+      "grad_norm": 0.5337746138339877,
+      "learning_rate": 9.578517335697065e-06,
+      "loss": 0.2313,
+      "step": 1963
+    },
+    {
+      "epoch": 0.655869093337786,
+      "grad_norm": 0.49279896240854576,
+      "learning_rate": 9.577736052362689e-06,
+      "loss": 0.2268,
+      "step": 1964
+    },
+    {
+      "epoch": 0.6562030389046586,
+      "grad_norm": 0.4302932121647986,
+      "learning_rate": 9.576954077507381e-06,
+      "loss": 0.2049,
+      "step": 1965
+    },
+    {
+      "epoch": 0.6565369844715312,
+      "grad_norm": 0.5127555964144664,
+      "learning_rate": 9.576171411249269e-06,
+      "loss": 0.23,
+      "step": 1966
+    },
+    {
+      "epoch": 0.6568709300384037,
+      "grad_norm": 0.5080459538592854,
+      "learning_rate": 9.575388053706582e-06,
+      "loss": 0.2317,
+      "step": 1967
+    },
+    {
+      "epoch": 0.6572048756052763,
+      "grad_norm": 0.4745192813209085,
+      "learning_rate": 9.574604004997654e-06,
+      "loss": 0.2154,
+      "step": 1968
+    },
+    {
+      "epoch": 0.6575388211721489,
+      "grad_norm": 0.5389965358570025,
+      "learning_rate": 9.57381926524093e-06,
+      "loss": 0.2433,
+      "step": 1969
+    },
+    {
+      "epoch": 0.6578727667390215,
+      "grad_norm": 0.4956040255633844,
+      "learning_rate": 9.57303383455495e-06,
+      "loss": 0.2234,
+      "step": 1970
+    },
+    {
+      "epoch": 0.6582067123058941,
+      "grad_norm": 0.48821226341668883,
+      "learning_rate": 9.572247713058362e-06,
+      "loss": 0.2225,
+      "step": 1971
+    },
+    {
+      "epoch": 0.6585406578727667,
+      "grad_norm": 0.48960509199150504,
+      "learning_rate": 9.571460900869923e-06,
+      "loss": 0.2243,
+      "step": 1972
+    },
+    {
+      "epoch": 0.6588746034396393,
+      "grad_norm": 0.5388856840457162,
+      "learning_rate": 9.570673398108485e-06,
+      "loss": 0.2191,
+      "step": 1973
+    },
+    {
+      "epoch": 0.6592085490065119,
+      "grad_norm": 0.5247531776272966,
+      "learning_rate": 9.569885204893015e-06,
+      "loss": 0.2438,
+      "step": 1974
+    },
+    {
+      "epoch": 0.6595424945733845,
+      "grad_norm": 0.4636309459992003,
+      "learning_rate": 9.569096321342574e-06,
+      "loss": 0.2205,
+      "step": 1975
+    },
+    {
+      "epoch": 0.6598764401402571,
+      "grad_norm": 0.5647518684704627,
+      "learning_rate": 9.568306747576335e-06,
+      "loss": 0.2424,
+      "step": 1976
+    },
+    {
+      "epoch": 0.6602103857071298,
+      "grad_norm": 0.4632908967911912,
+      "learning_rate": 9.567516483713572e-06,
+      "loss": 0.2183,
+      "step": 1977
+    },
+    {
+      "epoch": 0.6605443312740024,
+      "grad_norm": 0.4806235309010455,
+      "learning_rate": 9.566725529873664e-06,
+      "loss": 0.2129,
+      "step": 1978
+    },
+    {
+      "epoch": 0.660878276840875,
+      "grad_norm": 0.47812898299179163,
+      "learning_rate": 9.565933886176093e-06,
+      "loss": 0.2226,
+      "step": 1979
+    },
+    {
+      "epoch": 0.6612122224077476,
+      "grad_norm": 0.4790050168212765,
+      "learning_rate": 9.565141552740445e-06,
+      "loss": 0.2127,
+      "step": 1980
+    },
+    {
+      "epoch": 0.6615461679746202,
+      "grad_norm": 0.47301797232701465,
+      "learning_rate": 9.564348529686413e-06,
+      "loss": 0.2276,
+      "step": 1981
+    },
+    {
+      "epoch": 0.6618801135414928,
+      "grad_norm": 0.47800269002423895,
+      "learning_rate": 9.563554817133794e-06,
+      "loss": 0.2276,
+      "step": 1982
+    },
+    {
+      "epoch": 0.6622140591083653,
+      "grad_norm": 0.47697326231641335,
+      "learning_rate": 9.562760415202483e-06,
+      "loss": 0.2075,
+      "step": 1983
+    },
+    {
+      "epoch": 0.6625480046752379,
+      "grad_norm": 0.45114047498178445,
+      "learning_rate": 9.56196532401249e-06,
+      "loss": 0.1993,
+      "step": 1984
+    },
+    {
+      "epoch": 0.6628819502421105,
+      "grad_norm": 0.6767131550163791,
+      "learning_rate": 9.561169543683917e-06,
+      "loss": 0.2323,
+      "step": 1985
+    },
+    {
+      "epoch": 0.6632158958089831,
+      "grad_norm": 0.46779536182015385,
+      "learning_rate": 9.560373074336977e-06,
+      "loss": 0.2219,
+      "step": 1986
+    },
+    {
+      "epoch": 0.6635498413758557,
+      "grad_norm": 0.5689668029517431,
+      "learning_rate": 9.55957591609199e-06,
+      "loss": 0.2294,
+      "step": 1987
+    },
+    {
+      "epoch": 0.6638837869427283,
+      "grad_norm": 0.5473530266696686,
+      "learning_rate": 9.558778069069373e-06,
+      "loss": 0.2187,
+      "step": 1988
+    },
+    {
+      "epoch": 0.6642177325096009,
+      "grad_norm": 0.49049488313910705,
+      "learning_rate": 9.55797953338965e-06,
+      "loss": 0.2159,
+      "step": 1989
+    },
+    {
+      "epoch": 0.6645516780764735,
+      "grad_norm": 0.6069581055887437,
+      "learning_rate": 9.55718030917345e-06,
+      "loss": 0.2333,
+      "step": 1990
+    },
+    {
+      "epoch": 0.6648856236433461,
+      "grad_norm": 0.5169479730953396,
+      "learning_rate": 9.556380396541507e-06,
+      "loss": 0.2312,
+      "step": 1991
+    },
+    {
+      "epoch": 0.6652195692102187,
+      "grad_norm": 0.5700537116713167,
+      "learning_rate": 9.555579795614654e-06,
+      "loss": 0.247,
+      "step": 1992
+    },
+    {
+      "epoch": 0.6655535147770913,
+      "grad_norm": 0.6320463038320528,
+      "learning_rate": 9.554778506513834e-06,
+      "loss": 0.2578,
+      "step": 1993
+    },
+    {
+      "epoch": 0.665887460343964,
+      "grad_norm": 0.6242767829904925,
+      "learning_rate": 9.553976529360087e-06,
+      "loss": 0.2453,
+      "step": 1994
+    },
+    {
+      "epoch": 0.6662214059108366,
+      "grad_norm": 0.4657124319169294,
+      "learning_rate": 9.553173864274567e-06,
+      "loss": 0.2049,
+      "step": 1995
+    },
+    {
+      "epoch": 0.6665553514777092,
+      "grad_norm": 0.6911587103761502,
+      "learning_rate": 9.552370511378522e-06,
+      "loss": 0.2255,
+      "step": 1996
+    },
+    {
+      "epoch": 0.6668892970445818,
+      "grad_norm": 0.5680114058138519,
+      "learning_rate": 9.551566470793308e-06,
+      "loss": 0.2209,
+      "step": 1997
+    },
+    {
+      "epoch": 0.6672232426114544,
+      "grad_norm": 0.4878976742541729,
+      "learning_rate": 9.550761742640387e-06,
+      "loss": 0.2254,
+      "step": 1998
+    },
+    {
+      "epoch": 0.667557188178327,
+      "grad_norm": 0.5997891836824637,
+      "learning_rate": 9.549956327041318e-06,
+      "loss": 0.2323,
+      "step": 1999
+    },
+    {
+      "epoch": 0.6678911337451995,
+      "grad_norm": 0.4561949621578282,
+      "learning_rate": 9.549150224117776e-06,
+      "loss": 0.2078,
+      "step": 2000
+    },
+    {
+      "epoch": 0.6682250793120721,
+      "grad_norm": 0.4842209833350863,
+      "learning_rate": 9.548343433991524e-06,
+      "loss": 0.2096,
+      "step": 2001
+    },
+    {
+      "epoch": 0.6685590248789447,
+      "grad_norm": 0.5321439301519651,
+      "learning_rate": 9.547535956784445e-06,
+      "loss": 0.2498,
+      "step": 2002
+    },
+    {
+      "epoch": 0.6688929704458173,
+      "grad_norm": 0.5345226511901285,
+      "learning_rate": 9.546727792618512e-06,
+      "loss": 0.2149,
+      "step": 2003
+    },
+    {
+      "epoch": 0.6692269160126899,
+      "grad_norm": 0.5948754984549891,
+      "learning_rate": 9.545918941615811e-06,
+      "loss": 0.2268,
+      "step": 2004
+    },
+    {
+      "epoch": 0.6695608615795625,
+      "grad_norm": 0.5249741131297062,
+      "learning_rate": 9.545109403898527e-06,
+      "loss": 0.2332,
+      "step": 2005
+    },
+    {
+      "epoch": 0.6698948071464351,
+      "grad_norm": 0.448439782885968,
+      "learning_rate": 9.544299179588952e-06,
+      "loss": 0.2197,
+      "step": 2006
+    },
+    {
+      "epoch": 0.6702287527133077,
+      "grad_norm": 0.5320016792358742,
+      "learning_rate": 9.543488268809478e-06,
+      "loss": 0.2217,
+      "step": 2007
+    },
+    {
+      "epoch": 0.6705626982801803,
+      "grad_norm": 0.5033575128331692,
+      "learning_rate": 9.542676671682601e-06,
+      "loss": 0.2294,
+      "step": 2008
+    },
+    {
+      "epoch": 0.6708966438470529,
+      "grad_norm": 0.5338603551058775,
+      "learning_rate": 9.541864388330926e-06,
+      "loss": 0.2359,
+      "step": 2009
+    },
+    {
+      "epoch": 0.6712305894139255,
+      "grad_norm": 0.4960343876895645,
+      "learning_rate": 9.541051418877156e-06,
+      "loss": 0.23,
+      "step": 2010
+    },
+    {
+      "epoch": 0.6715645349807982,
+      "grad_norm": 0.5019233736697615,
+      "learning_rate": 9.5402377634441e-06,
+      "loss": 0.2184,
+      "step": 2011
+    },
+    {
+      "epoch": 0.6718984805476708,
+      "grad_norm": 0.5116242184595458,
+      "learning_rate": 9.539423422154672e-06,
+      "loss": 0.2057,
+      "step": 2012
+    },
+    {
+      "epoch": 0.6722324261145434,
+      "grad_norm": 0.4581151499736462,
+      "learning_rate": 9.538608395131884e-06,
+      "loss": 0.219,
+      "step": 2013
+    },
+    {
+      "epoch": 0.672566371681416,
+      "grad_norm": 0.5263573724429383,
+      "learning_rate": 9.537792682498859e-06,
+      "loss": 0.2345,
+      "step": 2014
+    },
+    {
+      "epoch": 0.6729003172482886,
+      "grad_norm": 0.5486956203361241,
+      "learning_rate": 9.536976284378818e-06,
+      "loss": 0.2365,
+      "step": 2015
+    },
+    {
+      "epoch": 0.6732342628151611,
+      "grad_norm": 0.48552662228072885,
+      "learning_rate": 9.536159200895088e-06,
+      "loss": 0.2472,
+      "step": 2016
+    },
+    {
+      "epoch": 0.6735682083820337,
+      "grad_norm": 0.4807730164835203,
+      "learning_rate": 9.535341432171098e-06,
+      "loss": 0.2158,
+      "step": 2017
+    },
+    {
+      "epoch": 0.6739021539489063,
+      "grad_norm": 0.5082241482128,
+      "learning_rate": 9.534522978330384e-06,
+      "loss": 0.2273,
+      "step": 2018
+    },
+    {
+      "epoch": 0.6742360995157789,
+      "grad_norm": 0.5247052847237837,
+      "learning_rate": 9.533703839496581e-06,
+      "loss": 0.2122,
+      "step": 2019
+    },
+    {
+      "epoch": 0.6745700450826515,
+      "grad_norm": 0.4996722623100839,
+      "learning_rate": 9.532884015793432e-06,
+      "loss": 0.2144,
+      "step": 2020
+    },
+    {
+      "epoch": 0.6749039906495241,
+      "grad_norm": 0.4864919237009365,
+      "learning_rate": 9.532063507344777e-06,
+      "loss": 0.2246,
+      "step": 2021
+    },
+    {
+      "epoch": 0.6752379362163967,
+      "grad_norm": 0.5055062622420079,
+      "learning_rate": 9.53124231427457e-06,
+      "loss": 0.2329,
+      "step": 2022
+    },
+    {
+      "epoch": 0.6755718817832693,
+      "grad_norm": 0.5127546113587057,
+      "learning_rate": 9.530420436706853e-06,
+      "loss": 0.225,
+      "step": 2023
+    },
+    {
+      "epoch": 0.6759058273501419,
+      "grad_norm": 0.5199847442243274,
+      "learning_rate": 9.529597874765788e-06,
+      "loss": 0.2224,
+      "step": 2024
+    },
+    {
+      "epoch": 0.6762397729170145,
+      "grad_norm": 0.5520828526591032,
+      "learning_rate": 9.528774628575628e-06,
+      "loss": 0.2329,
+      "step": 2025
+    },
+    {
+      "epoch": 0.6765737184838871,
+      "grad_norm": 0.4690377580828556,
+      "learning_rate": 9.527950698260737e-06,
+      "loss": 0.2162,
+      "step": 2026
+    },
+    {
+      "epoch": 0.6769076640507597,
+      "grad_norm": 0.7257663653006456,
+      "learning_rate": 9.527126083945578e-06,
+      "loss": 0.2423,
+      "step": 2027
+    },
+    {
+      "epoch": 0.6772416096176324,
+      "grad_norm": 0.592781750687045,
+      "learning_rate": 9.526300785754719e-06,
+      "loss": 0.2328,
+      "step": 2028
+    },
+    {
+      "epoch": 0.677575555184505,
+      "grad_norm": 0.6355902972521256,
+      "learning_rate": 9.525474803812831e-06,
+      "loss": 0.2281,
+      "step": 2029
+    },
+    {
+      "epoch": 0.6779095007513776,
+      "grad_norm": 0.5201719503575588,
+      "learning_rate": 9.524648138244688e-06,
+      "loss": 0.2376,
+      "step": 2030
+    },
+    {
+      "epoch": 0.6782434463182502,
+      "grad_norm": 0.5216385214418539,
+      "learning_rate": 9.523820789175167e-06,
+      "loss": 0.2329,
+      "step": 2031
+    },
+    {
+      "epoch": 0.6785773918851227,
+      "grad_norm": 0.4860538015687635,
+      "learning_rate": 9.52299275672925e-06,
+      "loss": 0.2239,
+      "step": 2032
+    },
+    {
+      "epoch": 0.6789113374519953,
+      "grad_norm": 0.5065750638589931,
+      "learning_rate": 9.52216404103202e-06,
+      "loss": 0.2204,
+      "step": 2033
+    },
+    {
+      "epoch": 0.6792452830188679,
+      "grad_norm": 0.5465674778090972,
+      "learning_rate": 9.521334642208666e-06,
+      "loss": 0.2149,
+      "step": 2034
+    },
+    {
+      "epoch": 0.6795792285857405,
+      "grad_norm": 0.530148424285913,
+      "learning_rate": 9.520504560384476e-06,
+      "loss": 0.2291,
+      "step": 2035
+    },
+    {
+      "epoch": 0.6799131741526131,
+      "grad_norm": 0.439419070166103,
+      "learning_rate": 9.519673795684845e-06,
+      "loss": 0.2246,
+      "step": 2036
+    },
+    {
+      "epoch": 0.6802471197194857,
+      "grad_norm": 0.5110159600833178,
+      "learning_rate": 9.518842348235271e-06,
+      "loss": 0.2156,
+      "step": 2037
+    },
+    {
+      "epoch": 0.6805810652863583,
+      "grad_norm": 0.5080312664365709,
+      "learning_rate": 9.51801021816135e-06,
+      "loss": 0.2282,
+      "step": 2038
+    },
+    {
+      "epoch": 0.6809150108532309,
+      "grad_norm": 0.4405972025652121,
+      "learning_rate": 9.51717740558879e-06,
+      "loss": 0.2287,
+      "step": 2039
+    },
+    {
+      "epoch": 0.6812489564201035,
+      "grad_norm": 0.4846966375847971,
+      "learning_rate": 9.516343910643395e-06,
+      "loss": 0.2304,
+      "step": 2040
+    },
+    {
+      "epoch": 0.6815829019869761,
+      "grad_norm": 0.4458254730841546,
+      "learning_rate": 9.515509733451074e-06,
+      "loss": 0.2266,
+      "step": 2041
+    },
+    {
+      "epoch": 0.6819168475538487,
+      "grad_norm": 0.5081979817054557,
+      "learning_rate": 9.514674874137838e-06,
+      "loss": 0.2212,
+      "step": 2042
+    },
+    {
+      "epoch": 0.6822507931207213,
+      "grad_norm": 0.44506787175264934,
+      "learning_rate": 9.513839332829806e-06,
+      "loss": 0.217,
+      "step": 2043
+    },
+    {
+      "epoch": 0.682584738687594,
+      "grad_norm": 0.4711214808235664,
+      "learning_rate": 9.513003109653192e-06,
+      "loss": 0.214,
+      "step": 2044
+    },
+    {
+      "epoch": 0.6829186842544666,
+      "grad_norm": 0.4690699744568473,
+      "learning_rate": 9.512166204734322e-06,
+      "loss": 0.2265,
+      "step": 2045
+    },
+    {
+      "epoch": 0.6832526298213392,
+      "grad_norm": 0.47218778632793423,
+      "learning_rate": 9.511328618199614e-06,
+      "loss": 0.222,
+      "step": 2046
+    },
+    {
+      "epoch": 0.6835865753882118,
+      "grad_norm": 0.5360979263013558,
+      "learning_rate": 9.510490350175602e-06,
+      "loss": 0.2571,
+      "step": 2047
+    },
+    {
+      "epoch": 0.6839205209550844,
+      "grad_norm": 0.4708991844847955,
+      "learning_rate": 9.50965140078891e-06,
+      "loss": 0.2347,
+      "step": 2048
+    },
+    {
+      "epoch": 0.6842544665219569,
+      "grad_norm": 0.4819450196925043,
+      "learning_rate": 9.508811770166277e-06,
+      "loss": 0.2322,
+      "step": 2049
+    },
+    {
+      "epoch": 0.6845884120888295,
+      "grad_norm": 0.45770077260472547,
+      "learning_rate": 9.507971458434538e-06,
+      "loss": 0.2217,
+      "step": 2050
+    },
+    {
+      "epoch": 0.6849223576557021,
+      "grad_norm": 0.4682650978661971,
+      "learning_rate": 9.507130465720628e-06,
+      "loss": 0.2258,
+      "step": 2051
+    },
+    {
+      "epoch": 0.6852563032225747,
+      "grad_norm": 0.4225589686793817,
+      "learning_rate": 9.506288792151592e-06,
+      "loss": 0.2082,
+      "step": 2052
+    },
+    {
+      "epoch": 0.6855902487894473,
+      "grad_norm": 0.43243099858850814,
+      "learning_rate": 9.505446437854574e-06,
+      "loss": 0.2192,
+      "step": 2053
+    },
+    {
+      "epoch": 0.6859241943563199,
+      "grad_norm": 0.48545888804912246,
+      "learning_rate": 9.504603402956823e-06,
+      "loss": 0.232,
+      "step": 2054
+    },
+    {
+      "epoch": 0.6862581399231925,
+      "grad_norm": 0.4784145024105088,
+      "learning_rate": 9.503759687585686e-06,
+      "loss": 0.2257,
+      "step": 2055
+    },
+    {
+      "epoch": 0.6865920854900651,
+      "grad_norm": 0.47809150718111393,
+      "learning_rate": 9.50291529186862e-06,
+      "loss": 0.2264,
+      "step": 2056
+    },
+    {
+      "epoch": 0.6869260310569377,
+      "grad_norm": 0.45443749767645863,
+      "learning_rate": 9.502070215933177e-06,
+      "loss": 0.2247,
+      "step": 2057
+    },
+    {
+      "epoch": 0.6872599766238103,
+      "grad_norm": 0.5001191908128457,
+      "learning_rate": 9.501224459907019e-06,
+      "loss": 0.229,
+      "step": 2058
+    },
+    {
+      "epoch": 0.6875939221906829,
+      "grad_norm": 0.4682122413633237,
+      "learning_rate": 9.500378023917906e-06,
+      "loss": 0.2206,
+      "step": 2059
+    },
+    {
+      "epoch": 0.6879278677575555,
+      "grad_norm": 0.48110266017390535,
+      "learning_rate": 9.499530908093702e-06,
+      "loss": 0.2237,
+      "step": 2060
+    },
+    {
+      "epoch": 0.6882618133244282,
+      "grad_norm": 0.4758868687878399,
+      "learning_rate": 9.498683112562374e-06,
+      "loss": 0.2292,
+      "step": 2061
+    },
+    {
+      "epoch": 0.6885957588913008,
+      "grad_norm": 0.4398159757318568,
+      "learning_rate": 9.497834637451992e-06,
+      "loss": 0.2194,
+      "step": 2062
+    },
+    {
+      "epoch": 0.6889297044581734,
+      "grad_norm": 0.4806638031099439,
+      "learning_rate": 9.496985482890728e-06,
+      "loss": 0.2204,
+      "step": 2063
+    },
+    {
+      "epoch": 0.689263650025046,
+      "grad_norm": 0.7573298106478715,
+      "learning_rate": 9.496135649006857e-06,
+      "loss": 0.2451,
+      "step": 2064
+    },
+    {
+      "epoch": 0.6895975955919185,
+      "grad_norm": 0.4398493907855471,
+      "learning_rate": 9.495285135928755e-06,
+      "loss": 0.2357,
+      "step": 2065
+    },
+    {
+      "epoch": 0.6899315411587911,
+      "grad_norm": 0.4722521081195024,
+      "learning_rate": 9.494433943784901e-06,
+      "loss": 0.2284,
+      "step": 2066
+    },
+    {
+      "epoch": 0.6902654867256637,
+      "grad_norm": 0.44358844114458357,
+      "learning_rate": 9.493582072703883e-06,
+      "loss": 0.2094,
+      "step": 2067
+    },
+    {
+      "epoch": 0.6905994322925363,
+      "grad_norm": 0.5388979510188426,
+      "learning_rate": 9.49272952281438e-06,
+      "loss": 0.2498,
+      "step": 2068
+    },
+    {
+      "epoch": 0.6909333778594089,
+      "grad_norm": 0.4401449144477547,
+      "learning_rate": 9.491876294245184e-06,
+      "loss": 0.2288,
+      "step": 2069
+    },
+    {
+      "epoch": 0.6912673234262815,
+      "grad_norm": 0.5258022574749537,
+      "learning_rate": 9.491022387125183e-06,
+      "loss": 0.2227,
+      "step": 2070
+    },
+    {
+      "epoch": 0.6916012689931541,
+      "grad_norm": 0.484024323700997,
+      "learning_rate": 9.490167801583373e-06,
+      "loss": 0.2139,
+      "step": 2071
+    },
+    {
+      "epoch": 0.6919352145600267,
+      "grad_norm": 0.506106087923025,
+      "learning_rate": 9.489312537748843e-06,
+      "loss": 0.2431,
+      "step": 2072
+    },
+    {
+      "epoch": 0.6922691601268993,
+      "grad_norm": 0.4585978711248216,
+      "learning_rate": 9.488456595750795e-06,
+      "loss": 0.2339,
+      "step": 2073
+    },
+    {
+      "epoch": 0.6926031056937719,
+      "grad_norm": 0.6031424608851983,
+      "learning_rate": 9.487599975718529e-06,
+      "loss": 0.2433,
+      "step": 2074
+    },
+    {
+      "epoch": 0.6929370512606445,
+      "grad_norm": 0.4556413453192436,
+      "learning_rate": 9.486742677781446e-06,
+      "loss": 0.2189,
+      "step": 2075
+    },
+    {
+      "epoch": 0.6932709968275171,
+      "grad_norm": 0.687156099033192,
+      "learning_rate": 9.485884702069053e-06,
+      "loss": 0.2343,
+      "step": 2076
+    },
+    {
+      "epoch": 0.6936049423943897,
+      "grad_norm": 0.4659994646594298,
+      "learning_rate": 9.485026048710957e-06,
+      "loss": 0.2295,
+      "step": 2077
+    },
+    {
+      "epoch": 0.6939388879612624,
+      "grad_norm": 0.5014688431544414,
+      "learning_rate": 9.484166717836865e-06,
+      "loss": 0.234,
+      "step": 2078
+    },
+    {
+      "epoch": 0.694272833528135,
+      "grad_norm": 0.5334695156505862,
+      "learning_rate": 9.48330670957659e-06,
+      "loss": 0.2192,
+      "step": 2079
+    },
+    {
+      "epoch": 0.6946067790950076,
+      "grad_norm": 0.5223706657750462,
+      "learning_rate": 9.48244602406005e-06,
+      "loss": 0.2254,
+      "step": 2080
+    },
+    {
+      "epoch": 0.6949407246618801,
+      "grad_norm": 0.5463595012524518,
+      "learning_rate": 9.481584661417258e-06,
+      "loss": 0.2254,
+      "step": 2081
+    },
+    {
+      "epoch": 0.6952746702287527,
+      "grad_norm": 0.48561380690299644,
+      "learning_rate": 9.480722621778334e-06,
+      "loss": 0.2277,
+      "step": 2082
+    },
+    {
+      "epoch": 0.6956086157956253,
+      "grad_norm": 0.4492219269705461,
+      "learning_rate": 9.479859905273498e-06,
+      "loss": 0.2236,
+      "step": 2083
+    },
+    {
+      "epoch": 0.6959425613624979,
+      "grad_norm": 0.5504216916036948,
+      "learning_rate": 9.478996512033074e-06,
+      "loss": 0.2467,
+      "step": 2084
+    },
+    {
+      "epoch": 0.6962765069293705,
+      "grad_norm": 0.5111070325627567,
+      "learning_rate": 9.478132442187491e-06,
+      "loss": 0.2112,
+      "step": 2085
+    },
+    {
+      "epoch": 0.6966104524962431,
+      "grad_norm": 0.47186041559580094,
+      "learning_rate": 9.477267695867275e-06,
+      "loss": 0.2229,
+      "step": 2086
+    },
+    {
+      "epoch": 0.6969443980631157,
+      "grad_norm": 0.4488571924469654,
+      "learning_rate": 9.476402273203052e-06,
+      "loss": 0.2225,
+      "step": 2087
+    },
+    {
+      "epoch": 0.6972783436299883,
+      "grad_norm": 0.5787574166594714,
+      "learning_rate": 9.47553617432556e-06,
+      "loss": 0.2335,
+      "step": 2088
+    },
+    {
+      "epoch": 0.6976122891968609,
+      "grad_norm": 0.49866311539288444,
+      "learning_rate": 9.47466939936563e-06,
+      "loss": 0.2237,
+      "step": 2089
+    },
+    {
+      "epoch": 0.6979462347637335,
+      "grad_norm": 0.7548546423555135,
+      "learning_rate": 9.473801948454199e-06,
+      "loss": 0.23,
+      "step": 2090
+    },
+    {
+      "epoch": 0.6982801803306061,
+      "grad_norm": 0.49007999916257805,
+      "learning_rate": 9.472933821722307e-06,
+      "loss": 0.2166,
+      "step": 2091
+    },
+    {
+      "epoch": 0.6986141258974787,
+      "grad_norm": 0.47535667454395897,
+      "learning_rate": 9.472065019301095e-06,
+      "loss": 0.2184,
+      "step": 2092
+    },
+    {
+      "epoch": 0.6989480714643513,
+      "grad_norm": 0.5211167728844137,
+      "learning_rate": 9.471195541321805e-06,
+      "loss": 0.2176,
+      "step": 2093
+    },
+    {
+      "epoch": 0.699282017031224,
+      "grad_norm": 0.5003555535014772,
+      "learning_rate": 9.470325387915782e-06,
+      "loss": 0.2353,
+      "step": 2094
+    },
+    {
+      "epoch": 0.6996159625980966,
+      "grad_norm": 0.48728850394662204,
+      "learning_rate": 9.469454559214473e-06,
+      "loss": 0.2278,
+      "step": 2095
+    },
+    {
+      "epoch": 0.6999499081649692,
+      "grad_norm": 0.5088758728966357,
+      "learning_rate": 9.468583055349425e-06,
+      "loss": 0.22,
+      "step": 2096
+    },
+    {
+      "epoch": 0.7002838537318417,
+      "grad_norm": 0.4368236017414324,
+      "learning_rate": 9.467710876452292e-06,
+      "loss": 0.2233,
+      "step": 2097
+    },
+    {
+      "epoch": 0.7006177992987143,
+      "grad_norm": 0.5077021290184639,
+      "learning_rate": 9.466838022654826e-06,
+      "loss": 0.2352,
+      "step": 2098
+    },
+    {
+      "epoch": 0.7009517448655869,
+      "grad_norm": 0.5513449938345952,
+      "learning_rate": 9.465964494088879e-06,
+      "loss": 0.2327,
+      "step": 2099
+    },
+    {
+      "epoch": 0.7012856904324595,
+      "grad_norm": 0.4771101109092869,
+      "learning_rate": 9.465090290886411e-06,
+      "loss": 0.2305,
+      "step": 2100
+    },
+    {
+      "epoch": 0.7016196359993321,
+      "grad_norm": 0.5415328425413375,
+      "learning_rate": 9.464215413179483e-06,
+      "loss": 0.2424,
+      "step": 2101
+    },
+    {
+      "epoch": 0.7019535815662047,
+      "grad_norm": 0.49226725525209947,
+      "learning_rate": 9.46333986110025e-06,
+      "loss": 0.2095,
+      "step": 2102
+    },
+    {
+      "epoch": 0.7022875271330773,
+      "grad_norm": 0.47931978869160313,
+      "learning_rate": 9.462463634780977e-06,
+      "loss": 0.2221,
+      "step": 2103
+    },
+    {
+      "epoch": 0.7026214726999499,
+      "grad_norm": 0.4655415271549972,
+      "learning_rate": 9.461586734354027e-06,
+      "loss": 0.2201,
+      "step": 2104
+    },
+    {
+      "epoch": 0.7029554182668225,
+      "grad_norm": 0.5668783713703968,
+      "learning_rate": 9.460709159951867e-06,
+      "loss": 0.2385,
+      "step": 2105
+    },
+    {
+      "epoch": 0.7032893638336951,
+      "grad_norm": 0.4442504668426384,
+      "learning_rate": 9.459830911707066e-06,
+      "loss": 0.2174,
+      "step": 2106
+    },
+    {
+      "epoch": 0.7036233094005677,
+      "grad_norm": 0.5148802971653738,
+      "learning_rate": 9.458951989752295e-06,
+      "loss": 0.2298,
+      "step": 2107
+    },
+    {
+      "epoch": 0.7039572549674403,
+      "grad_norm": 0.4961289354471199,
+      "learning_rate": 9.458072394220321e-06,
+      "loss": 0.2365,
+      "step": 2108
+    },
+    {
+      "epoch": 0.7042912005343129,
+      "grad_norm": 0.39859284746010964,
+      "learning_rate": 9.457192125244021e-06,
+      "loss": 0.2026,
+      "step": 2109
+    },
+    {
+      "epoch": 0.7046251461011855,
+      "grad_norm": 0.43207299712712405,
+      "learning_rate": 9.456311182956368e-06,
+      "loss": 0.2191,
+      "step": 2110
+    },
+    {
+      "epoch": 0.7049590916680581,
+      "grad_norm": 0.4752148941522198,
+      "learning_rate": 9.45542956749044e-06,
+      "loss": 0.2325,
+      "step": 2111
+    },
+    {
+      "epoch": 0.7052930372349308,
+      "grad_norm": 0.5010644723261528,
+      "learning_rate": 9.454547278979415e-06,
+      "loss": 0.2191,
+      "step": 2112
+    },
+    {
+      "epoch": 0.7056269828018034,
+      "grad_norm": 0.40752873703821485,
+      "learning_rate": 9.453664317556572e-06,
+      "loss": 0.2106,
+      "step": 2113
+    },
+    {
+      "epoch": 0.7059609283686759,
+      "grad_norm": 0.5576876661906841,
+      "learning_rate": 9.452780683355295e-06,
+      "loss": 0.222,
+      "step": 2114
+    },
+    {
+      "epoch": 0.7062948739355485,
+      "grad_norm": 0.432810708164213,
+      "learning_rate": 9.451896376509065e-06,
+      "loss": 0.2113,
+      "step": 2115
+    },
+    {
+      "epoch": 0.7066288195024211,
+      "grad_norm": 0.4482096027677475,
+      "learning_rate": 9.451011397151469e-06,
+      "loss": 0.2209,
+      "step": 2116
+    },
+    {
+      "epoch": 0.7069627650692937,
+      "grad_norm": 0.4366216644067192,
+      "learning_rate": 9.450125745416191e-06,
+      "loss": 0.2107,
+      "step": 2117
+    },
+    {
+      "epoch": 0.7072967106361663,
+      "grad_norm": 0.4948001961193805,
+      "learning_rate": 9.44923942143702e-06,
+      "loss": 0.2346,
+      "step": 2118
+    },
+    {
+      "epoch": 0.7076306562030389,
+      "grad_norm": 0.8894874375807659,
+      "learning_rate": 9.448352425347848e-06,
+      "loss": 0.2458,
+      "step": 2119
+    },
+    {
+      "epoch": 0.7079646017699115,
+      "grad_norm": 0.5299843319037437,
+      "learning_rate": 9.447464757282665e-06,
+      "loss": 0.2367,
+      "step": 2120
+    },
+    {
+      "epoch": 0.7082985473367841,
+      "grad_norm": 0.4727096807635542,
+      "learning_rate": 9.44657641737556e-06,
+      "loss": 0.2327,
+      "step": 2121
+    },
+    {
+      "epoch": 0.7086324929036567,
+      "grad_norm": 0.4600930804291522,
+      "learning_rate": 9.445687405760735e-06,
+      "loss": 0.2226,
+      "step": 2122
+    },
+    {
+      "epoch": 0.7089664384705293,
+      "grad_norm": 0.4332929425486284,
+      "learning_rate": 9.444797722572479e-06,
+      "loss": 0.2107,
+      "step": 2123
+    },
+    {
+      "epoch": 0.7093003840374019,
+      "grad_norm": 0.731912468149352,
+      "learning_rate": 9.44390736794519e-06,
+      "loss": 0.238,
+      "step": 2124
+    },
+    {
+      "epoch": 0.7096343296042745,
+      "grad_norm": 0.5081533596669653,
+      "learning_rate": 9.443016342013369e-06,
+      "loss": 0.2351,
+      "step": 2125
+    },
+    {
+      "epoch": 0.7099682751711471,
+      "grad_norm": 0.5148768327746467,
+      "learning_rate": 9.442124644911614e-06,
+      "loss": 0.2228,
+      "step": 2126
+    },
+    {
+      "epoch": 0.7103022207380197,
+      "grad_norm": 0.4889377936506926,
+      "learning_rate": 9.441232276774629e-06,
+      "loss": 0.2185,
+      "step": 2127
+    },
+    {
+      "epoch": 0.7106361663048923,
+      "grad_norm": 0.48641943614807776,
+      "learning_rate": 9.440339237737213e-06,
+      "loss": 0.2201,
+      "step": 2128
+    },
+    {
+      "epoch": 0.710970111871765,
+      "grad_norm": 0.5242633602293272,
+      "learning_rate": 9.439445527934272e-06,
+      "loss": 0.2176,
+      "step": 2129
+    },
+    {
+      "epoch": 0.7113040574386374,
+      "grad_norm": 0.5762365105646677,
+      "learning_rate": 9.438551147500812e-06,
+      "loss": 0.2338,
+      "step": 2130
+    },
+    {
+      "epoch": 0.7116380030055101,
+      "grad_norm": 0.48698836807254486,
+      "learning_rate": 9.437656096571938e-06,
+      "loss": 0.2287,
+      "step": 2131
+    },
+    {
+      "epoch": 0.7119719485723827,
+      "grad_norm": 0.5525740892373583,
+      "learning_rate": 9.436760375282858e-06,
+      "loss": 0.2395,
+      "step": 2132
+    },
+    {
+      "epoch": 0.7123058941392553,
+      "grad_norm": 0.6121087031250623,
+      "learning_rate": 9.435863983768884e-06,
+      "loss": 0.2327,
+      "step": 2133
+    },
+    {
+      "epoch": 0.7126398397061279,
+      "grad_norm": 0.5393868118238357,
+      "learning_rate": 9.434966922165424e-06,
+      "loss": 0.242,
+      "step": 2134
+    },
+    {
+      "epoch": 0.7129737852730005,
+      "grad_norm": 0.5338938325124174,
+      "learning_rate": 9.43406919060799e-06,
+      "loss": 0.2094,
+      "step": 2135
+    },
+    {
+      "epoch": 0.7133077308398731,
+      "grad_norm": 0.4834589668720248,
+      "learning_rate": 9.433170789232196e-06,
+      "loss": 0.207,
+      "step": 2136
+    },
+    {
+      "epoch": 0.7136416764067457,
+      "grad_norm": 0.636163814834721,
+      "learning_rate": 9.432271718173756e-06,
+      "loss": 0.2373,
+      "step": 2137
+    },
+    {
+      "epoch": 0.7139756219736183,
+      "grad_norm": 0.5423717816603074,
+      "learning_rate": 9.431371977568483e-06,
+      "loss": 0.2398,
+      "step": 2138
+    },
+    {
+      "epoch": 0.7143095675404909,
+      "grad_norm": 0.6024062328191949,
+      "learning_rate": 9.430471567552295e-06,
+      "loss": 0.2215,
+      "step": 2139
+    },
+    {
+      "epoch": 0.7146435131073635,
+      "grad_norm": 0.5675673827510521,
+      "learning_rate": 9.42957048826121e-06,
+      "loss": 0.2296,
+      "step": 2140
+    },
+    {
+      "epoch": 0.7149774586742361,
+      "grad_norm": 0.47911769123994113,
+      "learning_rate": 9.428668739831349e-06,
+      "loss": 0.2133,
+      "step": 2141
+    },
+    {
+      "epoch": 0.7153114042411087,
+      "grad_norm": 0.5032255216977924,
+      "learning_rate": 9.427766322398926e-06,
+      "loss": 0.228,
+      "step": 2142
+    },
+    {
+      "epoch": 0.7156453498079813,
+      "grad_norm": 0.4892738328424283,
+      "learning_rate": 9.426863236100266e-06,
+      "loss": 0.2315,
+      "step": 2143
+    },
+    {
+      "epoch": 0.7159792953748539,
+      "grad_norm": 0.5488139396113335,
+      "learning_rate": 9.425959481071787e-06,
+      "loss": 0.2283,
+      "step": 2144
+    },
+    {
+      "epoch": 0.7163132409417265,
+      "grad_norm": 0.46181208491623815,
+      "learning_rate": 9.425055057450017e-06,
+      "loss": 0.2199,
+      "step": 2145
+    },
+    {
+      "epoch": 0.716647186508599,
+      "grad_norm": 0.47958048701662137,
+      "learning_rate": 9.424149965371576e-06,
+      "loss": 0.2377,
+      "step": 2146
+    },
+    {
+      "epoch": 0.7169811320754716,
+      "grad_norm": 0.533708758771216,
+      "learning_rate": 9.423244204973191e-06,
+      "loss": 0.2263,
+      "step": 2147
+    },
+    {
+      "epoch": 0.7173150776423443,
+      "grad_norm": 0.47032392487456925,
+      "learning_rate": 9.422337776391686e-06,
+      "loss": 0.2139,
+      "step": 2148
+    },
+    {
+      "epoch": 0.7176490232092169,
+      "grad_norm": 0.4681306879448958,
+      "learning_rate": 9.421430679763989e-06,
+      "loss": 0.2217,
+      "step": 2149
+    },
+    {
+      "epoch": 0.7179829687760895,
+      "grad_norm": 0.4611059292110921,
+      "learning_rate": 9.420522915227129e-06,
+      "loss": 0.2262,
+      "step": 2150
+    },
+    {
+      "epoch": 0.7183169143429621,
+      "grad_norm": 0.4649061259338331,
+      "learning_rate": 9.419614482918229e-06,
+      "loss": 0.2194,
+      "step": 2151
+    },
+    {
+      "epoch": 0.7186508599098347,
+      "grad_norm": 0.49767814980354536,
+      "learning_rate": 9.418705382974524e-06,
+      "loss": 0.2165,
+      "step": 2152
+    },
+    {
+      "epoch": 0.7189848054767073,
+      "grad_norm": 0.4876893264124481,
+      "learning_rate": 9.417795615533343e-06,
+      "loss": 0.2306,
+      "step": 2153
+    },
+    {
+      "epoch": 0.7193187510435799,
+      "grad_norm": 0.5573835824314209,
+      "learning_rate": 9.416885180732115e-06,
+      "loss": 0.241,
+      "step": 2154
+    },
+    {
+      "epoch": 0.7196526966104525,
+      "grad_norm": 0.42980907750226544,
+      "learning_rate": 9.415974078708375e-06,
+      "loss": 0.2142,
+      "step": 2155
+    },
+    {
+      "epoch": 0.7199866421773251,
+      "grad_norm": 0.4581280120482188,
+      "learning_rate": 9.415062309599751e-06,
+      "loss": 0.225,
+      "step": 2156
+    },
+    {
+      "epoch": 0.7203205877441977,
+      "grad_norm": 0.4793267456929462,
+      "learning_rate": 9.414149873543983e-06,
+      "loss": 0.2216,
+      "step": 2157
+    },
+    {
+      "epoch": 0.7206545333110703,
+      "grad_norm": 0.5360592011567692,
+      "learning_rate": 9.4132367706789e-06,
+      "loss": 0.207,
+      "step": 2158
+    },
+    {
+      "epoch": 0.7209884788779429,
+      "grad_norm": 0.422519463252475,
+      "learning_rate": 9.412323001142438e-06,
+      "loss": 0.2048,
+      "step": 2159
+    },
+    {
+      "epoch": 0.7213224244448155,
+      "grad_norm": 0.47418431358564955,
+      "learning_rate": 9.411408565072635e-06,
+      "loss": 0.2353,
+      "step": 2160
+    },
+    {
+      "epoch": 0.7216563700116881,
+      "grad_norm": 0.43728181066282573,
+      "learning_rate": 9.410493462607623e-06,
+      "loss": 0.2095,
+      "step": 2161
+    },
+    {
+      "epoch": 0.7219903155785607,
+      "grad_norm": 0.5017096018356688,
+      "learning_rate": 9.409577693885642e-06,
+      "loss": 0.2326,
+      "step": 2162
+    },
+    {
+      "epoch": 0.7223242611454332,
+      "grad_norm": 0.45202500297613973,
+      "learning_rate": 9.408661259045032e-06,
+      "loss": 0.215,
+      "step": 2163
+    },
+    {
+      "epoch": 0.7226582067123059,
+      "grad_norm": 0.49163901550448014,
+      "learning_rate": 9.407744158224227e-06,
+      "loss": 0.2149,
+      "step": 2164
+    },
+    {
+      "epoch": 0.7229921522791785,
+      "grad_norm": 0.49708508765871556,
+      "learning_rate": 9.406826391561767e-06,
+      "loss": 0.228,
+      "step": 2165
+    },
+    {
+      "epoch": 0.7233260978460511,
+      "grad_norm": 0.5530402316677213,
+      "learning_rate": 9.405907959196293e-06,
+      "loss": 0.2534,
+      "step": 2166
+    },
+    {
+      "epoch": 0.7236600434129237,
+      "grad_norm": 0.4590331097109294,
+      "learning_rate": 9.404988861266543e-06,
+      "loss": 0.2127,
+      "step": 2167
+    },
+    {
+      "epoch": 0.7239939889797963,
+      "grad_norm": 0.5061377928038171,
+      "learning_rate": 9.404069097911358e-06,
+      "loss": 0.2181,
+      "step": 2168
+    },
+    {
+      "epoch": 0.7243279345466689,
+      "grad_norm": 0.4485152013068261,
+      "learning_rate": 9.40314866926968e-06,
+      "loss": 0.2113,
+      "step": 2169
+    },
+    {
+      "epoch": 0.7246618801135415,
+      "grad_norm": 0.4916280062892611,
+      "learning_rate": 9.402227575480549e-06,
+      "loss": 0.2363,
+      "step": 2170
+    },
+    {
+      "epoch": 0.7249958256804141,
+      "grad_norm": 0.4369101484429515,
+      "learning_rate": 9.401305816683111e-06,
+      "loss": 0.2022,
+      "step": 2171
+    },
+    {
+      "epoch": 0.7253297712472867,
+      "grad_norm": 0.4506814138302595,
+      "learning_rate": 9.400383393016604e-06,
+      "loss": 0.2273,
+      "step": 2172
+    },
+    {
+      "epoch": 0.7256637168141593,
+      "grad_norm": 0.4383138721900733,
+      "learning_rate": 9.39946030462037e-06,
+      "loss": 0.2058,
+      "step": 2173
+    },
+    {
+      "epoch": 0.7259976623810319,
+      "grad_norm": 0.4733775066633754,
+      "learning_rate": 9.39853655163386e-06,
+      "loss": 0.2033,
+      "step": 2174
+    },
+    {
+      "epoch": 0.7263316079479045,
+      "grad_norm": 0.44987005012925274,
+      "learning_rate": 9.39761213419661e-06,
+      "loss": 0.2093,
+      "step": 2175
+    },
+    {
+      "epoch": 0.7266655535147771,
+      "grad_norm": 0.44081375935995787,
+      "learning_rate": 9.396687052448267e-06,
+      "loss": 0.2151,
+      "step": 2176
+    },
+    {
+      "epoch": 0.7269994990816497,
+      "grad_norm": 0.4668970976585449,
+      "learning_rate": 9.395761306528576e-06,
+      "loss": 0.2111,
+      "step": 2177
+    },
+    {
+      "epoch": 0.7273334446485223,
+      "grad_norm": 0.4762785397500771,
+      "learning_rate": 9.39483489657738e-06,
+      "loss": 0.212,
+      "step": 2178
+    },
+    {
+      "epoch": 0.7276673902153948,
+      "grad_norm": 0.447677116890413,
+      "learning_rate": 9.393907822734627e-06,
+      "loss": 0.22,
+      "step": 2179
+    },
+    {
+      "epoch": 0.7280013357822674,
+      "grad_norm": 0.535442818777305,
+      "learning_rate": 9.39298008514036e-06,
+      "loss": 0.2372,
+      "step": 2180
+    },
+    {
+      "epoch": 0.72833528134914,
+      "grad_norm": 0.5188386933388953,
+      "learning_rate": 9.392051683934726e-06,
+      "loss": 0.2268,
+      "step": 2181
+    },
+    {
+      "epoch": 0.7286692269160127,
+      "grad_norm": 0.4727820750476286,
+      "learning_rate": 9.39112261925797e-06,
+      "loss": 0.2179,
+      "step": 2182
+    },
+    {
+      "epoch": 0.7290031724828853,
+      "grad_norm": 0.4854734035330415,
+      "learning_rate": 9.390192891250439e-06,
+      "loss": 0.2406,
+      "step": 2183
+    },
+    {
+      "epoch": 0.7293371180497579,
+      "grad_norm": 1.4538057381250347,
+      "learning_rate": 9.389262500052578e-06,
+      "loss": 0.226,
+      "step": 2184
+    },
+    {
+      "epoch": 0.7296710636166305,
+      "grad_norm": 0.511345482598057,
+      "learning_rate": 9.388331445804935e-06,
+      "loss": 0.2238,
+      "step": 2185
+    },
+    {
+      "epoch": 0.7300050091835031,
+      "grad_norm": 0.49329394070154936,
+      "learning_rate": 9.387399728648156e-06,
+      "loss": 0.2164,
+      "step": 2186
+    },
+    {
+      "epoch": 0.7303389547503757,
+      "grad_norm": 0.47261545953275097,
+      "learning_rate": 9.386467348722989e-06,
+      "loss": 0.2369,
+      "step": 2187
+    },
+    {
+      "epoch": 0.7306729003172483,
+      "grad_norm": 0.4970024372138262,
+      "learning_rate": 9.385534306170279e-06,
+      "loss": 0.2337,
+      "step": 2188
+    },
+    {
+      "epoch": 0.7310068458841209,
+      "grad_norm": 0.4407037790976595,
+      "learning_rate": 9.384600601130973e-06,
+      "loss": 0.2181,
+      "step": 2189
+    },
+    {
+      "epoch": 0.7313407914509935,
+      "grad_norm": 0.4612359667383596,
+      "learning_rate": 9.383666233746121e-06,
+      "loss": 0.213,
+      "step": 2190
+    },
+    {
+      "epoch": 0.7316747370178661,
+      "grad_norm": 0.5683151991367129,
+      "learning_rate": 9.382731204156869e-06,
+      "loss": 0.2403,
+      "step": 2191
+    },
+    {
+      "epoch": 0.7320086825847387,
+      "grad_norm": 0.6075346410739748,
+      "learning_rate": 9.381795512504461e-06,
+      "loss": 0.2251,
+      "step": 2192
+    },
+    {
+      "epoch": 0.7323426281516113,
+      "grad_norm": 0.7259845103570428,
+      "learning_rate": 9.380859158930249e-06,
+      "loss": 0.2379,
+      "step": 2193
+    },
+    {
+      "epoch": 0.7326765737184839,
+      "grad_norm": 0.47741648545629145,
+      "learning_rate": 9.379922143575678e-06,
+      "loss": 0.2191,
+      "step": 2194
+    },
+    {
+      "epoch": 0.7330105192853564,
+      "grad_norm": 0.45871484382317196,
+      "learning_rate": 9.378984466582294e-06,
+      "loss": 0.207,
+      "step": 2195
+    },
+    {
+      "epoch": 0.733344464852229,
+      "grad_norm": 0.4698346774760905,
+      "learning_rate": 9.378046128091748e-06,
+      "loss": 0.2156,
+      "step": 2196
+    },
+    {
+      "epoch": 0.7336784104191016,
+      "grad_norm": 0.517374409193277,
+      "learning_rate": 9.377107128245782e-06,
+      "loss": 0.2168,
+      "step": 2197
+    },
+    {
+      "epoch": 0.7340123559859743,
+      "grad_norm": 0.4960261055253891,
+      "learning_rate": 9.376167467186246e-06,
+      "loss": 0.2302,
+      "step": 2198
+    },
+    {
+      "epoch": 0.7343463015528469,
+      "grad_norm": 0.5756540347478746,
+      "learning_rate": 9.375227145055085e-06,
+      "loss": 0.2132,
+      "step": 2199
+    },
+    {
+      "epoch": 0.7346802471197195,
+      "grad_norm": 0.5091682769372302,
+      "learning_rate": 9.374286161994351e-06,
+      "loss": 0.2282,
+      "step": 2200
+    },
+    {
+      "epoch": 0.7350141926865921,
+      "grad_norm": 0.5067877827964193,
+      "learning_rate": 9.373344518146184e-06,
+      "loss": 0.2359,
+      "step": 2201
+    },
+    {
+      "epoch": 0.7353481382534647,
+      "grad_norm": 0.5423740124921783,
+      "learning_rate": 9.372402213652833e-06,
+      "loss": 0.2315,
+      "step": 2202
+    },
+    {
+      "epoch": 0.7356820838203373,
+      "grad_norm": 0.453426255813076,
+      "learning_rate": 9.371459248656645e-06,
+      "loss": 0.2123,
+      "step": 2203
+    },
+    {
+      "epoch": 0.7360160293872099,
+      "grad_norm": 0.5465347054793437,
+      "learning_rate": 9.370515623300066e-06,
+      "loss": 0.2203,
+      "step": 2204
+    },
+    {
+      "epoch": 0.7363499749540825,
+      "grad_norm": 0.5653821901957052,
+      "learning_rate": 9.369571337725638e-06,
+      "loss": 0.2308,
+      "step": 2205
+    },
+    {
+      "epoch": 0.7366839205209551,
+      "grad_norm": 0.5246706912753366,
+      "learning_rate": 9.368626392076013e-06,
+      "loss": 0.2288,
+      "step": 2206
+    },
+    {
+      "epoch": 0.7370178660878277,
+      "grad_norm": 0.42226917655168017,
+      "learning_rate": 9.367680786493929e-06,
+      "loss": 0.2245,
+      "step": 2207
+    },
+    {
+      "epoch": 0.7373518116547003,
+      "grad_norm": 0.5148652520612048,
+      "learning_rate": 9.366734521122236e-06,
+      "loss": 0.2121,
+      "step": 2208
+    },
+    {
+      "epoch": 0.7376857572215729,
+      "grad_norm": 0.49636706629870225,
+      "learning_rate": 9.365787596103877e-06,
+      "loss": 0.2422,
+      "step": 2209
+    },
+    {
+      "epoch": 0.7380197027884455,
+      "grad_norm": 0.5272484506883985,
+      "learning_rate": 9.364840011581896e-06,
+      "loss": 0.2361,
+      "step": 2210
+    },
+    {
+      "epoch": 0.7383536483553181,
+      "grad_norm": 0.48013110860491137,
+      "learning_rate": 9.363891767699437e-06,
+      "loss": 0.2289,
+      "step": 2211
+    },
+    {
+      "epoch": 0.7386875939221906,
+      "grad_norm": 0.4464703818992004,
+      "learning_rate": 9.362942864599746e-06,
+      "loss": 0.2189,
+      "step": 2212
+    },
+    {
+      "epoch": 0.7390215394890632,
+      "grad_norm": 0.49475924194998067,
+      "learning_rate": 9.36199330242616e-06,
+      "loss": 0.2301,
+      "step": 2213
+    },
+    {
+      "epoch": 0.7393554850559358,
+      "grad_norm": 0.46063739425578615,
+      "learning_rate": 9.361043081322125e-06,
+      "loss": 0.2268,
+      "step": 2214
+    },
+    {
+      "epoch": 0.7396894306228085,
+      "grad_norm": 0.4741690736245642,
+      "learning_rate": 9.360092201431186e-06,
+      "loss": 0.2333,
+      "step": 2215
+    },
+    {
+      "epoch": 0.7400233761896811,
+      "grad_norm": 0.5013668916419637,
+      "learning_rate": 9.359140662896978e-06,
+      "loss": 0.2202,
+      "step": 2216
+    },
+    {
+      "epoch": 0.7403573217565537,
+      "grad_norm": 0.4308948496107637,
+      "learning_rate": 9.358188465863247e-06,
+      "loss": 0.2086,
+      "step": 2217
+    },
+    {
+      "epoch": 0.7406912673234263,
+      "grad_norm": 0.4175975145931895,
+      "learning_rate": 9.357235610473833e-06,
+      "loss": 0.2091,
+      "step": 2218
+    },
+    {
+      "epoch": 0.7410252128902989,
+      "grad_norm": 0.4655604424669319,
+      "learning_rate": 9.356282096872673e-06,
+      "loss": 0.23,
+      "step": 2219
+    },
+    {
+      "epoch": 0.7413591584571715,
+      "grad_norm": 0.4740711739925384,
+      "learning_rate": 9.355327925203811e-06,
+      "loss": 0.2176,
+      "step": 2220
+    },
+    {
+      "epoch": 0.7416931040240441,
+      "grad_norm": 0.4748639281311392,
+      "learning_rate": 9.354373095611383e-06,
+      "loss": 0.2359,
+      "step": 2221
+    },
+    {
+      "epoch": 0.7420270495909167,
+      "grad_norm": 0.4368036977894943,
+      "learning_rate": 9.353417608239627e-06,
+      "loss": 0.2171,
+      "step": 2222
+    },
+    {
+      "epoch": 0.7423609951577893,
+      "grad_norm": 0.4666842076480527,
+      "learning_rate": 9.352461463232882e-06,
+      "loss": 0.2324,
+      "step": 2223
+    },
+    {
+      "epoch": 0.7426949407246619,
+      "grad_norm": 0.48937622343211834,
+      "learning_rate": 9.351504660735583e-06,
+      "loss": 0.2247,
+      "step": 2224
+    },
+    {
+      "epoch": 0.7430288862915345,
+      "grad_norm": 0.4711175442547071,
+      "learning_rate": 9.350547200892271e-06,
+      "loss": 0.2225,
+      "step": 2225
+    },
+    {
+      "epoch": 0.7433628318584071,
+      "grad_norm": 0.47189505898784784,
+      "learning_rate": 9.349589083847577e-06,
+      "loss": 0.2336,
+      "step": 2226
+    },
+    {
+      "epoch": 0.7436967774252797,
+      "grad_norm": 0.49998037587574656,
+      "learning_rate": 9.348630309746236e-06,
+      "loss": 0.242,
+      "step": 2227
+    },
+    {
+      "epoch": 0.7440307229921522,
+      "grad_norm": 0.4389783954854127,
+      "learning_rate": 9.347670878733084e-06,
+      "loss": 0.2179,
+      "step": 2228
+    },
+    {
+      "epoch": 0.7443646685590248,
+      "grad_norm": 0.46169101406158664,
+      "learning_rate": 9.346710790953053e-06,
+      "loss": 0.2259,
+      "step": 2229
+    },
+    {
+      "epoch": 0.7446986141258974,
+      "grad_norm": 0.43050129747366644,
+      "learning_rate": 9.345750046551177e-06,
+      "loss": 0.211,
+      "step": 2230
+    },
+    {
+      "epoch": 0.74503255969277,
+      "grad_norm": 0.4296786556030227,
+      "learning_rate": 9.344788645672585e-06,
+      "loss": 0.2237,
+      "step": 2231
+    },
+    {
+      "epoch": 0.7453665052596427,
+      "grad_norm": 0.5686148097924658,
+      "learning_rate": 9.343826588462513e-06,
+      "loss": 0.2344,
+      "step": 2232
+    },
+    {
+      "epoch": 0.7457004508265153,
+      "grad_norm": 0.4233257630785193,
+      "learning_rate": 9.342863875066284e-06,
+      "loss": 0.2123,
+      "step": 2233
+    },
+    {
+      "epoch": 0.7460343963933879,
+      "grad_norm": 0.437653238005159,
+      "learning_rate": 9.341900505629333e-06,
+      "loss": 0.2154,
+      "step": 2234
+    },
+    {
+      "epoch": 0.7463683419602605,
+      "grad_norm": 0.45698025308598117,
+      "learning_rate": 9.340936480297187e-06,
+      "loss": 0.2295,
+      "step": 2235
+    },
+    {
+      "epoch": 0.7467022875271331,
+      "grad_norm": 0.4688987813684971,
+      "learning_rate": 9.339971799215472e-06,
+      "loss": 0.226,
+      "step": 2236
+    },
+    {
+      "epoch": 0.7470362330940057,
+      "grad_norm": 0.4760421598172396,
+      "learning_rate": 9.339006462529916e-06,
+      "loss": 0.2334,
+      "step": 2237
+    },
+    {
+      "epoch": 0.7473701786608783,
+      "grad_norm": 0.4192485301306118,
+      "learning_rate": 9.338040470386344e-06,
+      "loss": 0.2122,
+      "step": 2238
+    },
+    {
+      "epoch": 0.7477041242277509,
+      "grad_norm": 0.45176630982507726,
+      "learning_rate": 9.337073822930681e-06,
+      "loss": 0.2387,
+      "step": 2239
+    },
+    {
+      "epoch": 0.7480380697946235,
+      "grad_norm": 0.4486924188323239,
+      "learning_rate": 9.336106520308948e-06,
+      "loss": 0.1943,
+      "step": 2240
+    },
+    {
+      "epoch": 0.7483720153614961,
+      "grad_norm": 0.4646813003017934,
+      "learning_rate": 9.335138562667267e-06,
+      "loss": 0.2308,
+      "step": 2241
+    },
+    {
+      "epoch": 0.7487059609283687,
+      "grad_norm": 0.46477838126704796,
+      "learning_rate": 9.334169950151866e-06,
+      "loss": 0.2234,
+      "step": 2242
+    },
+    {
+      "epoch": 0.7490399064952413,
+      "grad_norm": 0.5008667979340363,
+      "learning_rate": 9.333200682909059e-06,
+      "loss": 0.2236,
+      "step": 2243
+    },
+    {
+      "epoch": 0.7493738520621138,
+      "grad_norm": 0.48975993721972527,
+      "learning_rate": 9.332230761085265e-06,
+      "loss": 0.2269,
+      "step": 2244
+    },
+    {
+      "epoch": 0.7497077976289864,
+      "grad_norm": 0.43081722939141126,
+      "learning_rate": 9.331260184827006e-06,
+      "loss": 0.2189,
+      "step": 2245
+    },
+    {
+      "epoch": 0.750041743195859,
+      "grad_norm": 0.4627663439303731,
+      "learning_rate": 9.330288954280898e-06,
+      "loss": 0.2105,
+      "step": 2246
+    },
+    {
+      "epoch": 0.7503756887627316,
+      "grad_norm": 0.5521700146824564,
+      "learning_rate": 9.329317069593654e-06,
+      "loss": 0.2273,
+      "step": 2247
+    },
+    {
+      "epoch": 0.7507096343296042,
+      "grad_norm": 0.518812087453654,
+      "learning_rate": 9.328344530912093e-06,
+      "loss": 0.2518,
+      "step": 2248
+    },
+    {
+      "epoch": 0.7510435798964769,
+      "grad_norm": 0.5221575866002434,
+      "learning_rate": 9.327371338383124e-06,
+      "loss": 0.2458,
+      "step": 2249
+    },
+    {
+      "epoch": 0.7513775254633495,
+      "grad_norm": 0.47799468782489984,
+      "learning_rate": 9.326397492153762e-06,
+      "loss": 0.2268,
+      "step": 2250
+    },
+    {
+      "epoch": 0.7517114710302221,
+      "grad_norm": 0.8073887593343466,
+      "learning_rate": 9.325422992371117e-06,
+      "loss": 0.2399,
+      "step": 2251
+    },
+    {
+      "epoch": 0.7520454165970947,
+      "grad_norm": 0.5105733251703505,
+      "learning_rate": 9.324447839182397e-06,
+      "loss": 0.2248,
+      "step": 2252
+    },
+    {
+      "epoch": 0.7523793621639673,
+      "grad_norm": 0.5480491125068885,
+      "learning_rate": 9.323472032734915e-06,
+      "loss": 0.2464,
+      "step": 2253
+    },
+    {
+      "epoch": 0.7527133077308399,
+      "grad_norm": 0.48116501479820156,
+      "learning_rate": 9.322495573176073e-06,
+      "loss": 0.226,
+      "step": 2254
+    },
+    {
+      "epoch": 0.7530472532977125,
+      "grad_norm": 0.4210936718940355,
+      "learning_rate": 9.321518460653381e-06,
+      "loss": 0.2125,
+      "step": 2255
+    },
+    {
+      "epoch": 0.7533811988645851,
+      "grad_norm": 0.48326083685217036,
+      "learning_rate": 9.32054069531444e-06,
+      "loss": 0.2262,
+      "step": 2256
+    },
+    {
+      "epoch": 0.7537151444314577,
+      "grad_norm": 0.9265740345041341,
+      "learning_rate": 9.319562277306955e-06,
+      "loss": 0.2241,
+      "step": 2257
+    },
+    {
+      "epoch": 0.7540490899983303,
+      "grad_norm": 0.45048145099511155,
+      "learning_rate": 9.318583206778726e-06,
+      "loss": 0.216,
+      "step": 2258
+    },
+    {
+      "epoch": 0.7543830355652029,
+      "grad_norm": 0.578061264472536,
+      "learning_rate": 9.317603483877654e-06,
+      "loss": 0.2202,
+      "step": 2259
+    },
+    {
+      "epoch": 0.7547169811320755,
+      "grad_norm": 0.4307754856617888,
+      "learning_rate": 9.316623108751739e-06,
+      "loss": 0.2132,
+      "step": 2260
+    },
+    {
+      "epoch": 0.755050926698948,
+      "grad_norm": 0.486997614653093,
+      "learning_rate": 9.315642081549074e-06,
+      "loss": 0.2318,
+      "step": 2261
+    },
+    {
+      "epoch": 0.7553848722658206,
+      "grad_norm": 0.5251137171531048,
+      "learning_rate": 9.31466040241786e-06,
+      "loss": 0.2263,
+      "step": 2262
+    },
+    {
+      "epoch": 0.7557188178326932,
+      "grad_norm": 0.4664038869726607,
+      "learning_rate": 9.313678071506388e-06,
+      "loss": 0.2252,
+      "step": 2263
+    },
+    {
+      "epoch": 0.7560527633995658,
+      "grad_norm": 0.4349295324939173,
+      "learning_rate": 9.31269508896305e-06,
+      "loss": 0.2191,
+      "step": 2264
+    },
+    {
+      "epoch": 0.7563867089664384,
+      "grad_norm": 0.5160541914139507,
+      "learning_rate": 9.31171145493634e-06,
+      "loss": 0.2304,
+      "step": 2265
+    },
+    {
+      "epoch": 0.756720654533311,
+      "grad_norm": 0.48044683366908697,
+      "learning_rate": 9.310727169574847e-06,
+      "loss": 0.2313,
+      "step": 2266
+    },
+    {
+      "epoch": 0.7570546001001837,
+      "grad_norm": 0.44340511774161134,
+      "learning_rate": 9.309742233027258e-06,
+      "loss": 0.2178,
+      "step": 2267
+    },
+    {
+      "epoch": 0.7573885456670563,
+      "grad_norm": 0.4401405973485924,
+      "learning_rate": 9.308756645442356e-06,
+      "loss": 0.225,
+      "step": 2268
+    },
+    {
+      "epoch": 0.7577224912339289,
+      "grad_norm": 0.47966482086980605,
+      "learning_rate": 9.307770406969032e-06,
+      "loss": 0.23,
+      "step": 2269
+    },
+    {
+      "epoch": 0.7580564368008015,
+      "grad_norm": 0.43564783263314855,
+      "learning_rate": 9.306783517756264e-06,
+      "loss": 0.2131,
+      "step": 2270
+    },
+    {
+      "epoch": 0.7583903823676741,
+      "grad_norm": 0.45364809207582024,
+      "learning_rate": 9.305795977953134e-06,
+      "loss": 0.2276,
+      "step": 2271
+    },
+    {
+      "epoch": 0.7587243279345467,
+      "grad_norm": 0.5198386100281404,
+      "learning_rate": 9.304807787708825e-06,
+      "loss": 0.2259,
+      "step": 2272
+    },
+    {
+      "epoch": 0.7590582735014193,
+      "grad_norm": 0.46342370795905347,
+      "learning_rate": 9.303818947172611e-06,
+      "loss": 0.2179,
+      "step": 2273
+    },
+    {
+      "epoch": 0.7593922190682919,
+      "grad_norm": 0.4731835740602336,
+      "learning_rate": 9.302829456493868e-06,
+      "loss": 0.2333,
+      "step": 2274
+    },
+    {
+      "epoch": 0.7597261646351645,
+      "grad_norm": 0.4216435732249339,
+      "learning_rate": 9.301839315822072e-06,
+      "loss": 0.2196,
+      "step": 2275
+    },
+    {
+      "epoch": 0.7600601102020371,
+      "grad_norm": 0.4499998322284951,
+      "learning_rate": 9.300848525306797e-06,
+      "loss": 0.2208,
+      "step": 2276
+    },
+    {
+      "epoch": 0.7603940557689096,
+      "grad_norm": 0.4504516858457847,
+      "learning_rate": 9.299857085097708e-06,
+      "loss": 0.2258,
+      "step": 2277
+    },
+    {
+      "epoch": 0.7607280013357822,
+      "grad_norm": 0.5090935883272614,
+      "learning_rate": 9.298864995344579e-06,
+      "loss": 0.2151,
+      "step": 2278
+    },
+    {
+      "epoch": 0.7610619469026548,
+      "grad_norm": 0.485130607846563,
+      "learning_rate": 9.297872256197276e-06,
+      "loss": 0.231,
+      "step": 2279
+    },
+    {
+      "epoch": 0.7613958924695274,
+      "grad_norm": 0.3951173987462743,
+      "learning_rate": 9.296878867805762e-06,
+      "loss": 0.2119,
+      "step": 2280
+    },
+    {
+      "epoch": 0.7617298380364,
+      "grad_norm": 0.42718739210129064,
+      "learning_rate": 9.2958848303201e-06,
+      "loss": 0.2259,
+      "step": 2281
+    },
+    {
+      "epoch": 0.7620637836032726,
+      "grad_norm": 0.40877281318732694,
+      "learning_rate": 9.294890143890451e-06,
+      "loss": 0.217,
+      "step": 2282
+    },
+    {
+      "epoch": 0.7623977291701453,
+      "grad_norm": 0.44623672624981436,
+      "learning_rate": 9.293894808667077e-06,
+      "loss": 0.2216,
+      "step": 2283
+    },
+    {
+      "epoch": 0.7627316747370179,
+      "grad_norm": 0.5236867845873939,
+      "learning_rate": 9.292898824800333e-06,
+      "loss": 0.2561,
+      "step": 2284
+    },
+    {
+      "epoch": 0.7630656203038905,
+      "grad_norm": 0.46275656758815503,
+      "learning_rate": 9.291902192440673e-06,
+      "loss": 0.2191,
+      "step": 2285
+    },
+    {
+      "epoch": 0.7633995658707631,
+      "grad_norm": 0.44986183401594326,
+      "learning_rate": 9.290904911738653e-06,
+      "loss": 0.2196,
+      "step": 2286
+    },
+    {
+      "epoch": 0.7637335114376357,
+      "grad_norm": 0.47312386838896714,
+      "learning_rate": 9.289906982844923e-06,
+      "loss": 0.2311,
+      "step": 2287
+    },
+    {
+      "epoch": 0.7640674570045083,
+      "grad_norm": 0.4955825415082542,
+      "learning_rate": 9.288908405910228e-06,
+      "loss": 0.2187,
+      "step": 2288
+    },
+    {
+      "epoch": 0.7644014025713809,
+      "grad_norm": 0.4529969289979694,
+      "learning_rate": 9.287909181085421e-06,
+      "loss": 0.2211,
+      "step": 2289
+    },
+    {
+      "epoch": 0.7647353481382535,
+      "grad_norm": 0.4576506572434161,
+      "learning_rate": 9.286909308521443e-06,
+      "loss": 0.2095,
+      "step": 2290
+    },
+    {
+      "epoch": 0.7650692937051261,
+      "grad_norm": 0.4508998678478078,
+      "learning_rate": 9.285908788369336e-06,
+      "loss": 0.2287,
+      "step": 2291
+    },
+    {
+      "epoch": 0.7654032392719987,
+      "grad_norm": 0.45715461051048567,
+      "learning_rate": 9.284907620780244e-06,
+      "loss": 0.2042,
+      "step": 2292
+    },
+    {
+      "epoch": 0.7657371848388712,
+      "grad_norm": 0.4180093603700725,
+      "learning_rate": 9.2839058059054e-06,
+      "loss": 0.2125,
+      "step": 2293
+    },
+    {
+      "epoch": 0.7660711304057438,
+      "grad_norm": 0.43093767255958393,
+      "learning_rate": 9.282903343896144e-06,
+      "loss": 0.2238,
+      "step": 2294
+    },
+    {
+      "epoch": 0.7664050759726164,
+      "grad_norm": 0.5483814770622409,
+      "learning_rate": 9.281900234903908e-06,
+      "loss": 0.2316,
+      "step": 2295
+    },
+    {
+      "epoch": 0.766739021539489,
+      "grad_norm": 0.43771157015292517,
+      "learning_rate": 9.280896479080224e-06,
+      "loss": 0.2138,
+      "step": 2296
+    },
+    {
+      "epoch": 0.7670729671063616,
+      "grad_norm": 0.4574659911330641,
+      "learning_rate": 9.27989207657672e-06,
+      "loss": 0.226,
+      "step": 2297
+    },
+    {
+      "epoch": 0.7674069126732342,
+      "grad_norm": 0.42221388094567286,
+      "learning_rate": 9.278887027545125e-06,
+      "loss": 0.2209,
+      "step": 2298
+    },
+    {
+      "epoch": 0.7677408582401068,
+      "grad_norm": 0.5150950388506087,
+      "learning_rate": 9.277881332137261e-06,
+      "loss": 0.2349,
+      "step": 2299
+    },
+    {
+      "epoch": 0.7680748038069795,
+      "grad_norm": 0.48245297149523503,
+      "learning_rate": 9.276874990505053e-06,
+      "loss": 0.23,
+      "step": 2300
+    },
+    {
+      "epoch": 0.7684087493738521,
+      "grad_norm": 0.548683322704859,
+      "learning_rate": 9.27586800280052e-06,
+      "loss": 0.2135,
+      "step": 2301
+    },
+    {
+      "epoch": 0.7687426949407247,
+      "grad_norm": 0.4994150747196339,
+      "learning_rate": 9.274860369175775e-06,
+      "loss": 0.2303,
+      "step": 2302
+    },
+    {
+      "epoch": 0.7690766405075973,
+      "grad_norm": 0.5378040458428275,
+      "learning_rate": 9.27385208978304e-06,
+      "loss": 0.2461,
+      "step": 2303
+    },
+    {
+      "epoch": 0.7694105860744699,
+      "grad_norm": 0.45033554985182755,
+      "learning_rate": 9.272843164774622e-06,
+      "loss": 0.2131,
+      "step": 2304
+    },
+    {
+      "epoch": 0.7697445316413425,
+      "grad_norm": 0.46757095199070065,
+      "learning_rate": 9.27183359430293e-06,
+      "loss": 0.2232,
+      "step": 2305
+    },
+    {
+      "epoch": 0.7700784772082151,
+      "grad_norm": 0.39788086669142664,
+      "learning_rate": 9.270823378520478e-06,
+      "loss": 0.1993,
+      "step": 2306
+    },
+    {
+      "epoch": 0.7704124227750877,
+      "grad_norm": 0.5432380637027175,
+      "learning_rate": 9.269812517579867e-06,
+      "loss": 0.2409,
+      "step": 2307
+    },
+    {
+      "epoch": 0.7707463683419603,
+      "grad_norm": 0.5044905567763108,
+      "learning_rate": 9.268801011633799e-06,
+      "loss": 0.2263,
+      "step": 2308
+    },
+    {
+      "epoch": 0.7710803139088329,
+      "grad_norm": 0.4430694585924386,
+      "learning_rate": 9.267788860835076e-06,
+      "loss": 0.214,
+      "step": 2309
+    },
+    {
+      "epoch": 0.7714142594757054,
+      "grad_norm": 0.5203303582541469,
+      "learning_rate": 9.266776065336593e-06,
+      "loss": 0.246,
+      "step": 2310
+    },
+    {
+      "epoch": 0.771748205042578,
+      "grad_norm": 0.42657503652675616,
+      "learning_rate": 9.265762625291346e-06,
+      "loss": 0.2232,
+      "step": 2311
+    },
+    {
+      "epoch": 0.7720821506094506,
+      "grad_norm": 0.4594336261365237,
+      "learning_rate": 9.264748540852427e-06,
+      "loss": 0.2359,
+      "step": 2312
+    },
+    {
+      "epoch": 0.7724160961763232,
+      "grad_norm": 0.46882401712288807,
+      "learning_rate": 9.263733812173023e-06,
+      "loss": 0.2083,
+      "step": 2313
+    },
+    {
+      "epoch": 0.7727500417431958,
+      "grad_norm": 0.5028204129392393,
+      "learning_rate": 9.262718439406425e-06,
+      "loss": 0.2256,
+      "step": 2314
+    },
+    {
+      "epoch": 0.7730839873100684,
+      "grad_norm": 0.49213998524386604,
+      "learning_rate": 9.261702422706014e-06,
+      "loss": 0.2436,
+      "step": 2315
+    },
+    {
+      "epoch": 0.773417932876941,
+      "grad_norm": 0.43231620796153136,
+      "learning_rate": 9.260685762225273e-06,
+      "loss": 0.233,
+      "step": 2316
+    },
+    {
+      "epoch": 0.7737518784438137,
+      "grad_norm": 0.4434494887987508,
+      "learning_rate": 9.25966845811778e-06,
+      "loss": 0.2301,
+      "step": 2317
+    },
+    {
+      "epoch": 0.7740858240106863,
+      "grad_norm": 0.4370287988017438,
+      "learning_rate": 9.258650510537208e-06,
+      "loss": 0.1969,
+      "step": 2318
+    },
+    {
+      "epoch": 0.7744197695775589,
+      "grad_norm": 0.47263084992349913,
+      "learning_rate": 9.257631919637333e-06,
+      "loss": 0.2157,
+      "step": 2319
+    },
+    {
+      "epoch": 0.7747537151444315,
+      "grad_norm": 0.48339688653267926,
+      "learning_rate": 9.256612685572027e-06,
+      "loss": 0.2201,
+      "step": 2320
+    },
+    {
+      "epoch": 0.7750876607113041,
+      "grad_norm": 0.5110737778673129,
+      "learning_rate": 9.255592808495254e-06,
+      "loss": 0.2405,
+      "step": 2321
+    },
+    {
+      "epoch": 0.7754216062781767,
+      "grad_norm": 0.45441898821062743,
+      "learning_rate": 9.254572288561077e-06,
+      "loss": 0.2124,
+      "step": 2322
+    },
+    {
+      "epoch": 0.7757555518450493,
+      "grad_norm": 0.4588311491582138,
+      "learning_rate": 9.253551125923662e-06,
+      "loss": 0.2201,
+      "step": 2323
+    },
+    {
+      "epoch": 0.7760894974119219,
+      "grad_norm": 0.5177735740891841,
+      "learning_rate": 9.252529320737265e-06,
+      "loss": 0.2429,
+      "step": 2324
+    },
+    {
+      "epoch": 0.7764234429787945,
+      "grad_norm": 0.5022245852427304,
+      "learning_rate": 9.251506873156242e-06,
+      "loss": 0.2188,
+      "step": 2325
+    },
+    {
+      "epoch": 0.776757388545667,
+      "grad_norm": 0.4537744070437574,
+      "learning_rate": 9.250483783335046e-06,
+      "loss": 0.2106,
+      "step": 2326
+    },
+    {
+      "epoch": 0.7770913341125396,
+      "grad_norm": 0.6192354840406287,
+      "learning_rate": 9.249460051428226e-06,
+      "loss": 0.222,
+      "step": 2327
+    },
+    {
+      "epoch": 0.7774252796794122,
+      "grad_norm": 0.44129467986532117,
+      "learning_rate": 9.24843567759043e-06,
+      "loss": 0.206,
+      "step": 2328
+    },
+    {
+      "epoch": 0.7777592252462848,
+      "grad_norm": 0.4770974165507659,
+      "learning_rate": 9.247410661976402e-06,
+      "loss": 0.2309,
+      "step": 2329
+    },
+    {
+      "epoch": 0.7780931708131574,
+      "grad_norm": 0.48716858018195214,
+      "learning_rate": 9.246385004740981e-06,
+      "loss": 0.2411,
+      "step": 2330
+    },
+    {
+      "epoch": 0.77842711638003,
+      "grad_norm": 0.5243113960671338,
+      "learning_rate": 9.245358706039105e-06,
+      "loss": 0.234,
+      "step": 2331
+    },
+    {
+      "epoch": 0.7787610619469026,
+      "grad_norm": 0.41801045846912316,
+      "learning_rate": 9.244331766025812e-06,
+      "loss": 0.2239,
+      "step": 2332
+    },
+    {
+      "epoch": 0.7790950075137753,
+      "grad_norm": 0.46394404487526614,
+      "learning_rate": 9.243304184856226e-06,
+      "loss": 0.2148,
+      "step": 2333
+    },
+    {
+      "epoch": 0.7794289530806479,
+      "grad_norm": 0.4277415195601737,
+      "learning_rate": 9.242275962685584e-06,
+      "loss": 0.212,
+      "step": 2334
+    },
+    {
+      "epoch": 0.7797628986475205,
+      "grad_norm": 0.5185901645423077,
+      "learning_rate": 9.241247099669202e-06,
+      "loss": 0.245,
+      "step": 2335
+    },
+    {
+      "epoch": 0.7800968442143931,
+      "grad_norm": 0.44824621316470326,
+      "learning_rate": 9.24021759596251e-06,
+      "loss": 0.2255,
+      "step": 2336
+    },
+    {
+      "epoch": 0.7804307897812657,
+      "grad_norm": 0.4812361814852991,
+      "learning_rate": 9.239187451721021e-06,
+      "loss": 0.2352,
+      "step": 2337
+    },
+    {
+      "epoch": 0.7807647353481383,
+      "grad_norm": 0.4830179394244599,
+      "learning_rate": 9.238156667100354e-06,
+      "loss": 0.2291,
+      "step": 2338
+    },
+    {
+      "epoch": 0.7810986809150109,
+      "grad_norm": 0.44869641118909687,
+      "learning_rate": 9.237125242256219e-06,
+      "loss": 0.2121,
+      "step": 2339
+    },
+    {
+      "epoch": 0.7814326264818835,
+      "grad_norm": 0.4183573673022089,
+      "learning_rate": 9.236093177344427e-06,
+      "loss": 0.2165,
+      "step": 2340
+    },
+    {
+      "epoch": 0.7817665720487561,
+      "grad_norm": 0.44158451252351555,
+      "learning_rate": 9.23506047252088e-06,
+      "loss": 0.2193,
+      "step": 2341
+    },
+    {
+      "epoch": 0.7821005176156286,
+      "grad_norm": 0.4988366633490991,
+      "learning_rate": 9.234027127941585e-06,
+      "loss": 0.23,
+      "step": 2342
+    },
+    {
+      "epoch": 0.7824344631825012,
+      "grad_norm": 0.47542697794018896,
+      "learning_rate": 9.232993143762637e-06,
+      "loss": 0.2205,
+      "step": 2343
+    },
+    {
+      "epoch": 0.7827684087493738,
+      "grad_norm": 0.4307912169479183,
+      "learning_rate": 9.231958520140232e-06,
+      "loss": 0.2193,
+      "step": 2344
+    },
+    {
+      "epoch": 0.7831023543162464,
+      "grad_norm": 0.4686420725641428,
+      "learning_rate": 9.230923257230663e-06,
+      "loss": 0.2207,
+      "step": 2345
+    },
+    {
+      "epoch": 0.783436299883119,
+      "grad_norm": 0.7093627249782076,
+      "learning_rate": 9.22988735519032e-06,
+      "loss": 0.211,
+      "step": 2346
+    },
+    {
+      "epoch": 0.7837702454499916,
+      "grad_norm": 0.4449712797327876,
+      "learning_rate": 9.228850814175684e-06,
+      "loss": 0.2203,
+      "step": 2347
+    },
+    {
+      "epoch": 0.7841041910168642,
+      "grad_norm": 0.4224110582979993,
+      "learning_rate": 9.22781363434334e-06,
+      "loss": 0.2164,
+      "step": 2348
+    },
+    {
+      "epoch": 0.7844381365837368,
+      "grad_norm": 0.4453644666775102,
+      "learning_rate": 9.226775815849969e-06,
+      "loss": 0.2184,
+      "step": 2349
+    },
+    {
+      "epoch": 0.7847720821506095,
+      "grad_norm": 0.47882639139667954,
+      "learning_rate": 9.225737358852339e-06,
+      "loss": 0.2259,
+      "step": 2350
+    },
+    {
+      "epoch": 0.7851060277174821,
+      "grad_norm": 0.4614230944404661,
+      "learning_rate": 9.224698263507326e-06,
+      "loss": 0.2437,
+      "step": 2351
+    },
+    {
+      "epoch": 0.7854399732843547,
+      "grad_norm": 0.46240361961793186,
+      "learning_rate": 9.223658529971896e-06,
+      "loss": 0.2159,
+      "step": 2352
+    },
+    {
+      "epoch": 0.7857739188512273,
+      "grad_norm": 0.4326362693521754,
+      "learning_rate": 9.222618158403111e-06,
+      "loss": 0.2111,
+      "step": 2353
+    },
+    {
+      "epoch": 0.7861078644180999,
+      "grad_norm": 0.4294245191748408,
+      "learning_rate": 9.221577148958137e-06,
+      "loss": 0.2179,
+      "step": 2354
+    },
+    {
+      "epoch": 0.7864418099849725,
+      "grad_norm": 0.47430465166835356,
+      "learning_rate": 9.220535501794224e-06,
+      "loss": 0.2148,
+      "step": 2355
+    },
+    {
+      "epoch": 0.7867757555518451,
+      "grad_norm": 0.40903100342304954,
+      "learning_rate": 9.21949321706873e-06,
+      "loss": 0.2111,
+      "step": 2356
+    },
+    {
+      "epoch": 0.7871097011187177,
+      "grad_norm": 0.4871994961911349,
+      "learning_rate": 9.218450294939103e-06,
+      "loss": 0.2329,
+      "step": 2357
+    },
+    {
+      "epoch": 0.7874436466855903,
+      "grad_norm": 0.4802859353134735,
+      "learning_rate": 9.217406735562887e-06,
+      "loss": 0.2148,
+      "step": 2358
+    },
+    {
+      "epoch": 0.7877775922524628,
+      "grad_norm": 0.4829538927703034,
+      "learning_rate": 9.216362539097726e-06,
+      "loss": 0.2109,
+      "step": 2359
+    },
+    {
+      "epoch": 0.7881115378193354,
+      "grad_norm": 0.46957086759196354,
+      "learning_rate": 9.215317705701356e-06,
+      "loss": 0.2158,
+      "step": 2360
+    },
+    {
+      "epoch": 0.788445483386208,
+      "grad_norm": 0.4415507689008121,
+      "learning_rate": 9.214272235531615e-06,
+      "loss": 0.22,
+      "step": 2361
+    },
+    {
+      "epoch": 0.7887794289530806,
+      "grad_norm": 0.4784953125828448,
+      "learning_rate": 9.213226128746431e-06,
+      "loss": 0.2112,
+      "step": 2362
+    },
+    {
+      "epoch": 0.7891133745199532,
+      "grad_norm": 0.4748308372416173,
+      "learning_rate": 9.21217938550383e-06,
+      "loss": 0.2267,
+      "step": 2363
+    },
+    {
+      "epoch": 0.7894473200868258,
+      "grad_norm": 0.42786072729815783,
+      "learning_rate": 9.211132005961936e-06,
+      "loss": 0.2182,
+      "step": 2364
+    },
+    {
+      "epoch": 0.7897812656536984,
+      "grad_norm": 0.46379390215134986,
+      "learning_rate": 9.210083990278968e-06,
+      "loss": 0.2159,
+      "step": 2365
+    },
+    {
+      "epoch": 0.790115211220571,
+      "grad_norm": 0.434672422878081,
+      "learning_rate": 9.209035338613242e-06,
+      "loss": 0.2176,
+      "step": 2366
+    },
+    {
+      "epoch": 0.7904491567874437,
+      "grad_norm": 0.4712397323015266,
+      "learning_rate": 9.207986051123167e-06,
+      "loss": 0.2169,
+      "step": 2367
+    },
+    {
+      "epoch": 0.7907831023543163,
+      "grad_norm": 0.4643471410872069,
+      "learning_rate": 9.206936127967254e-06,
+      "loss": 0.2231,
+      "step": 2368
+    },
+    {
+      "epoch": 0.7911170479211889,
+      "grad_norm": 0.4832625761947329,
+      "learning_rate": 9.205885569304103e-06,
+      "loss": 0.2346,
+      "step": 2369
+    },
+    {
+      "epoch": 0.7914509934880615,
+      "grad_norm": 0.5120653341402449,
+      "learning_rate": 9.204834375292413e-06,
+      "loss": 0.2248,
+      "step": 2370
+    },
+    {
+      "epoch": 0.7917849390549341,
+      "grad_norm": 0.4595895504768996,
+      "learning_rate": 9.20378254609098e-06,
+      "loss": 0.212,
+      "step": 2371
+    },
+    {
+      "epoch": 0.7921188846218067,
+      "grad_norm": 0.4464850753967196,
+      "learning_rate": 9.202730081858697e-06,
+      "loss": 0.2216,
+      "step": 2372
+    },
+    {
+      "epoch": 0.7924528301886793,
+      "grad_norm": 0.4465507871460174,
+      "learning_rate": 9.201676982754549e-06,
+      "loss": 0.2175,
+      "step": 2373
+    },
+    {
+      "epoch": 0.7927867757555519,
+      "grad_norm": 0.7186089123657715,
+      "learning_rate": 9.200623248937619e-06,
+      "loss": 0.2217,
+      "step": 2374
+    },
+    {
+      "epoch": 0.7931207213224244,
+      "grad_norm": 0.48065369107795447,
+      "learning_rate": 9.199568880567085e-06,
+      "loss": 0.2346,
+      "step": 2375
+    },
+    {
+      "epoch": 0.793454666889297,
+      "grad_norm": 0.46022889936302425,
+      "learning_rate": 9.198513877802226e-06,
+      "loss": 0.2185,
+      "step": 2376
+    },
+    {
+      "epoch": 0.7937886124561696,
+      "grad_norm": 0.48578689235218536,
+      "learning_rate": 9.19745824080241e-06,
+      "loss": 0.2325,
+      "step": 2377
+    },
+    {
+      "epoch": 0.7941225580230422,
+      "grad_norm": 0.4201772420752296,
+      "learning_rate": 9.196401969727101e-06,
+      "loss": 0.2219,
+      "step": 2378
+    },
+    {
+      "epoch": 0.7944565035899148,
+      "grad_norm": 0.5148645151535987,
+      "learning_rate": 9.195345064735865e-06,
+      "loss": 0.2187,
+      "step": 2379
+    },
+    {
+      "epoch": 0.7947904491567874,
+      "grad_norm": 0.4793318370535307,
+      "learning_rate": 9.194287525988358e-06,
+      "loss": 0.2206,
+      "step": 2380
+    },
+    {
+      "epoch": 0.79512439472366,
+      "grad_norm": 0.44812143862309217,
+      "learning_rate": 9.193229353644336e-06,
+      "loss": 0.2066,
+      "step": 2381
+    },
+    {
+      "epoch": 0.7954583402905326,
+      "grad_norm": 0.48309758118279134,
+      "learning_rate": 9.192170547863644e-06,
+      "loss": 0.2272,
+      "step": 2382
+    },
+    {
+      "epoch": 0.7957922858574052,
+      "grad_norm": 0.4955575076751468,
+      "learning_rate": 9.191111108806228e-06,
+      "loss": 0.2152,
+      "step": 2383
+    },
+    {
+      "epoch": 0.7961262314242779,
+      "grad_norm": 0.5483835825302927,
+      "learning_rate": 9.190051036632133e-06,
+      "loss": 0.2155,
+      "step": 2384
+    },
+    {
+      "epoch": 0.7964601769911505,
+      "grad_norm": 0.43256677860584014,
+      "learning_rate": 9.188990331501493e-06,
+      "loss": 0.2172,
+      "step": 2385
+    },
+    {
+      "epoch": 0.7967941225580231,
+      "grad_norm": 0.47777695666620773,
+      "learning_rate": 9.187928993574537e-06,
+      "loss": 0.218,
+      "step": 2386
+    },
+    {
+      "epoch": 0.7971280681248957,
+      "grad_norm": 0.4857124920043452,
+      "learning_rate": 9.186867023011598e-06,
+      "loss": 0.2269,
+      "step": 2387
+    },
+    {
+      "epoch": 0.7974620136917683,
+      "grad_norm": 0.46097558597069427,
+      "learning_rate": 9.185804419973096e-06,
+      "loss": 0.2262,
+      "step": 2388
+    },
+    {
+      "epoch": 0.7977959592586409,
+      "grad_norm": 0.45451151055369127,
+      "learning_rate": 9.18474118461955e-06,
+      "loss": 0.2283,
+      "step": 2389
+    },
+    {
+      "epoch": 0.7981299048255135,
+      "grad_norm": 0.45002263688863814,
+      "learning_rate": 9.183677317111572e-06,
+      "loss": 0.2264,
+      "step": 2390
+    },
+    {
+      "epoch": 0.798463850392386,
+      "grad_norm": 0.4351929657301828,
+      "learning_rate": 9.182612817609877e-06,
+      "loss": 0.2162,
+      "step": 2391
+    },
+    {
+      "epoch": 0.7987977959592586,
+      "grad_norm": 0.4829125840864148,
+      "learning_rate": 9.181547686275266e-06,
+      "loss": 0.2209,
+      "step": 2392
+    },
+    {
+      "epoch": 0.7991317415261312,
+      "grad_norm": 0.4617414472574589,
+      "learning_rate": 9.180481923268641e-06,
+      "loss": 0.2317,
+      "step": 2393
+    },
+    {
+      "epoch": 0.7994656870930038,
+      "grad_norm": 0.4715677223409943,
+      "learning_rate": 9.179415528750998e-06,
+      "loss": 0.2324,
+      "step": 2394
+    },
+    {
+      "epoch": 0.7997996326598764,
+      "grad_norm": 0.445753677713308,
+      "learning_rate": 9.178348502883428e-06,
+      "loss": 0.2086,
+      "step": 2395
+    },
+    {
+      "epoch": 0.800133578226749,
+      "grad_norm": 0.4739163161049927,
+      "learning_rate": 9.17728084582712e-06,
+      "loss": 0.2299,
+      "step": 2396
+    },
+    {
+      "epoch": 0.8004675237936216,
+      "grad_norm": 0.4444338702698356,
+      "learning_rate": 9.176212557743352e-06,
+      "loss": 0.2179,
+      "step": 2397
+    },
+    {
+      "epoch": 0.8008014693604942,
+      "grad_norm": 0.5521009272187888,
+      "learning_rate": 9.175143638793504e-06,
+      "loss": 0.2332,
+      "step": 2398
+    },
+    {
+      "epoch": 0.8011354149273668,
+      "grad_norm": 0.4529565320102755,
+      "learning_rate": 9.174074089139048e-06,
+      "loss": 0.2188,
+      "step": 2399
+    },
+    {
+      "epoch": 0.8014693604942394,
+      "grad_norm": 0.4655241748444513,
+      "learning_rate": 9.173003908941555e-06,
+      "loss": 0.231,
+      "step": 2400
+    },
+    {
+      "epoch": 0.801803306061112,
+      "grad_norm": 0.46546430337015227,
+      "learning_rate": 9.171933098362685e-06,
+      "loss": 0.2245,
+      "step": 2401
+    },
+    {
+      "epoch": 0.8021372516279847,
+      "grad_norm": 0.4291232236423889,
+      "learning_rate": 9.170861657564197e-06,
+      "loss": 0.1983,
+      "step": 2402
+    },
+    {
+      "epoch": 0.8024711971948573,
+      "grad_norm": 0.47959703405449,
+      "learning_rate": 9.169789586707947e-06,
+      "loss": 0.2225,
+      "step": 2403
+    },
+    {
+      "epoch": 0.8028051427617299,
+      "grad_norm": 0.4705338430027695,
+      "learning_rate": 9.16871688595588e-06,
+      "loss": 0.2174,
+      "step": 2404
+    },
+    {
+      "epoch": 0.8031390883286025,
+      "grad_norm": 0.4542925186202159,
+      "learning_rate": 9.167643555470044e-06,
+      "loss": 0.221,
+      "step": 2405
+    },
+    {
+      "epoch": 0.8034730338954751,
+      "grad_norm": 0.5053346494758011,
+      "learning_rate": 9.166569595412576e-06,
+      "loss": 0.2214,
+      "step": 2406
+    },
+    {
+      "epoch": 0.8038069794623477,
+      "grad_norm": 0.518802140242786,
+      "learning_rate": 9.16549500594571e-06,
+      "loss": 0.203,
+      "step": 2407
+    },
+    {
+      "epoch": 0.8041409250292202,
+      "grad_norm": 0.4830630204730151,
+      "learning_rate": 9.164419787231778e-06,
+      "loss": 0.2334,
+      "step": 2408
+    },
+    {
+      "epoch": 0.8044748705960928,
+      "grad_norm": 0.5058947808268289,
+      "learning_rate": 9.163343939433202e-06,
+      "loss": 0.2303,
+      "step": 2409
+    },
+    {
+      "epoch": 0.8048088161629654,
+      "grad_norm": 0.5481350305947928,
+      "learning_rate": 9.162267462712502e-06,
+      "loss": 0.227,
+      "step": 2410
+    },
+    {
+      "epoch": 0.805142761729838,
+      "grad_norm": 0.40346017951021734,
+      "learning_rate": 9.161190357232292e-06,
+      "loss": 0.2078,
+      "step": 2411
+    },
+    {
+      "epoch": 0.8054767072967106,
+      "grad_norm": 0.4012932942269892,
+      "learning_rate": 9.160112623155282e-06,
+      "loss": 0.2099,
+      "step": 2412
+    },
+    {
+      "epoch": 0.8058106528635832,
+      "grad_norm": 0.41692385024259015,
+      "learning_rate": 9.159034260644277e-06,
+      "loss": 0.2084,
+      "step": 2413
+    },
+    {
+      "epoch": 0.8061445984304558,
+      "grad_norm": 0.5463744356480256,
+      "learning_rate": 9.157955269862176e-06,
+      "loss": 0.2286,
+      "step": 2414
+    },
+    {
+      "epoch": 0.8064785439973284,
+      "grad_norm": 0.49138767972709996,
+      "learning_rate": 9.156875650971974e-06,
+      "loss": 0.2223,
+      "step": 2415
+    },
+    {
+      "epoch": 0.806812489564201,
+      "grad_norm": 0.46139748034889444,
+      "learning_rate": 9.155795404136757e-06,
+      "loss": 0.2143,
+      "step": 2416
+    },
+    {
+      "epoch": 0.8071464351310736,
+      "grad_norm": 0.41387074512700345,
+      "learning_rate": 9.154714529519715e-06,
+      "loss": 0.2094,
+      "step": 2417
+    },
+    {
+      "epoch": 0.8074803806979463,
+      "grad_norm": 0.4711974425999777,
+      "learning_rate": 9.15363302728412e-06,
+      "loss": 0.2252,
+      "step": 2418
+    },
+    {
+      "epoch": 0.8078143262648189,
+      "grad_norm": 0.43803798636984687,
+      "learning_rate": 9.15255089759335e-06,
+      "loss": 0.2097,
+      "step": 2419
+    },
+    {
+      "epoch": 0.8081482718316915,
+      "grad_norm": 0.4315283690951102,
+      "learning_rate": 9.151468140610872e-06,
+      "loss": 0.2029,
+      "step": 2420
+    },
+    {
+      "epoch": 0.8084822173985641,
+      "grad_norm": 0.43314565060866766,
+      "learning_rate": 9.150384756500249e-06,
+      "loss": 0.2129,
+      "step": 2421
+    },
+    {
+      "epoch": 0.8088161629654367,
+      "grad_norm": 0.4992267726729122,
+      "learning_rate": 9.14930074542514e-06,
+      "loss": 0.2276,
+      "step": 2422
+    },
+    {
+      "epoch": 0.8091501085323093,
+      "grad_norm": 0.8734400531790079,
+      "learning_rate": 9.148216107549297e-06,
+      "loss": 0.2276,
+      "step": 2423
+    },
+    {
+      "epoch": 0.8094840540991818,
+      "grad_norm": 0.4618645547769956,
+      "learning_rate": 9.147130843036567e-06,
+      "loss": 0.2216,
+      "step": 2424
+    },
+    {
+      "epoch": 0.8098179996660544,
+      "grad_norm": 0.503229948469187,
+      "learning_rate": 9.146044952050891e-06,
+      "loss": 0.2266,
+      "step": 2425
+    },
+    {
+      "epoch": 0.810151945232927,
+      "grad_norm": 0.48234269217820075,
+      "learning_rate": 9.144958434756308e-06,
+      "loss": 0.246,
+      "step": 2426
+    },
+    {
+      "epoch": 0.8104858907997996,
+      "grad_norm": 0.45799158973946075,
+      "learning_rate": 9.14387129131695e-06,
+      "loss": 0.2241,
+      "step": 2427
+    },
+    {
+      "epoch": 0.8108198363666722,
+      "grad_norm": 0.47473856709077045,
+      "learning_rate": 9.142783521897038e-06,
+      "loss": 0.2315,
+      "step": 2428
+    },
+    {
+      "epoch": 0.8111537819335448,
+      "grad_norm": 0.46232253841106646,
+      "learning_rate": 9.141695126660896e-06,
+      "loss": 0.2146,
+      "step": 2429
+    },
+    {
+      "epoch": 0.8114877275004174,
+      "grad_norm": 0.45675763317544865,
+      "learning_rate": 9.14060610577294e-06,
+      "loss": 0.2196,
+      "step": 2430
+    },
+    {
+      "epoch": 0.81182167306729,
+      "grad_norm": 0.45024530888195646,
+      "learning_rate": 9.139516459397675e-06,
+      "loss": 0.2087,
+      "step": 2431
+    },
+    {
+      "epoch": 0.8121556186341626,
+      "grad_norm": 0.5389654482824952,
+      "learning_rate": 9.13842618769971e-06,
+      "loss": 0.231,
+      "step": 2432
+    },
+    {
+      "epoch": 0.8124895642010352,
+      "grad_norm": 0.48436355868941494,
+      "learning_rate": 9.13733529084374e-06,
+      "loss": 0.2434,
+      "step": 2433
+    },
+    {
+      "epoch": 0.8128235097679078,
+      "grad_norm": 0.46078224853375155,
+      "learning_rate": 9.13624376899456e-06,
+      "loss": 0.2338,
+      "step": 2434
+    },
+    {
+      "epoch": 0.8131574553347805,
+      "grad_norm": 0.41677648160432945,
+      "learning_rate": 9.135151622317054e-06,
+      "loss": 0.224,
+      "step": 2435
+    },
+    {
+      "epoch": 0.8134914009016531,
+      "grad_norm": 0.4755376042102644,
+      "learning_rate": 9.134058850976205e-06,
+      "loss": 0.2247,
+      "step": 2436
+    },
+    {
+      "epoch": 0.8138253464685257,
+      "grad_norm": 0.436828747543142,
+      "learning_rate": 9.132965455137092e-06,
+      "loss": 0.2056,
+      "step": 2437
+    },
+    {
+      "epoch": 0.8141592920353983,
+      "grad_norm": 0.5091572548534862,
+      "learning_rate": 9.13187143496488e-06,
+      "loss": 0.2249,
+      "step": 2438
+    },
+    {
+      "epoch": 0.8144932376022709,
+      "grad_norm": 0.4430402243350094,
+      "learning_rate": 9.13077679062484e-06,
+      "loss": 0.2222,
+      "step": 2439
+    },
+    {
+      "epoch": 0.8148271831691434,
+      "grad_norm": 0.47005559131312463,
+      "learning_rate": 9.129681522282326e-06,
+      "loss": 0.2224,
+      "step": 2440
+    },
+    {
+      "epoch": 0.815161128736016,
+      "grad_norm": 0.46240395621992303,
+      "learning_rate": 9.128585630102793e-06,
+      "loss": 0.218,
+      "step": 2441
+    },
+    {
+      "epoch": 0.8154950743028886,
+      "grad_norm": 0.4229967980030704,
+      "learning_rate": 9.127489114251787e-06,
+      "loss": 0.2088,
+      "step": 2442
+    },
+    {
+      "epoch": 0.8158290198697612,
+      "grad_norm": 0.4359587185147543,
+      "learning_rate": 9.12639197489495e-06,
+      "loss": 0.2226,
+      "step": 2443
+    },
+    {
+      "epoch": 0.8161629654366338,
+      "grad_norm": 0.47401382528724656,
+      "learning_rate": 9.125294212198022e-06,
+      "loss": 0.2334,
+      "step": 2444
+    },
+    {
+      "epoch": 0.8164969110035064,
+      "grad_norm": 0.4838938684515212,
+      "learning_rate": 9.124195826326827e-06,
+      "loss": 0.2328,
+      "step": 2445
+    },
+    {
+      "epoch": 0.816830856570379,
+      "grad_norm": 0.5430695760268881,
+      "learning_rate": 9.12309681744729e-06,
+      "loss": 0.2175,
+      "step": 2446
+    },
+    {
+      "epoch": 0.8171648021372516,
+      "grad_norm": 0.5297232531969129,
+      "learning_rate": 9.121997185725433e-06,
+      "loss": 0.2239,
+      "step": 2447
+    },
+    {
+      "epoch": 0.8174987477041242,
+      "grad_norm": 0.4606818496718444,
+      "learning_rate": 9.120896931327366e-06,
+      "loss": 0.2322,
+      "step": 2448
+    },
+    {
+      "epoch": 0.8178326932709968,
+      "grad_norm": 0.43326888876665187,
+      "learning_rate": 9.119796054419295e-06,
+      "loss": 0.2352,
+      "step": 2449
+    },
+    {
+      "epoch": 0.8181666388378694,
+      "grad_norm": 0.45191071938809313,
+      "learning_rate": 9.118694555167521e-06,
+      "loss": 0.2334,
+      "step": 2450
+    },
+    {
+      "epoch": 0.818500584404742,
+      "grad_norm": 0.4137197400642333,
+      "learning_rate": 9.117592433738439e-06,
+      "loss": 0.2209,
+      "step": 2451
+    },
+    {
+      "epoch": 0.8188345299716147,
+      "grad_norm": 0.4726434520191131,
+      "learning_rate": 9.116489690298536e-06,
+      "loss": 0.2307,
+      "step": 2452
+    },
+    {
+      "epoch": 0.8191684755384873,
+      "grad_norm": 0.45335955621754376,
+      "learning_rate": 9.115386325014396e-06,
+      "loss": 0.2149,
+      "step": 2453
+    },
+    {
+      "epoch": 0.8195024211053599,
+      "grad_norm": 0.5966832339899147,
+      "learning_rate": 9.114282338052695e-06,
+      "loss": 0.2494,
+      "step": 2454
+    },
+    {
+      "epoch": 0.8198363666722325,
+      "grad_norm": 0.507628757751821,
+      "learning_rate": 9.113177729580203e-06,
+      "loss": 0.224,
+      "step": 2455
+    },
+    {
+      "epoch": 0.820170312239105,
+      "grad_norm": 0.41348330095186164,
+      "learning_rate": 9.112072499763783e-06,
+      "loss": 0.2139,
+      "step": 2456
+    },
+    {
+      "epoch": 0.8205042578059776,
+      "grad_norm": 0.4445659097139337,
+      "learning_rate": 9.110966648770392e-06,
+      "loss": 0.2098,
+      "step": 2457
+    },
+    {
+      "epoch": 0.8208382033728502,
+      "grad_norm": 0.49639772274003185,
+      "learning_rate": 9.109860176767085e-06,
+      "loss": 0.2242,
+      "step": 2458
+    },
+    {
+      "epoch": 0.8211721489397228,
+      "grad_norm": 0.4692548380202328,
+      "learning_rate": 9.108753083921007e-06,
+      "loss": 0.2223,
+      "step": 2459
+    },
+    {
+      "epoch": 0.8215060945065954,
+      "grad_norm": 0.4857245701306937,
+      "learning_rate": 9.107645370399395e-06,
+      "loss": 0.2282,
+      "step": 2460
+    },
+    {
+      "epoch": 0.821840040073468,
+      "grad_norm": 0.4350617485594562,
+      "learning_rate": 9.106537036369587e-06,
+      "loss": 0.2057,
+      "step": 2461
+    },
+    {
+      "epoch": 0.8221739856403406,
+      "grad_norm": 0.45938249982968987,
+      "learning_rate": 9.105428081999004e-06,
+      "loss": 0.2297,
+      "step": 2462
+    },
+    {
+      "epoch": 0.8225079312072132,
+      "grad_norm": 0.4086313343960967,
+      "learning_rate": 9.10431850745517e-06,
+      "loss": 0.2036,
+      "step": 2463
+    },
+    {
+      "epoch": 0.8228418767740858,
+      "grad_norm": 0.44187113165377345,
+      "learning_rate": 9.103208312905698e-06,
+      "loss": 0.2149,
+      "step": 2464
+    },
+    {
+      "epoch": 0.8231758223409584,
+      "grad_norm": 0.4456715017463197,
+      "learning_rate": 9.102097498518299e-06,
+      "loss": 0.2157,
+      "step": 2465
+    },
+    {
+      "epoch": 0.823509767907831,
+      "grad_norm": 0.3795807225141392,
+      "learning_rate": 9.100986064460769e-06,
+      "loss": 0.1994,
+      "step": 2466
+    },
+    {
+      "epoch": 0.8238437134747036,
+      "grad_norm": 0.5015498235452791,
+      "learning_rate": 9.099874010901009e-06,
+      "loss": 0.2067,
+      "step": 2467
+    },
+    {
+      "epoch": 0.8241776590415762,
+      "grad_norm": 0.46873528314568835,
+      "learning_rate": 9.098761338007003e-06,
+      "loss": 0.2342,
+      "step": 2468
+    },
+    {
+      "epoch": 0.8245116046084489,
+      "grad_norm": 0.4802429384639126,
+      "learning_rate": 9.097648045946837e-06,
+      "loss": 0.2077,
+      "step": 2469
+    },
+    {
+      "epoch": 0.8248455501753215,
+      "grad_norm": 0.47674620695731734,
+      "learning_rate": 9.096534134888685e-06,
+      "loss": 0.2311,
+      "step": 2470
+    },
+    {
+      "epoch": 0.8251794957421941,
+      "grad_norm": 0.4741718011929554,
+      "learning_rate": 9.095419605000817e-06,
+      "loss": 0.2223,
+      "step": 2471
+    },
+    {
+      "epoch": 0.8255134413090667,
+      "grad_norm": 0.4614907153038838,
+      "learning_rate": 9.094304456451596e-06,
+      "loss": 0.2301,
+      "step": 2472
+    },
+    {
+      "epoch": 0.8258473868759392,
+      "grad_norm": 0.4517767126346064,
+      "learning_rate": 9.093188689409477e-06,
+      "loss": 0.2222,
+      "step": 2473
+    },
+    {
+      "epoch": 0.8261813324428118,
+      "grad_norm": 0.652118941163761,
+      "learning_rate": 9.09207230404301e-06,
+      "loss": 0.2331,
+      "step": 2474
+    },
+    {
+      "epoch": 0.8265152780096844,
+      "grad_norm": 0.45586670064349893,
+      "learning_rate": 9.090955300520842e-06,
+      "loss": 0.2111,
+      "step": 2475
+    },
+    {
+      "epoch": 0.826849223576557,
+      "grad_norm": 0.4899208936605748,
+      "learning_rate": 9.089837679011704e-06,
+      "loss": 0.2282,
+      "step": 2476
+    },
+    {
+      "epoch": 0.8271831691434296,
+      "grad_norm": 0.4964827628532168,
+      "learning_rate": 9.08871943968443e-06,
+      "loss": 0.2269,
+      "step": 2477
+    },
+    {
+      "epoch": 0.8275171147103022,
+      "grad_norm": 0.4552900965289199,
+      "learning_rate": 9.08760058270794e-06,
+      "loss": 0.2356,
+      "step": 2478
+    },
+    {
+      "epoch": 0.8278510602771748,
+      "grad_norm": 0.4918378116260361,
+      "learning_rate": 9.086481108251253e-06,
+      "loss": 0.237,
+      "step": 2479
+    },
+    {
+      "epoch": 0.8281850058440474,
+      "grad_norm": 0.4607573617794218,
+      "learning_rate": 9.085361016483477e-06,
+      "loss": 0.2197,
+      "step": 2480
+    },
+    {
+      "epoch": 0.82851895141092,
+      "grad_norm": 0.5491934418016965,
+      "learning_rate": 9.084240307573816e-06,
+      "loss": 0.2145,
+      "step": 2481
+    },
+    {
+      "epoch": 0.8288528969777926,
+      "grad_norm": 0.48107467425045347,
+      "learning_rate": 9.083118981691567e-06,
+      "loss": 0.2198,
+      "step": 2482
+    },
+    {
+      "epoch": 0.8291868425446652,
+      "grad_norm": 0.48450480401984153,
+      "learning_rate": 9.081997039006117e-06,
+      "loss": 0.2226,
+      "step": 2483
+    },
+    {
+      "epoch": 0.8295207881115378,
+      "grad_norm": 0.4454234793986314,
+      "learning_rate": 9.080874479686952e-06,
+      "loss": 0.2091,
+      "step": 2484
+    },
+    {
+      "epoch": 0.8298547336784105,
+      "grad_norm": 0.5397372637338255,
+      "learning_rate": 9.079751303903646e-06,
+      "loss": 0.2302,
+      "step": 2485
+    },
+    {
+      "epoch": 0.8301886792452831,
+      "grad_norm": 0.49058908175007737,
+      "learning_rate": 9.078627511825866e-06,
+      "loss": 0.2339,
+      "step": 2486
+    },
+    {
+      "epoch": 0.8305226248121557,
+      "grad_norm": 0.49048769393241404,
+      "learning_rate": 9.077503103623379e-06,
+      "loss": 0.2263,
+      "step": 2487
+    },
+    {
+      "epoch": 0.8308565703790283,
+      "grad_norm": 0.4704762203253689,
+      "learning_rate": 9.076378079466036e-06,
+      "loss": 0.2132,
+      "step": 2488
+    },
+    {
+      "epoch": 0.8311905159459008,
+      "grad_norm": 0.4149169942291882,
+      "learning_rate": 9.075252439523785e-06,
+      "loss": 0.2185,
+      "step": 2489
+    },
+    {
+      "epoch": 0.8315244615127734,
+      "grad_norm": 0.4750547835779951,
+      "learning_rate": 9.074126183966669e-06,
+      "loss": 0.2136,
+      "step": 2490
+    },
+    {
+      "epoch": 0.831858407079646,
+      "grad_norm": 0.4757858563906054,
+      "learning_rate": 9.072999312964823e-06,
+      "loss": 0.2256,
+      "step": 2491
+    },
+    {
+      "epoch": 0.8321923526465186,
+      "grad_norm": 0.4588296740580707,
+      "learning_rate": 9.071871826688472e-06,
+      "loss": 0.2182,
+      "step": 2492
+    },
+    {
+      "epoch": 0.8325262982133912,
+      "grad_norm": 0.4501152802206452,
+      "learning_rate": 9.070743725307937e-06,
+      "loss": 0.2126,
+      "step": 2493
+    },
+    {
+      "epoch": 0.8328602437802638,
+      "grad_norm": 0.5107092439533395,
+      "learning_rate": 9.06961500899363e-06,
+      "loss": 0.2385,
+      "step": 2494
+    },
+    {
+      "epoch": 0.8331941893471364,
+      "grad_norm": 0.46172068053365517,
+      "learning_rate": 9.068485677916059e-06,
+      "loss": 0.2233,
+      "step": 2495
+    },
+    {
+      "epoch": 0.833528134914009,
+      "grad_norm": 0.4559896293320187,
+      "learning_rate": 9.06735573224582e-06,
+      "loss": 0.2156,
+      "step": 2496
+    },
+    {
+      "epoch": 0.8338620804808816,
+      "grad_norm": 0.48841992918673477,
+      "learning_rate": 9.066225172153607e-06,
+      "loss": 0.2509,
+      "step": 2497
+    },
+    {
+      "epoch": 0.8341960260477542,
+      "grad_norm": 0.48807810736665075,
+      "learning_rate": 9.065093997810204e-06,
+      "loss": 0.2297,
+      "step": 2498
+    },
+    {
+      "epoch": 0.8345299716146268,
+      "grad_norm": 0.45944899131100364,
+      "learning_rate": 9.063962209386485e-06,
+      "loss": 0.234,
+      "step": 2499
+    },
+    {
+      "epoch": 0.8348639171814994,
+      "grad_norm": 0.4787228349593665,
+      "learning_rate": 9.062829807053426e-06,
+      "loss": 0.2188,
+      "step": 2500
+    },
+    {
+      "epoch": 0.835197862748372,
+      "grad_norm": 0.5036077547144578,
+      "learning_rate": 9.061696790982086e-06,
+      "loss": 0.2366,
+      "step": 2501
+    },
+    {
+      "epoch": 0.8355318083152447,
+      "grad_norm": 0.544975155082703,
+      "learning_rate": 9.060563161343618e-06,
+      "loss": 0.214,
+      "step": 2502
+    },
+    {
+      "epoch": 0.8358657538821173,
+      "grad_norm": 0.4434329818632235,
+      "learning_rate": 9.059428918309276e-06,
+      "loss": 0.2141,
+      "step": 2503
+    },
+    {
+      "epoch": 0.8361996994489899,
+      "grad_norm": 0.4498540488388689,
+      "learning_rate": 9.058294062050396e-06,
+      "loss": 0.2178,
+      "step": 2504
+    },
+    {
+      "epoch": 0.8365336450158624,
+      "grad_norm": 0.42897565416363814,
+      "learning_rate": 9.057158592738414e-06,
+      "loss": 0.2147,
+      "step": 2505
+    },
+    {
+      "epoch": 0.836867590582735,
+      "grad_norm": 0.42112860630504073,
+      "learning_rate": 9.056022510544855e-06,
+      "loss": 0.2132,
+      "step": 2506
+    },
+    {
+      "epoch": 0.8372015361496076,
+      "grad_norm": 0.4408094729636434,
+      "learning_rate": 9.054885815641336e-06,
+      "loss": 0.2215,
+      "step": 2507
+    },
+    {
+      "epoch": 0.8375354817164802,
+      "grad_norm": 0.45258060889972634,
+      "learning_rate": 9.05374850819957e-06,
+      "loss": 0.2132,
+      "step": 2508
+    },
+    {
+      "epoch": 0.8378694272833528,
+      "grad_norm": 0.47132515883640624,
+      "learning_rate": 9.052610588391363e-06,
+      "loss": 0.2233,
+      "step": 2509
+    },
+    {
+      "epoch": 0.8382033728502254,
+      "grad_norm": 0.4767783733828431,
+      "learning_rate": 9.051472056388606e-06,
+      "loss": 0.2418,
+      "step": 2510
+    },
+    {
+      "epoch": 0.838537318417098,
+      "grad_norm": 0.46137862596916196,
+      "learning_rate": 9.050332912363292e-06,
+      "loss": 0.237,
+      "step": 2511
+    },
+    {
+      "epoch": 0.8388712639839706,
+      "grad_norm": 0.49409067455142913,
+      "learning_rate": 9.049193156487501e-06,
+      "loss": 0.2046,
+      "step": 2512
+    },
+    {
+      "epoch": 0.8392052095508432,
+      "grad_norm": 0.49205788033890396,
+      "learning_rate": 9.048052788933405e-06,
+      "loss": 0.2206,
+      "step": 2513
+    },
+    {
+      "epoch": 0.8395391551177158,
+      "grad_norm": 0.4667335941199234,
+      "learning_rate": 9.046911809873271e-06,
+      "loss": 0.2158,
+      "step": 2514
+    },
+    {
+      "epoch": 0.8398731006845884,
+      "grad_norm": 0.4121890688466787,
+      "learning_rate": 9.045770219479457e-06,
+      "loss": 0.2021,
+      "step": 2515
+    },
+    {
+      "epoch": 0.840207046251461,
+      "grad_norm": 0.42556930558028944,
+      "learning_rate": 9.044628017924415e-06,
+      "loss": 0.2226,
+      "step": 2516
+    },
+    {
+      "epoch": 0.8405409918183336,
+      "grad_norm": 0.5288410131785884,
+      "learning_rate": 9.043485205380685e-06,
+      "loss": 0.214,
+      "step": 2517
+    },
+    {
+      "epoch": 0.8408749373852062,
+      "grad_norm": 0.435346841429691,
+      "learning_rate": 9.042341782020906e-06,
+      "loss": 0.2163,
+      "step": 2518
+    },
+    {
+      "epoch": 0.8412088829520789,
+      "grad_norm": 0.4680554061197634,
+      "learning_rate": 9.041197748017802e-06,
+      "loss": 0.2162,
+      "step": 2519
+    },
+    {
+      "epoch": 0.8415428285189515,
+      "grad_norm": 0.6331729472174026,
+      "learning_rate": 9.040053103544196e-06,
+      "loss": 0.2171,
+      "step": 2520
+    },
+    {
+      "epoch": 0.8418767740858241,
+      "grad_norm": 0.46302434709386225,
+      "learning_rate": 9.038907848772999e-06,
+      "loss": 0.2313,
+      "step": 2521
+    },
+    {
+      "epoch": 0.8422107196526966,
+      "grad_norm": 0.5418674061177113,
+      "learning_rate": 9.037761983877214e-06,
+      "loss": 0.2314,
+      "step": 2522
+    },
+    {
+      "epoch": 0.8425446652195692,
+      "grad_norm": 0.422599498634045,
+      "learning_rate": 9.036615509029939e-06,
+      "loss": 0.2148,
+      "step": 2523
+    },
+    {
+      "epoch": 0.8428786107864418,
+      "grad_norm": 0.47415409143410847,
+      "learning_rate": 9.035468424404362e-06,
+      "loss": 0.2271,
+      "step": 2524
+    },
+    {
+      "epoch": 0.8432125563533144,
+      "grad_norm": 0.43769467339845197,
+      "learning_rate": 9.034320730173762e-06,
+      "loss": 0.2078,
+      "step": 2525
+    },
+    {
+      "epoch": 0.843546501920187,
+      "grad_norm": 0.4579042598091712,
+      "learning_rate": 9.033172426511515e-06,
+      "loss": 0.2275,
+      "step": 2526
+    },
+    {
+      "epoch": 0.8438804474870596,
+      "grad_norm": 0.472853158936843,
+      "learning_rate": 9.032023513591081e-06,
+      "loss": 0.2163,
+      "step": 2527
+    },
+    {
+      "epoch": 0.8442143930539322,
+      "grad_norm": 0.47883958606918925,
+      "learning_rate": 9.030873991586021e-06,
+      "loss": 0.2362,
+      "step": 2528
+    },
+    {
+      "epoch": 0.8445483386208048,
+      "grad_norm": 0.4244077632216657,
+      "learning_rate": 9.029723860669983e-06,
+      "loss": 0.2135,
+      "step": 2529
+    },
+    {
+      "epoch": 0.8448822841876774,
+      "grad_norm": 0.5703843399860384,
+      "learning_rate": 9.028573121016707e-06,
+      "loss": 0.2595,
+      "step": 2530
+    },
+    {
+      "epoch": 0.84521622975455,
+      "grad_norm": 0.5540172877400666,
+      "learning_rate": 9.027421772800027e-06,
+      "loss": 0.2268,
+      "step": 2531
+    },
+    {
+      "epoch": 0.8455501753214226,
+      "grad_norm": 0.46860925608211035,
+      "learning_rate": 9.026269816193867e-06,
+      "loss": 0.2231,
+      "step": 2532
+    },
+    {
+      "epoch": 0.8458841208882952,
+      "grad_norm": 0.4536040554309427,
+      "learning_rate": 9.025117251372242e-06,
+      "loss": 0.2264,
+      "step": 2533
+    },
+    {
+      "epoch": 0.8462180664551678,
+      "grad_norm": 0.48291163229730805,
+      "learning_rate": 9.023964078509263e-06,
+      "loss": 0.2268,
+      "step": 2534
+    },
+    {
+      "epoch": 0.8465520120220404,
+      "grad_norm": 0.47777184837544556,
+      "learning_rate": 9.022810297779129e-06,
+      "loss": 0.2208,
+      "step": 2535
+    },
+    {
+      "epoch": 0.846885957588913,
+      "grad_norm": 0.46448916819982217,
+      "learning_rate": 9.021655909356133e-06,
+      "loss": 0.2424,
+      "step": 2536
+    },
+    {
+      "epoch": 0.8472199031557857,
+      "grad_norm": 0.47316820975138824,
+      "learning_rate": 9.020500913414658e-06,
+      "loss": 0.2248,
+      "step": 2537
+    },
+    {
+      "epoch": 0.8475538487226582,
+      "grad_norm": 0.4767805328678842,
+      "learning_rate": 9.019345310129179e-06,
+      "loss": 0.2338,
+      "step": 2538
+    },
+    {
+      "epoch": 0.8478877942895308,
+      "grad_norm": 0.4397760724625773,
+      "learning_rate": 9.018189099674266e-06,
+      "loss": 0.2208,
+      "step": 2539
+    },
+    {
+      "epoch": 0.8482217398564034,
+      "grad_norm": 0.4862443013166264,
+      "learning_rate": 9.017032282224577e-06,
+      "loss": 0.2299,
+      "step": 2540
+    },
+    {
+      "epoch": 0.848555685423276,
+      "grad_norm": 0.44379757315440904,
+      "learning_rate": 9.015874857954863e-06,
+      "loss": 0.2192,
+      "step": 2541
+    },
+    {
+      "epoch": 0.8488896309901486,
+      "grad_norm": 0.42679243339558226,
+      "learning_rate": 9.014716827039965e-06,
+      "loss": 0.2222,
+      "step": 2542
+    },
+    {
+      "epoch": 0.8492235765570212,
+      "grad_norm": 0.41430931012260835,
+      "learning_rate": 9.013558189654819e-06,
+      "loss": 0.2161,
+      "step": 2543
+    },
+    {
+      "epoch": 0.8495575221238938,
+      "grad_norm": 0.48620083773360206,
+      "learning_rate": 9.01239894597445e-06,
+      "loss": 0.2191,
+      "step": 2544
+    },
+    {
+      "epoch": 0.8498914676907664,
+      "grad_norm": 0.4914872230641687,
+      "learning_rate": 9.011239096173977e-06,
+      "loss": 0.2205,
+      "step": 2545
+    },
+    {
+      "epoch": 0.850225413257639,
+      "grad_norm": 0.4726040998570721,
+      "learning_rate": 9.010078640428606e-06,
+      "loss": 0.2288,
+      "step": 2546
+    },
+    {
+      "epoch": 0.8505593588245116,
+      "grad_norm": 0.46491513024652575,
+      "learning_rate": 9.00891757891364e-06,
+      "loss": 0.2277,
+      "step": 2547
+    },
+    {
+      "epoch": 0.8508933043913842,
+      "grad_norm": 0.45611845460077005,
+      "learning_rate": 9.007755911804471e-06,
+      "loss": 0.2251,
+      "step": 2548
+    },
+    {
+      "epoch": 0.8512272499582568,
+      "grad_norm": 0.5926560572888228,
+      "learning_rate": 9.006593639276582e-06,
+      "loss": 0.2243,
+      "step": 2549
+    },
+    {
+      "epoch": 0.8515611955251294,
+      "grad_norm": 0.496523833587818,
+      "learning_rate": 9.005430761505548e-06,
+      "loss": 0.2403,
+      "step": 2550
+    },
+    {
+      "epoch": 0.851895141092002,
+      "grad_norm": 0.4714784682154149,
+      "learning_rate": 9.004267278667032e-06,
+      "loss": 0.2149,
+      "step": 2551
+    },
+    {
+      "epoch": 0.8522290866588746,
+      "grad_norm": 0.4313656521047808,
+      "learning_rate": 9.003103190936797e-06,
+      "loss": 0.2187,
+      "step": 2552
+    },
+    {
+      "epoch": 0.8525630322257473,
+      "grad_norm": 0.46928566508359176,
+      "learning_rate": 9.00193849849069e-06,
+      "loss": 0.2208,
+      "step": 2553
+    },
+    {
+      "epoch": 0.8528969777926197,
+      "grad_norm": 0.4283171636804329,
+      "learning_rate": 9.00077320150465e-06,
+      "loss": 0.2112,
+      "step": 2554
+    },
+    {
+      "epoch": 0.8532309233594924,
+      "grad_norm": 0.5000190368184554,
+      "learning_rate": 8.999607300154712e-06,
+      "loss": 0.2192,
+      "step": 2555
+    },
+    {
+      "epoch": 0.853564868926365,
+      "grad_norm": 0.47355749853117557,
+      "learning_rate": 8.998440794616998e-06,
+      "loss": 0.2213,
+      "step": 2556
+    },
+    {
+      "epoch": 0.8538988144932376,
+      "grad_norm": 0.43176568872005033,
+      "learning_rate": 8.99727368506772e-06,
+      "loss": 0.2098,
+      "step": 2557
+    },
+    {
+      "epoch": 0.8542327600601102,
+      "grad_norm": 0.46982069301090845,
+      "learning_rate": 8.996105971683187e-06,
+      "loss": 0.2177,
+      "step": 2558
+    },
+    {
+      "epoch": 0.8545667056269828,
+      "grad_norm": 0.47512821563209395,
+      "learning_rate": 8.994937654639793e-06,
+      "loss": 0.2278,
+      "step": 2559
+    },
+    {
+      "epoch": 0.8549006511938554,
+      "grad_norm": 0.4000965489856469,
+      "learning_rate": 8.993768734114029e-06,
+      "loss": 0.2023,
+      "step": 2560
+    },
+    {
+      "epoch": 0.855234596760728,
+      "grad_norm": 0.43268062516172295,
+      "learning_rate": 8.992599210282471e-06,
+      "loss": 0.2202,
+      "step": 2561
+    },
+    {
+      "epoch": 0.8555685423276006,
+      "grad_norm": 0.42305087012502596,
+      "learning_rate": 8.991429083321792e-06,
+      "loss": 0.2189,
+      "step": 2562
+    },
+    {
+      "epoch": 0.8559024878944732,
+      "grad_norm": 0.4708842682056169,
+      "learning_rate": 8.990258353408754e-06,
+      "loss": 0.2295,
+      "step": 2563
+    },
+    {
+      "epoch": 0.8562364334613458,
+      "grad_norm": 0.4600816021897864,
+      "learning_rate": 8.989087020720204e-06,
+      "loss": 0.2193,
+      "step": 2564
+    },
+    {
+      "epoch": 0.8565703790282184,
+      "grad_norm": 0.4198834665228543,
+      "learning_rate": 8.987915085433092e-06,
+      "loss": 0.2195,
+      "step": 2565
+    },
+    {
+      "epoch": 0.856904324595091,
+      "grad_norm": 0.4489653916885646,
+      "learning_rate": 8.98674254772445e-06,
+      "loss": 0.2126,
+      "step": 2566
+    },
+    {
+      "epoch": 0.8572382701619636,
+      "grad_norm": 0.40367098282319236,
+      "learning_rate": 8.985569407771404e-06,
+      "loss": 0.1956,
+      "step": 2567
+    },
+    {
+      "epoch": 0.8575722157288362,
+      "grad_norm": 0.4732011179689396,
+      "learning_rate": 8.984395665751169e-06,
+      "loss": 0.2296,
+      "step": 2568
+    },
+    {
+      "epoch": 0.8579061612957088,
+      "grad_norm": 0.47840121907574584,
+      "learning_rate": 8.983221321841056e-06,
+      "loss": 0.2299,
+      "step": 2569
+    },
+    {
+      "epoch": 0.8582401068625815,
+      "grad_norm": 0.4465691457596304,
+      "learning_rate": 8.98204637621846e-06,
+      "loss": 0.2143,
+      "step": 2570
+    },
+    {
+      "epoch": 0.858574052429454,
+      "grad_norm": 0.47090014081524445,
+      "learning_rate": 8.980870829060872e-06,
+      "loss": 0.2255,
+      "step": 2571
+    },
+    {
+      "epoch": 0.8589079979963266,
+      "grad_norm": 0.46190459557837005,
+      "learning_rate": 8.979694680545872e-06,
+      "loss": 0.2164,
+      "step": 2572
+    },
+    {
+      "epoch": 0.8592419435631992,
+      "grad_norm": 0.3984766673387074,
+      "learning_rate": 8.978517930851132e-06,
+      "loss": 0.2025,
+      "step": 2573
+    },
+    {
+      "epoch": 0.8595758891300718,
+      "grad_norm": 0.4881647227380028,
+      "learning_rate": 8.977340580154411e-06,
+      "loss": 0.2222,
+      "step": 2574
+    },
+    {
+      "epoch": 0.8599098346969444,
+      "grad_norm": 0.4577249839436491,
+      "learning_rate": 8.976162628633565e-06,
+      "loss": 0.2284,
+      "step": 2575
+    },
+    {
+      "epoch": 0.860243780263817,
+      "grad_norm": 0.42843210433451573,
+      "learning_rate": 8.974984076466537e-06,
+      "loss": 0.228,
+      "step": 2576
+    },
+    {
+      "epoch": 0.8605777258306896,
+      "grad_norm": 0.46579319758112697,
+      "learning_rate": 8.97380492383136e-06,
+      "loss": 0.2117,
+      "step": 2577
+    },
+    {
+      "epoch": 0.8609116713975622,
+      "grad_norm": 0.4552220225147081,
+      "learning_rate": 8.972625170906157e-06,
+      "loss": 0.2316,
+      "step": 2578
+    },
+    {
+      "epoch": 0.8612456169644348,
+      "grad_norm": 0.42726184700574793,
+      "learning_rate": 8.971444817869148e-06,
+      "loss": 0.2198,
+      "step": 2579
+    },
+    {
+      "epoch": 0.8615795625313074,
+      "grad_norm": 0.45374200275965654,
+      "learning_rate": 8.970263864898636e-06,
+      "loss": 0.2261,
+      "step": 2580
+    },
+    {
+      "epoch": 0.86191350809818,
+      "grad_norm": 0.5381715281038784,
+      "learning_rate": 8.969082312173021e-06,
+      "loss": 0.2344,
+      "step": 2581
+    },
+    {
+      "epoch": 0.8622474536650526,
+      "grad_norm": 0.4754027335097857,
+      "learning_rate": 8.967900159870787e-06,
+      "loss": 0.2293,
+      "step": 2582
+    },
+    {
+      "epoch": 0.8625813992319252,
+      "grad_norm": 0.4460243245841365,
+      "learning_rate": 8.966717408170512e-06,
+      "loss": 0.2129,
+      "step": 2583
+    },
+    {
+      "epoch": 0.8629153447987978,
+      "grad_norm": 0.471562541587052,
+      "learning_rate": 8.965534057250866e-06,
+      "loss": 0.232,
+      "step": 2584
+    },
+    {
+      "epoch": 0.8632492903656704,
+      "grad_norm": 0.4426034976537733,
+      "learning_rate": 8.964350107290609e-06,
+      "loss": 0.2099,
+      "step": 2585
+    },
+    {
+      "epoch": 0.863583235932543,
+      "grad_norm": 0.4588021928348187,
+      "learning_rate": 8.96316555846859e-06,
+      "loss": 0.2205,
+      "step": 2586
+    },
+    {
+      "epoch": 0.8639171814994155,
+      "grad_norm": 0.42630016568299534,
+      "learning_rate": 8.961980410963749e-06,
+      "loss": 0.2135,
+      "step": 2587
+    },
+    {
+      "epoch": 0.8642511270662881,
+      "grad_norm": 0.45810183824218453,
+      "learning_rate": 8.960794664955115e-06,
+      "loss": 0.2229,
+      "step": 2588
+    },
+    {
+      "epoch": 0.8645850726331608,
+      "grad_norm": 0.42559767116343933,
+      "learning_rate": 8.95960832062181e-06,
+      "loss": 0.2156,
+      "step": 2589
+    },
+    {
+      "epoch": 0.8649190182000334,
+      "grad_norm": 0.4245683101087079,
+      "learning_rate": 8.958421378143046e-06,
+      "loss": 0.22,
+      "step": 2590
+    },
+    {
+      "epoch": 0.865252963766906,
+      "grad_norm": 0.5033515035437922,
+      "learning_rate": 8.957233837698122e-06,
+      "loss": 0.2298,
+      "step": 2591
+    },
+    {
+      "epoch": 0.8655869093337786,
+      "grad_norm": 0.46700010068824005,
+      "learning_rate": 8.956045699466433e-06,
+      "loss": 0.2344,
+      "step": 2592
+    },
+    {
+      "epoch": 0.8659208549006512,
+      "grad_norm": 0.4434776982903602,
+      "learning_rate": 8.95485696362746e-06,
+      "loss": 0.2143,
+      "step": 2593
+    },
+    {
+      "epoch": 0.8662548004675238,
+      "grad_norm": 0.432504424021459,
+      "learning_rate": 8.953667630360778e-06,
+      "loss": 0.2135,
+      "step": 2594
+    },
+    {
+      "epoch": 0.8665887460343964,
+      "grad_norm": 0.4761778540131483,
+      "learning_rate": 8.952477699846044e-06,
+      "loss": 0.2311,
+      "step": 2595
+    },
+    {
+      "epoch": 0.866922691601269,
+      "grad_norm": 0.41699684183795516,
+      "learning_rate": 8.951287172263018e-06,
+      "loss": 0.2181,
+      "step": 2596
+    },
+    {
+      "epoch": 0.8672566371681416,
+      "grad_norm": 0.4546797953736065,
+      "learning_rate": 8.950096047791539e-06,
+      "loss": 0.2311,
+      "step": 2597
+    },
+    {
+      "epoch": 0.8675905827350142,
+      "grad_norm": 0.4793027535037006,
+      "learning_rate": 8.94890432661154e-06,
+      "loss": 0.2416,
+      "step": 2598
+    },
+    {
+      "epoch": 0.8679245283018868,
+      "grad_norm": 0.4259729344924639,
+      "learning_rate": 8.947712008903045e-06,
+      "loss": 0.2164,
+      "step": 2599
+    },
+    {
+      "epoch": 0.8682584738687594,
+      "grad_norm": 0.45367870809652466,
+      "learning_rate": 8.946519094846169e-06,
+      "loss": 0.2168,
+      "step": 2600
+    },
+    {
+      "epoch": 0.868592419435632,
+      "grad_norm": 0.4283930933516754,
+      "learning_rate": 8.945325584621116e-06,
+      "loss": 0.2048,
+      "step": 2601
+    },
+    {
+      "epoch": 0.8689263650025046,
+      "grad_norm": 0.47811696623877187,
+      "learning_rate": 8.944131478408177e-06,
+      "loss": 0.2256,
+      "step": 2602
+    },
+    {
+      "epoch": 0.8692603105693771,
+      "grad_norm": 0.474177095776076,
+      "learning_rate": 8.942936776387739e-06,
+      "loss": 0.2199,
+      "step": 2603
+    },
+    {
+      "epoch": 0.8695942561362497,
+      "grad_norm": 0.449120897392552,
+      "learning_rate": 8.941741478740272e-06,
+      "loss": 0.2251,
+      "step": 2604
+    },
+    {
+      "epoch": 0.8699282017031224,
+      "grad_norm": 0.3831811961730255,
+      "learning_rate": 8.940545585646344e-06,
+      "loss": 0.1982,
+      "step": 2605
+    },
+    {
+      "epoch": 0.870262147269995,
+      "grad_norm": 0.46919762256557285,
+      "learning_rate": 8.939349097286608e-06,
+      "loss": 0.237,
+      "step": 2606
+    },
+    {
+      "epoch": 0.8705960928368676,
+      "grad_norm": 0.4595681283636294,
+      "learning_rate": 8.938152013841803e-06,
+      "loss": 0.2359,
+      "step": 2607
+    },
+    {
+      "epoch": 0.8709300384037402,
+      "grad_norm": 0.46244221082714243,
+      "learning_rate": 8.93695433549277e-06,
+      "loss": 0.2172,
+      "step": 2608
+    },
+    {
+      "epoch": 0.8712639839706128,
+      "grad_norm": 0.44144435300671014,
+      "learning_rate": 8.935756062420426e-06,
+      "loss": 0.2259,
+      "step": 2609
+    },
+    {
+      "epoch": 0.8715979295374854,
+      "grad_norm": 0.4735100176435913,
+      "learning_rate": 8.934557194805787e-06,
+      "loss": 0.2221,
+      "step": 2610
+    },
+    {
+      "epoch": 0.871931875104358,
+      "grad_norm": 0.44284418340652504,
+      "learning_rate": 8.933357732829957e-06,
+      "loss": 0.2265,
+      "step": 2611
+    },
+    {
+      "epoch": 0.8722658206712306,
+      "grad_norm": 0.4659415166073965,
+      "learning_rate": 8.932157676674126e-06,
+      "loss": 0.2276,
+      "step": 2612
+    },
+    {
+      "epoch": 0.8725997662381032,
+      "grad_norm": 0.41233515606202303,
+      "learning_rate": 8.93095702651958e-06,
+      "loss": 0.2074,
+      "step": 2613
+    },
+    {
+      "epoch": 0.8729337118049758,
+      "grad_norm": 0.4877780049978787,
+      "learning_rate": 8.92975578254769e-06,
+      "loss": 0.2313,
+      "step": 2614
+    },
+    {
+      "epoch": 0.8732676573718484,
+      "grad_norm": 0.5150208995785533,
+      "learning_rate": 8.928553944939915e-06,
+      "loss": 0.2156,
+      "step": 2615
+    },
+    {
+      "epoch": 0.873601602938721,
+      "grad_norm": 0.4168813001248241,
+      "learning_rate": 8.92735151387781e-06,
+      "loss": 0.2222,
+      "step": 2616
+    },
+    {
+      "epoch": 0.8739355485055936,
+      "grad_norm": 0.4098574760494883,
+      "learning_rate": 8.926148489543018e-06,
+      "loss": 0.2175,
+      "step": 2617
+    },
+    {
+      "epoch": 0.8742694940724662,
+      "grad_norm": 0.4412115939247315,
+      "learning_rate": 8.924944872117264e-06,
+      "loss": 0.2195,
+      "step": 2618
+    },
+    {
+      "epoch": 0.8746034396393388,
+      "grad_norm": 0.4630857578664083,
+      "learning_rate": 8.923740661782376e-06,
+      "loss": 0.2195,
+      "step": 2619
+    },
+    {
+      "epoch": 0.8749373852062113,
+      "grad_norm": 0.48566643237103263,
+      "learning_rate": 8.92253585872026e-06,
+      "loss": 0.2351,
+      "step": 2620
+    },
+    {
+      "epoch": 0.8752713307730839,
+      "grad_norm": 0.43311552083576144,
+      "learning_rate": 8.921330463112915e-06,
+      "loss": 0.2288,
+      "step": 2621
+    },
+    {
+      "epoch": 0.8756052763399566,
+      "grad_norm": 0.39507171452378326,
+      "learning_rate": 8.92012447514243e-06,
+      "loss": 0.207,
+      "step": 2622
+    },
+    {
+      "epoch": 0.8759392219068292,
+      "grad_norm": 0.4452755906788476,
+      "learning_rate": 8.918917894990989e-06,
+      "loss": 0.1985,
+      "step": 2623
+    },
+    {
+      "epoch": 0.8762731674737018,
+      "grad_norm": 0.3910814697165605,
+      "learning_rate": 8.917710722840853e-06,
+      "loss": 0.2066,
+      "step": 2624
+    },
+    {
+      "epoch": 0.8766071130405744,
+      "grad_norm": 0.4239629843736424,
+      "learning_rate": 8.916502958874385e-06,
+      "loss": 0.2274,
+      "step": 2625
+    },
+    {
+      "epoch": 0.876941058607447,
+      "grad_norm": 0.45349007585563317,
+      "learning_rate": 8.915294603274027e-06,
+      "loss": 0.2205,
+      "step": 2626
+    },
+    {
+      "epoch": 0.8772750041743196,
+      "grad_norm": 0.42342259610566313,
+      "learning_rate": 8.91408565622232e-06,
+      "loss": 0.2065,
+      "step": 2627
+    },
+    {
+      "epoch": 0.8776089497411922,
+      "grad_norm": 0.44535591900172594,
+      "learning_rate": 8.912876117901887e-06,
+      "loss": 0.2203,
+      "step": 2628
+    },
+    {
+      "epoch": 0.8779428953080648,
+      "grad_norm": 0.46320851664376284,
+      "learning_rate": 8.911665988495446e-06,
+      "loss": 0.2432,
+      "step": 2629
+    },
+    {
+      "epoch": 0.8782768408749374,
+      "grad_norm": 0.4367734120566955,
+      "learning_rate": 8.910455268185795e-06,
+      "loss": 0.2166,
+      "step": 2630
+    },
+    {
+      "epoch": 0.87861078644181,
+      "grad_norm": 0.43971258610914693,
+      "learning_rate": 8.909243957155835e-06,
+      "loss": 0.2313,
+      "step": 2631
+    },
+    {
+      "epoch": 0.8789447320086826,
+      "grad_norm": 0.408475957094058,
+      "learning_rate": 8.908032055588544e-06,
+      "loss": 0.2282,
+      "step": 2632
+    },
+    {
+      "epoch": 0.8792786775755552,
+      "grad_norm": 0.4413318202987162,
+      "learning_rate": 8.906819563666997e-06,
+      "loss": 0.2325,
+      "step": 2633
+    },
+    {
+      "epoch": 0.8796126231424278,
+      "grad_norm": 0.4204393958934717,
+      "learning_rate": 8.905606481574351e-06,
+      "loss": 0.2212,
+      "step": 2634
+    },
+    {
+      "epoch": 0.8799465687093004,
+      "grad_norm": 0.43071251695265306,
+      "learning_rate": 8.90439280949386e-06,
+      "loss": 0.2153,
+      "step": 2635
+    },
+    {
+      "epoch": 0.8802805142761729,
+      "grad_norm": 0.44850361365375907,
+      "learning_rate": 8.903178547608863e-06,
+      "loss": 0.2221,
+      "step": 2636
+    },
+    {
+      "epoch": 0.8806144598430455,
+      "grad_norm": 0.42781213117927985,
+      "learning_rate": 8.901963696102788e-06,
+      "loss": 0.2306,
+      "step": 2637
+    },
+    {
+      "epoch": 0.8809484054099181,
+      "grad_norm": 0.42113339754508183,
+      "learning_rate": 8.900748255159152e-06,
+      "loss": 0.2111,
+      "step": 2638
+    },
+    {
+      "epoch": 0.8812823509767908,
+      "grad_norm": 0.47605309867063816,
+      "learning_rate": 8.899532224961562e-06,
+      "loss": 0.229,
+      "step": 2639
+    },
+    {
+      "epoch": 0.8816162965436634,
+      "grad_norm": 0.5527061292754034,
+      "learning_rate": 8.898315605693715e-06,
+      "loss": 0.217,
+      "step": 2640
+    },
+    {
+      "epoch": 0.881950242110536,
+      "grad_norm": 0.43906596005312143,
+      "learning_rate": 8.897098397539394e-06,
+      "loss": 0.2164,
+      "step": 2641
+    },
+    {
+      "epoch": 0.8822841876774086,
+      "grad_norm": 0.5067302891421397,
+      "learning_rate": 8.895880600682472e-06,
+      "loss": 0.2403,
+      "step": 2642
+    },
+    {
+      "epoch": 0.8826181332442812,
+      "grad_norm": 0.48526179484087406,
+      "learning_rate": 8.894662215306913e-06,
+      "loss": 0.216,
+      "step": 2643
+    },
+    {
+      "epoch": 0.8829520788111538,
+      "grad_norm": 0.5004998831458609,
+      "learning_rate": 8.89344324159677e-06,
+      "loss": 0.2182,
+      "step": 2644
+    },
+    {
+      "epoch": 0.8832860243780264,
+      "grad_norm": 0.4141565409924,
+      "learning_rate": 8.89222367973618e-06,
+      "loss": 0.2173,
+      "step": 2645
+    },
+    {
+      "epoch": 0.883619969944899,
+      "grad_norm": 0.451587274844467,
+      "learning_rate": 8.891003529909375e-06,
+      "loss": 0.2258,
+      "step": 2646
+    },
+    {
+      "epoch": 0.8839539155117716,
+      "grad_norm": 0.4357097873514739,
+      "learning_rate": 8.889782792300672e-06,
+      "loss": 0.2176,
+      "step": 2647
+    },
+    {
+      "epoch": 0.8842878610786442,
+      "grad_norm": 0.4424588073002416,
+      "learning_rate": 8.888561467094476e-06,
+      "loss": 0.2173,
+      "step": 2648
+    },
+    {
+      "epoch": 0.8846218066455168,
+      "grad_norm": 0.47374105210343537,
+      "learning_rate": 8.887339554475284e-06,
+      "loss": 0.2167,
+      "step": 2649
+    },
+    {
+      "epoch": 0.8849557522123894,
+      "grad_norm": 0.4314904279729543,
+      "learning_rate": 8.886117054627682e-06,
+      "loss": 0.2114,
+      "step": 2650
+    },
+    {
+      "epoch": 0.885289697779262,
+      "grad_norm": 0.48980056249973974,
+      "learning_rate": 8.88489396773634e-06,
+      "loss": 0.2344,
+      "step": 2651
+    },
+    {
+      "epoch": 0.8856236433461345,
+      "grad_norm": 0.38805445786960013,
+      "learning_rate": 8.883670293986019e-06,
+      "loss": 0.2056,
+      "step": 2652
+    },
+    {
+      "epoch": 0.8859575889130071,
+      "grad_norm": 0.4178809237058471,
+      "learning_rate": 8.882446033561576e-06,
+      "loss": 0.2242,
+      "step": 2653
+    },
+    {
+      "epoch": 0.8862915344798797,
+      "grad_norm": 0.4936359585377921,
+      "learning_rate": 8.881221186647941e-06,
+      "loss": 0.2183,
+      "step": 2654
+    },
+    {
+      "epoch": 0.8866254800467523,
+      "grad_norm": 0.4859800139169095,
+      "learning_rate": 8.879995753430148e-06,
+      "loss": 0.2178,
+      "step": 2655
+    },
+    {
+      "epoch": 0.886959425613625,
+      "grad_norm": 0.48946455778130216,
+      "learning_rate": 8.878769734093312e-06,
+      "loss": 0.2203,
+      "step": 2656
+    },
+    {
+      "epoch": 0.8872933711804976,
+      "grad_norm": 0.41725432475224355,
+      "learning_rate": 8.877543128822634e-06,
+      "loss": 0.2265,
+      "step": 2657
+    },
+    {
+      "epoch": 0.8876273167473702,
+      "grad_norm": 0.4663793748591142,
+      "learning_rate": 8.876315937803413e-06,
+      "loss": 0.2262,
+      "step": 2658
+    },
+    {
+      "epoch": 0.8879612623142428,
+      "grad_norm": 0.3916146858993435,
+      "learning_rate": 8.875088161221025e-06,
+      "loss": 0.2072,
+      "step": 2659
+    },
+    {
+      "epoch": 0.8882952078811154,
+      "grad_norm": 0.47716488660190576,
+      "learning_rate": 8.873859799260944e-06,
+      "loss": 0.2245,
+      "step": 2660
+    },
+    {
+      "epoch": 0.888629153447988,
+      "grad_norm": 0.42641030028705323,
+      "learning_rate": 8.872630852108725e-06,
+      "loss": 0.2301,
+      "step": 2661
+    },
+    {
+      "epoch": 0.8889630990148606,
+      "grad_norm": 0.4117929648291072,
+      "learning_rate": 8.87140131995002e-06,
+      "loss": 0.2303,
+      "step": 2662
+    },
+    {
+      "epoch": 0.8892970445817332,
+      "grad_norm": 0.4161638007577847,
+      "learning_rate": 8.870171202970559e-06,
+      "loss": 0.2121,
+      "step": 2663
+    },
+    {
+      "epoch": 0.8896309901486058,
+      "grad_norm": 0.4530085404187551,
+      "learning_rate": 8.868940501356169e-06,
+      "loss": 0.2248,
+      "step": 2664
+    },
+    {
+      "epoch": 0.8899649357154784,
+      "grad_norm": 0.40243303655245055,
+      "learning_rate": 8.86770921529276e-06,
+      "loss": 0.2135,
+      "step": 2665
+    },
+    {
+      "epoch": 0.890298881282351,
+      "grad_norm": 0.4473350481589006,
+      "learning_rate": 8.866477344966334e-06,
+      "loss": 0.216,
+      "step": 2666
+    },
+    {
+      "epoch": 0.8906328268492236,
+      "grad_norm": 0.41627719547498515,
+      "learning_rate": 8.865244890562978e-06,
+      "loss": 0.2174,
+      "step": 2667
+    },
+    {
+      "epoch": 0.8909667724160962,
+      "grad_norm": 0.5442020904005506,
+      "learning_rate": 8.864011852268872e-06,
+      "loss": 0.2261,
+      "step": 2668
+    },
+    {
+      "epoch": 0.8913007179829687,
+      "grad_norm": 0.4366841475309793,
+      "learning_rate": 8.862778230270276e-06,
+      "loss": 0.2166,
+      "step": 2669
+    },
+    {
+      "epoch": 0.8916346635498413,
+      "grad_norm": 0.5358299377296973,
+      "learning_rate": 8.861544024753545e-06,
+      "loss": 0.256,
+      "step": 2670
+    },
+    {
+      "epoch": 0.8919686091167139,
+      "grad_norm": 0.7520222436984267,
+      "learning_rate": 8.860309235905122e-06,
+      "loss": 0.2221,
+      "step": 2671
+    },
+    {
+      "epoch": 0.8923025546835865,
+      "grad_norm": 0.456864799154424,
+      "learning_rate": 8.859073863911536e-06,
+      "loss": 0.2284,
+      "step": 2672
+    },
+    {
+      "epoch": 0.8926365002504592,
+      "grad_norm": 0.4443603685723179,
+      "learning_rate": 8.857837908959404e-06,
+      "loss": 0.2192,
+      "step": 2673
+    },
+    {
+      "epoch": 0.8929704458173318,
+      "grad_norm": 0.44649147782698073,
+      "learning_rate": 8.856601371235429e-06,
+      "loss": 0.2153,
+      "step": 2674
+    },
+    {
+      "epoch": 0.8933043913842044,
+      "grad_norm": 0.5127058305862575,
+      "learning_rate": 8.855364250926409e-06,
+      "loss": 0.2328,
+      "step": 2675
+    },
+    {
+      "epoch": 0.893638336951077,
+      "grad_norm": 0.42241283375166006,
+      "learning_rate": 8.854126548219222e-06,
+      "loss": 0.2144,
+      "step": 2676
+    },
+    {
+      "epoch": 0.8939722825179496,
+      "grad_norm": 0.44645580116752953,
+      "learning_rate": 8.85288826330084e-06,
+      "loss": 0.2217,
+      "step": 2677
+    },
+    {
+      "epoch": 0.8943062280848222,
+      "grad_norm": 0.4843402606886858,
+      "learning_rate": 8.85164939635832e-06,
+      "loss": 0.2058,
+      "step": 2678
+    },
+    {
+      "epoch": 0.8946401736516948,
+      "grad_norm": 0.4401205110768855,
+      "learning_rate": 8.850409947578806e-06,
+      "loss": 0.2255,
+      "step": 2679
+    },
+    {
+      "epoch": 0.8949741192185674,
+      "grad_norm": 0.46319114729765054,
+      "learning_rate": 8.849169917149532e-06,
+      "loss": 0.2152,
+      "step": 2680
+    },
+    {
+      "epoch": 0.89530806478544,
+      "grad_norm": 0.42318735608832414,
+      "learning_rate": 8.847929305257821e-06,
+      "loss": 0.2072,
+      "step": 2681
+    },
+    {
+      "epoch": 0.8956420103523126,
+      "grad_norm": 0.4774593411729633,
+      "learning_rate": 8.846688112091078e-06,
+      "loss": 0.2164,
+      "step": 2682
+    },
+    {
+      "epoch": 0.8959759559191852,
+      "grad_norm": 0.5172363111160776,
+      "learning_rate": 8.845446337836805e-06,
+      "loss": 0.2199,
+      "step": 2683
+    },
+    {
+      "epoch": 0.8963099014860578,
+      "grad_norm": 0.47524030222731717,
+      "learning_rate": 8.844203982682583e-06,
+      "loss": 0.2118,
+      "step": 2684
+    },
+    {
+      "epoch": 0.8966438470529303,
+      "grad_norm": 0.48935917439328974,
+      "learning_rate": 8.842961046816085e-06,
+      "loss": 0.2246,
+      "step": 2685
+    },
+    {
+      "epoch": 0.8969777926198029,
+      "grad_norm": 0.42309667094002784,
+      "learning_rate": 8.841717530425071e-06,
+      "loss": 0.2059,
+      "step": 2686
+    },
+    {
+      "epoch": 0.8973117381866755,
+      "grad_norm": 0.43840157623404685,
+      "learning_rate": 8.84047343369739e-06,
+      "loss": 0.2208,
+      "step": 2687
+    },
+    {
+      "epoch": 0.8976456837535481,
+      "grad_norm": 0.4595498272411674,
+      "learning_rate": 8.839228756820977e-06,
+      "loss": 0.2178,
+      "step": 2688
+    },
+    {
+      "epoch": 0.8979796293204207,
+      "grad_norm": 0.5061728834639878,
+      "learning_rate": 8.837983499983856e-06,
+      "loss": 0.2409,
+      "step": 2689
+    },
+    {
+      "epoch": 0.8983135748872934,
+      "grad_norm": 0.4681121712319394,
+      "learning_rate": 8.836737663374135e-06,
+      "loss": 0.2213,
+      "step": 2690
+    },
+    {
+      "epoch": 0.898647520454166,
+      "grad_norm": 0.47022403674358887,
+      "learning_rate": 8.835491247180012e-06,
+      "loss": 0.2266,
+      "step": 2691
+    },
+    {
+      "epoch": 0.8989814660210386,
+      "grad_norm": 0.46339165662113857,
+      "learning_rate": 8.834244251589778e-06,
+      "loss": 0.2293,
+      "step": 2692
+    },
+    {
+      "epoch": 0.8993154115879112,
+      "grad_norm": 0.5508066598437971,
+      "learning_rate": 8.832996676791802e-06,
+      "loss": 0.2296,
+      "step": 2693
+    },
+    {
+      "epoch": 0.8996493571547838,
+      "grad_norm": 0.3974241538911554,
+      "learning_rate": 8.831748522974545e-06,
+      "loss": 0.2016,
+      "step": 2694
+    },
+    {
+      "epoch": 0.8999833027216564,
+      "grad_norm": 0.46507520916790673,
+      "learning_rate": 8.830499790326556e-06,
+      "loss": 0.2204,
+      "step": 2695
+    },
+    {
+      "epoch": 0.900317248288529,
+      "grad_norm": 0.4080434870525821,
+      "learning_rate": 8.829250479036473e-06,
+      "loss": 0.2098,
+      "step": 2696
+    },
+    {
+      "epoch": 0.9006511938554016,
+      "grad_norm": 0.4703823065746979,
+      "learning_rate": 8.828000589293016e-06,
+      "loss": 0.2096,
+      "step": 2697
+    },
+    {
+      "epoch": 0.9009851394222742,
+      "grad_norm": 0.42236956688946864,
+      "learning_rate": 8.826750121284998e-06,
+      "loss": 0.2083,
+      "step": 2698
+    },
+    {
+      "epoch": 0.9013190849891468,
+      "grad_norm": 0.44662974396124566,
+      "learning_rate": 8.825499075201314e-06,
+      "loss": 0.2245,
+      "step": 2699
+    },
+    {
+      "epoch": 0.9016530305560194,
+      "grad_norm": 0.43992141849148725,
+      "learning_rate": 8.824247451230949e-06,
+      "loss": 0.2171,
+      "step": 2700
+    },
+    {
+      "epoch": 0.9019869761228919,
+      "grad_norm": 0.5054985915126862,
+      "learning_rate": 8.82299524956298e-06,
+      "loss": 0.2374,
+      "step": 2701
+    },
+    {
+      "epoch": 0.9023209216897645,
+      "grad_norm": 0.384565909847055,
+      "learning_rate": 8.821742470386565e-06,
+      "loss": 0.1991,
+      "step": 2702
+    },
+    {
+      "epoch": 0.9026548672566371,
+      "grad_norm": 0.4712108137917473,
+      "learning_rate": 8.820489113890949e-06,
+      "loss": 0.227,
+      "step": 2703
+    },
+    {
+      "epoch": 0.9029888128235097,
+      "grad_norm": 0.4162460345682535,
+      "learning_rate": 8.819235180265468e-06,
+      "loss": 0.209,
+      "step": 2704
+    },
+    {
+      "epoch": 0.9033227583903823,
+      "grad_norm": 0.3951344465539853,
+      "learning_rate": 8.817980669699544e-06,
+      "loss": 0.1994,
+      "step": 2705
+    },
+    {
+      "epoch": 0.903656703957255,
+      "grad_norm": 0.49667700484753613,
+      "learning_rate": 8.816725582382681e-06,
+      "loss": 0.2251,
+      "step": 2706
+    },
+    {
+      "epoch": 0.9039906495241276,
+      "grad_norm": 0.444646491895196,
+      "learning_rate": 8.815469918504482e-06,
+      "loss": 0.2253,
+      "step": 2707
+    },
+    {
+      "epoch": 0.9043245950910002,
+      "grad_norm": 0.47116995697716324,
+      "learning_rate": 8.814213678254624e-06,
+      "loss": 0.2323,
+      "step": 2708
+    },
+    {
+      "epoch": 0.9046585406578728,
+      "grad_norm": 0.4294065077553617,
+      "learning_rate": 8.81295686182288e-06,
+      "loss": 0.2155,
+      "step": 2709
+    },
+    {
+      "epoch": 0.9049924862247454,
+      "grad_norm": 0.4643290288473964,
+      "learning_rate": 8.811699469399106e-06,
+      "loss": 0.2257,
+      "step": 2710
+    },
+    {
+      "epoch": 0.905326431791618,
+      "grad_norm": 0.5106255418157994,
+      "learning_rate": 8.810441501173245e-06,
+      "loss": 0.2198,
+      "step": 2711
+    },
+    {
+      "epoch": 0.9056603773584906,
+      "grad_norm": 0.47059697891525737,
+      "learning_rate": 8.809182957335329e-06,
+      "loss": 0.2193,
+      "step": 2712
+    },
+    {
+      "epoch": 0.9059943229253632,
+      "grad_norm": 0.4121796907160016,
+      "learning_rate": 8.807923838075476e-06,
+      "loss": 0.2255,
+      "step": 2713
+    },
+    {
+      "epoch": 0.9063282684922358,
+      "grad_norm": 0.4914052442580082,
+      "learning_rate": 8.80666414358389e-06,
+      "loss": 0.2386,
+      "step": 2714
+    },
+    {
+      "epoch": 0.9066622140591084,
+      "grad_norm": 0.5300383794010388,
+      "learning_rate": 8.805403874050864e-06,
+      "loss": 0.2275,
+      "step": 2715
+    },
+    {
+      "epoch": 0.906996159625981,
+      "grad_norm": 0.4420574700910713,
+      "learning_rate": 8.804143029666775e-06,
+      "loss": 0.2207,
+      "step": 2716
+    },
+    {
+      "epoch": 0.9073301051928536,
+      "grad_norm": 0.4569049390299293,
+      "learning_rate": 8.802881610622089e-06,
+      "loss": 0.2102,
+      "step": 2717
+    },
+    {
+      "epoch": 0.9076640507597261,
+      "grad_norm": 0.4138994917550044,
+      "learning_rate": 8.801619617107359e-06,
+      "loss": 0.2229,
+      "step": 2718
+    },
+    {
+      "epoch": 0.9079979963265987,
+      "grad_norm": 0.4794838835897948,
+      "learning_rate": 8.800357049313222e-06,
+      "loss": 0.2354,
+      "step": 2719
+    },
+    {
+      "epoch": 0.9083319418934713,
+      "grad_norm": 0.4589463845784919,
+      "learning_rate": 8.799093907430406e-06,
+      "loss": 0.2195,
+      "step": 2720
+    },
+    {
+      "epoch": 0.9086658874603439,
+      "grad_norm": 0.40781923211396,
+      "learning_rate": 8.797830191649721e-06,
+      "loss": 0.2192,
+      "step": 2721
+    },
+    {
+      "epoch": 0.9089998330272165,
+      "grad_norm": 0.4177098084042151,
+      "learning_rate": 8.796565902162069e-06,
+      "loss": 0.2096,
+      "step": 2722
+    },
+    {
+      "epoch": 0.9093337785940891,
+      "grad_norm": 0.4323255277508984,
+      "learning_rate": 8.795301039158433e-06,
+      "loss": 0.2233,
+      "step": 2723
+    },
+    {
+      "epoch": 0.9096677241609618,
+      "grad_norm": 0.4587198185489595,
+      "learning_rate": 8.794035602829887e-06,
+      "loss": 0.2419,
+      "step": 2724
+    },
+    {
+      "epoch": 0.9100016697278344,
+      "grad_norm": 0.4542463004544109,
+      "learning_rate": 8.792769593367591e-06,
+      "loss": 0.2179,
+      "step": 2725
+    },
+    {
+      "epoch": 0.910335615294707,
+      "grad_norm": 0.4990418712075424,
+      "learning_rate": 8.79150301096279e-06,
+      "loss": 0.2238,
+      "step": 2726
+    },
+    {
+      "epoch": 0.9106695608615796,
+      "grad_norm": 0.41162778959199753,
+      "learning_rate": 8.790235855806814e-06,
+      "loss": 0.2127,
+      "step": 2727
+    },
+    {
+      "epoch": 0.9110035064284522,
+      "grad_norm": 0.4377098192081709,
+      "learning_rate": 8.788968128091084e-06,
+      "loss": 0.2171,
+      "step": 2728
+    },
+    {
+      "epoch": 0.9113374519953248,
+      "grad_norm": 0.4511181347047479,
+      "learning_rate": 8.787699828007104e-06,
+      "loss": 0.2264,
+      "step": 2729
+    },
+    {
+      "epoch": 0.9116713975621974,
+      "grad_norm": 0.44193992386166336,
+      "learning_rate": 8.786430955746468e-06,
+      "loss": 0.2142,
+      "step": 2730
+    },
+    {
+      "epoch": 0.91200534312907,
+      "grad_norm": 0.5164407737903689,
+      "learning_rate": 8.78516151150085e-06,
+      "loss": 0.2258,
+      "step": 2731
+    },
+    {
+      "epoch": 0.9123392886959426,
+      "grad_norm": 0.45102390120947355,
+      "learning_rate": 8.783891495462018e-06,
+      "loss": 0.2182,
+      "step": 2732
+    },
+    {
+      "epoch": 0.9126732342628152,
+      "grad_norm": 0.4717630713459086,
+      "learning_rate": 8.782620907821823e-06,
+      "loss": 0.216,
+      "step": 2733
+    },
+    {
+      "epoch": 0.9130071798296877,
+      "grad_norm": 0.4192665748513033,
+      "learning_rate": 8.781349748772198e-06,
+      "loss": 0.2054,
+      "step": 2734
+    },
+    {
+      "epoch": 0.9133411253965603,
+      "grad_norm": 0.5708074422120173,
+      "learning_rate": 8.780078018505172e-06,
+      "loss": 0.227,
+      "step": 2735
+    },
+    {
+      "epoch": 0.9136750709634329,
+      "grad_norm": 0.5610844903221539,
+      "learning_rate": 8.778805717212853e-06,
+      "loss": 0.2202,
+      "step": 2736
+    },
+    {
+      "epoch": 0.9140090165303055,
+      "grad_norm": 0.44429928394757645,
+      "learning_rate": 8.777532845087434e-06,
+      "loss": 0.2224,
+      "step": 2737
+    },
+    {
+      "epoch": 0.9143429620971781,
+      "grad_norm": 0.5344338868572158,
+      "learning_rate": 8.776259402321201e-06,
+      "loss": 0.235,
+      "step": 2738
+    },
+    {
+      "epoch": 0.9146769076640507,
+      "grad_norm": 0.5063152059042075,
+      "learning_rate": 8.774985389106521e-06,
+      "loss": 0.2272,
+      "step": 2739
+    },
+    {
+      "epoch": 0.9150108532309233,
+      "grad_norm": 0.5051461254697261,
+      "learning_rate": 8.77371080563585e-06,
+      "loss": 0.2251,
+      "step": 2740
+    },
+    {
+      "epoch": 0.915344798797796,
+      "grad_norm": 0.41201433720845737,
+      "learning_rate": 8.772435652101726e-06,
+      "loss": 0.2084,
+      "step": 2741
+    },
+    {
+      "epoch": 0.9156787443646686,
+      "grad_norm": 0.42455946149967577,
+      "learning_rate": 8.771159928696779e-06,
+      "loss": 0.2091,
+      "step": 2742
+    },
+    {
+      "epoch": 0.9160126899315412,
+      "grad_norm": 0.4605627342858186,
+      "learning_rate": 8.76988363561372e-06,
+      "loss": 0.2197,
+      "step": 2743
+    },
+    {
+      "epoch": 0.9163466354984138,
+      "grad_norm": 0.482110449293083,
+      "learning_rate": 8.76860677304535e-06,
+      "loss": 0.216,
+      "step": 2744
+    },
+    {
+      "epoch": 0.9166805810652864,
+      "grad_norm": 0.46195236737899087,
+      "learning_rate": 8.767329341184552e-06,
+      "loss": 0.2331,
+      "step": 2745
+    },
+    {
+      "epoch": 0.917014526632159,
+      "grad_norm": 0.4413188092360609,
+      "learning_rate": 8.766051340224297e-06,
+      "loss": 0.2171,
+      "step": 2746
+    },
+    {
+      "epoch": 0.9173484721990316,
+      "grad_norm": 0.48280257812503025,
+      "learning_rate": 8.764772770357646e-06,
+      "loss": 0.2182,
+      "step": 2747
+    },
+    {
+      "epoch": 0.9176824177659042,
+      "grad_norm": 0.4210285641166794,
+      "learning_rate": 8.763493631777738e-06,
+      "loss": 0.2044,
+      "step": 2748
+    },
+    {
+      "epoch": 0.9180163633327768,
+      "grad_norm": 0.4496036936278124,
+      "learning_rate": 8.762213924677802e-06,
+      "loss": 0.2224,
+      "step": 2749
+    },
+    {
+      "epoch": 0.9183503088996493,
+      "grad_norm": 0.4402813014352695,
+      "learning_rate": 8.760933649251155e-06,
+      "loss": 0.2157,
+      "step": 2750
+    },
+    {
+      "epoch": 0.9186842544665219,
+      "grad_norm": 0.43235518297213915,
+      "learning_rate": 8.759652805691197e-06,
+      "loss": 0.221,
+      "step": 2751
+    },
+    {
+      "epoch": 0.9190182000333945,
+      "grad_norm": 0.4424479020307994,
+      "learning_rate": 8.758371394191415e-06,
+      "loss": 0.22,
+      "step": 2752
+    },
+    {
+      "epoch": 0.9193521456002671,
+      "grad_norm": 0.4281531056346567,
+      "learning_rate": 8.75708941494538e-06,
+      "loss": 0.2127,
+      "step": 2753
+    },
+    {
+      "epoch": 0.9196860911671397,
+      "grad_norm": 0.48598727252634466,
+      "learning_rate": 8.75580686814675e-06,
+      "loss": 0.2201,
+      "step": 2754
+    },
+    {
+      "epoch": 0.9200200367340123,
+      "grad_norm": 0.4900282146647412,
+      "learning_rate": 8.75452375398927e-06,
+      "loss": 0.2202,
+      "step": 2755
+    },
+    {
+      "epoch": 0.9203539823008849,
+      "grad_norm": 0.41487723575509605,
+      "learning_rate": 8.753240072666769e-06,
+      "loss": 0.2172,
+      "step": 2756
+    },
+    {
+      "epoch": 0.9206879278677575,
+      "grad_norm": 0.46744752038473614,
+      "learning_rate": 8.751955824373161e-06,
+      "loss": 0.2299,
+      "step": 2757
+    },
+    {
+      "epoch": 0.9210218734346302,
+      "grad_norm": 0.3913659048757199,
+      "learning_rate": 8.750671009302448e-06,
+      "loss": 0.2142,
+      "step": 2758
+    },
+    {
+      "epoch": 0.9213558190015028,
+      "grad_norm": 0.3858840470760089,
+      "learning_rate": 8.749385627648717e-06,
+      "loss": 0.1974,
+      "step": 2759
+    },
+    {
+      "epoch": 0.9216897645683754,
+      "grad_norm": 0.46012895388328423,
+      "learning_rate": 8.748099679606139e-06,
+      "loss": 0.2263,
+      "step": 2760
+    },
+    {
+      "epoch": 0.922023710135248,
+      "grad_norm": 0.5056054844807831,
+      "learning_rate": 8.746813165368973e-06,
+      "loss": 0.2164,
+      "step": 2761
+    },
+    {
+      "epoch": 0.9223576557021206,
+      "grad_norm": 0.4099051015952098,
+      "learning_rate": 8.745526085131559e-06,
+      "loss": 0.2087,
+      "step": 2762
+    },
+    {
+      "epoch": 0.9226916012689932,
+      "grad_norm": 0.44267862491947235,
+      "learning_rate": 8.744238439088328e-06,
+      "loss": 0.2143,
+      "step": 2763
+    },
+    {
+      "epoch": 0.9230255468358658,
+      "grad_norm": 0.5002499141140306,
+      "learning_rate": 8.742950227433795e-06,
+      "loss": 0.227,
+      "step": 2764
+    },
+    {
+      "epoch": 0.9233594924027384,
+      "grad_norm": 0.4609375168550936,
+      "learning_rate": 8.741661450362559e-06,
+      "loss": 0.222,
+      "step": 2765
+    },
+    {
+      "epoch": 0.923693437969611,
+      "grad_norm": 0.40652027628765436,
+      "learning_rate": 8.740372108069304e-06,
+      "loss": 0.2063,
+      "step": 2766
+    },
+    {
+      "epoch": 0.9240273835364835,
+      "grad_norm": 0.46845775301622733,
+      "learning_rate": 8.739082200748799e-06,
+      "loss": 0.2056,
+      "step": 2767
+    },
+    {
+      "epoch": 0.9243613291033561,
+      "grad_norm": 0.43583945959938114,
+      "learning_rate": 8.737791728595903e-06,
+      "loss": 0.221,
+      "step": 2768
+    },
+    {
+      "epoch": 0.9246952746702287,
+      "grad_norm": 0.4211490574278276,
+      "learning_rate": 8.736500691805554e-06,
+      "loss": 0.2249,
+      "step": 2769
+    },
+    {
+      "epoch": 0.9250292202371013,
+      "grad_norm": 0.4928325266258507,
+      "learning_rate": 8.73520909057278e-06,
+      "loss": 0.2268,
+      "step": 2770
+    },
+    {
+      "epoch": 0.9253631658039739,
+      "grad_norm": 0.47059734345795795,
+      "learning_rate": 8.733916925092691e-06,
+      "loss": 0.231,
+      "step": 2771
+    },
+    {
+      "epoch": 0.9256971113708465,
+      "grad_norm": 0.45417002632161574,
+      "learning_rate": 8.732624195560487e-06,
+      "loss": 0.2389,
+      "step": 2772
+    },
+    {
+      "epoch": 0.9260310569377191,
+      "grad_norm": 0.3945646807349342,
+      "learning_rate": 8.731330902171447e-06,
+      "loss": 0.2168,
+      "step": 2773
+    },
+    {
+      "epoch": 0.9263650025045918,
+      "grad_norm": 0.44248145973303443,
+      "learning_rate": 8.730037045120941e-06,
+      "loss": 0.2238,
+      "step": 2774
+    },
+    {
+      "epoch": 0.9266989480714644,
+      "grad_norm": 0.4674819994664436,
+      "learning_rate": 8.728742624604418e-06,
+      "loss": 0.2341,
+      "step": 2775
+    },
+    {
+      "epoch": 0.927032893638337,
+      "grad_norm": 0.64231489409082,
+      "learning_rate": 8.727447640817417e-06,
+      "loss": 0.2133,
+      "step": 2776
+    },
+    {
+      "epoch": 0.9273668392052096,
+      "grad_norm": 0.44146865573432903,
+      "learning_rate": 8.726152093955561e-06,
+      "loss": 0.2104,
+      "step": 2777
+    },
+    {
+      "epoch": 0.9277007847720822,
+      "grad_norm": 0.43933661617710457,
+      "learning_rate": 8.724855984214558e-06,
+      "loss": 0.2321,
+      "step": 2778
+    },
+    {
+      "epoch": 0.9280347303389548,
+      "grad_norm": 0.46754943764880913,
+      "learning_rate": 8.723559311790197e-06,
+      "loss": 0.227,
+      "step": 2779
+    },
+    {
+      "epoch": 0.9283686759058274,
+      "grad_norm": 0.41188188249517227,
+      "learning_rate": 8.722262076878361e-06,
+      "loss": 0.2002,
+      "step": 2780
+    },
+    {
+      "epoch": 0.9287026214727,
+      "grad_norm": 0.4430354887476343,
+      "learning_rate": 8.720964279675009e-06,
+      "loss": 0.2115,
+      "step": 2781
+    },
+    {
+      "epoch": 0.9290365670395726,
+      "grad_norm": 0.4164609143001292,
+      "learning_rate": 8.71966592037619e-06,
+      "loss": 0.2121,
+      "step": 2782
+    },
+    {
+      "epoch": 0.9293705126064451,
+      "grad_norm": 0.42842924540161714,
+      "learning_rate": 8.718366999178037e-06,
+      "loss": 0.2219,
+      "step": 2783
+    },
+    {
+      "epoch": 0.9297044581733177,
+      "grad_norm": 0.46668660218796776,
+      "learning_rate": 8.717067516276764e-06,
+      "loss": 0.2293,
+      "step": 2784
+    },
+    {
+      "epoch": 0.9300384037401903,
+      "grad_norm": 0.4467610399518897,
+      "learning_rate": 8.715767471868679e-06,
+      "loss": 0.2166,
+      "step": 2785
+    },
+    {
+      "epoch": 0.9303723493070629,
+      "grad_norm": 0.5306058118803246,
+      "learning_rate": 8.714466866150162e-06,
+      "loss": 0.2515,
+      "step": 2786
+    },
+    {
+      "epoch": 0.9307062948739355,
+      "grad_norm": 0.4538833004379203,
+      "learning_rate": 8.71316569931769e-06,
+      "loss": 0.2245,
+      "step": 2787
+    },
+    {
+      "epoch": 0.9310402404408081,
+      "grad_norm": 0.44120044003214,
+      "learning_rate": 8.71186397156782e-06,
+      "loss": 0.2191,
+      "step": 2788
+    },
+    {
+      "epoch": 0.9313741860076807,
+      "grad_norm": 0.4610836615076066,
+      "learning_rate": 8.710561683097189e-06,
+      "loss": 0.2178,
+      "step": 2789
+    },
+    {
+      "epoch": 0.9317081315745533,
+      "grad_norm": 0.4363676340228982,
+      "learning_rate": 8.709258834102525e-06,
+      "loss": 0.2242,
+      "step": 2790
+    },
+    {
+      "epoch": 0.932042077141426,
+      "grad_norm": 0.6756307435936366,
+      "learning_rate": 8.70795542478064e-06,
+      "loss": 0.231,
+      "step": 2791
+    },
+    {
+      "epoch": 0.9323760227082986,
+      "grad_norm": 0.5041996579273476,
+      "learning_rate": 8.706651455328427e-06,
+      "loss": 0.2163,
+      "step": 2792
+    },
+    {
+      "epoch": 0.9327099682751712,
+      "grad_norm": 0.4698362067299536,
+      "learning_rate": 8.70534692594287e-06,
+      "loss": 0.2216,
+      "step": 2793
+    },
+    {
+      "epoch": 0.9330439138420438,
+      "grad_norm": 0.4820217295468483,
+      "learning_rate": 8.704041836821029e-06,
+      "loss": 0.2287,
+      "step": 2794
+    },
+    {
+      "epoch": 0.9333778594089164,
+      "grad_norm": 0.4577816146622304,
+      "learning_rate": 8.702736188160055e-06,
+      "loss": 0.2085,
+      "step": 2795
+    },
+    {
+      "epoch": 0.933711804975789,
+      "grad_norm": 0.4804066031180158,
+      "learning_rate": 8.70142998015718e-06,
+      "loss": 0.2128,
+      "step": 2796
+    },
+    {
+      "epoch": 0.9340457505426616,
+      "grad_norm": 0.48646393253865444,
+      "learning_rate": 8.700123213009726e-06,
+      "loss": 0.2315,
+      "step": 2797
+    },
+    {
+      "epoch": 0.9343796961095342,
+      "grad_norm": 0.4392885717258637,
+      "learning_rate": 8.698815886915094e-06,
+      "loss": 0.2372,
+      "step": 2798
+    },
+    {
+      "epoch": 0.9347136416764067,
+      "grad_norm": 0.4573456253338256,
+      "learning_rate": 8.697508002070766e-06,
+      "loss": 0.2237,
+      "step": 2799
+    },
+    {
+      "epoch": 0.9350475872432793,
+      "grad_norm": 1.0167289235959724,
+      "learning_rate": 8.696199558674321e-06,
+      "loss": 0.2282,
+      "step": 2800
+    },
+    {
+      "epoch": 0.9353815328101519,
+      "grad_norm": 0.49862618680445076,
+      "learning_rate": 8.69489055692341e-06,
+      "loss": 0.2191,
+      "step": 2801
+    },
+    {
+      "epoch": 0.9357154783770245,
+      "grad_norm": 0.4071314628849528,
+      "learning_rate": 8.693580997015775e-06,
+      "loss": 0.2145,
+      "step": 2802
+    },
+    {
+      "epoch": 0.9360494239438971,
+      "grad_norm": 0.48212789403115414,
+      "learning_rate": 8.692270879149241e-06,
+      "loss": 0.2264,
+      "step": 2803
+    },
+    {
+      "epoch": 0.9363833695107697,
+      "grad_norm": 0.46241253052454173,
+      "learning_rate": 8.690960203521713e-06,
+      "loss": 0.2084,
+      "step": 2804
+    },
+    {
+      "epoch": 0.9367173150776423,
+      "grad_norm": 0.5117131305242669,
+      "learning_rate": 8.689648970331188e-06,
+      "loss": 0.2498,
+      "step": 2805
+    },
+    {
+      "epoch": 0.9370512606445149,
+      "grad_norm": 0.488625905932195,
+      "learning_rate": 8.68833717977574e-06,
+      "loss": 0.205,
+      "step": 2806
+    },
+    {
+      "epoch": 0.9373852062113875,
+      "grad_norm": 0.46934650535365496,
+      "learning_rate": 8.687024832053534e-06,
+      "loss": 0.209,
+      "step": 2807
+    },
+    {
+      "epoch": 0.9377191517782602,
+      "grad_norm": 0.4876878801995502,
+      "learning_rate": 8.685711927362815e-06,
+      "loss": 0.2166,
+      "step": 2808
+    },
+    {
+      "epoch": 0.9380530973451328,
+      "grad_norm": 0.47415720928263255,
+      "learning_rate": 8.68439846590191e-06,
+      "loss": 0.2019,
+      "step": 2809
+    },
+    {
+      "epoch": 0.9383870429120054,
+      "grad_norm": 0.5071405923836091,
+      "learning_rate": 8.683084447869234e-06,
+      "loss": 0.2193,
+      "step": 2810
+    },
+    {
+      "epoch": 0.938720988478878,
+      "grad_norm": 0.5342801311116051,
+      "learning_rate": 8.681769873463286e-06,
+      "loss": 0.2238,
+      "step": 2811
+    },
+    {
+      "epoch": 0.9390549340457506,
+      "grad_norm": 0.4967761833386368,
+      "learning_rate": 8.680454742882647e-06,
+      "loss": 0.2319,
+      "step": 2812
+    },
+    {
+      "epoch": 0.9393888796126232,
+      "grad_norm": 0.45077129590302223,
+      "learning_rate": 8.679139056325983e-06,
+      "loss": 0.2295,
+      "step": 2813
+    },
+    {
+      "epoch": 0.9397228251794958,
+      "grad_norm": 0.4979759329992662,
+      "learning_rate": 8.677822813992046e-06,
+      "loss": 0.2261,
+      "step": 2814
+    },
+    {
+      "epoch": 0.9400567707463683,
+      "grad_norm": 0.4950847343176679,
+      "learning_rate": 8.676506016079664e-06,
+      "loss": 0.2238,
+      "step": 2815
+    },
+    {
+      "epoch": 0.9403907163132409,
+      "grad_norm": 0.40708194153685967,
+      "learning_rate": 8.675188662787762e-06,
+      "loss": 0.2187,
+      "step": 2816
+    },
+    {
+      "epoch": 0.9407246618801135,
+      "grad_norm": 0.5521110478916056,
+      "learning_rate": 8.673870754315336e-06,
+      "loss": 0.2053,
+      "step": 2817
+    },
+    {
+      "epoch": 0.9410586074469861,
+      "grad_norm": 0.4442733755427678,
+      "learning_rate": 8.672552290861478e-06,
+      "loss": 0.2164,
+      "step": 2818
+    },
+    {
+      "epoch": 0.9413925530138587,
+      "grad_norm": 0.4553102853972935,
+      "learning_rate": 8.67123327262535e-06,
+      "loss": 0.2169,
+      "step": 2819
+    },
+    {
+      "epoch": 0.9417264985807313,
+      "grad_norm": 0.4306856509162607,
+      "learning_rate": 8.669913699806209e-06,
+      "loss": 0.2096,
+      "step": 2820
+    },
+    {
+      "epoch": 0.9420604441476039,
+      "grad_norm": 0.5401958009078948,
+      "learning_rate": 8.668593572603394e-06,
+      "loss": 0.2282,
+      "step": 2821
+    },
+    {
+      "epoch": 0.9423943897144765,
+      "grad_norm": 0.5107141764908304,
+      "learning_rate": 8.667272891216323e-06,
+      "loss": 0.2359,
+      "step": 2822
+    },
+    {
+      "epoch": 0.9427283352813491,
+      "grad_norm": 0.45683098172915115,
+      "learning_rate": 8.6659516558445e-06,
+      "loss": 0.2207,
+      "step": 2823
+    },
+    {
+      "epoch": 0.9430622808482217,
+      "grad_norm": 0.5311817625144637,
+      "learning_rate": 8.664629866687514e-06,
+      "loss": 0.2377,
+      "step": 2824
+    },
+    {
+      "epoch": 0.9433962264150944,
+      "grad_norm": 0.441871608821732,
+      "learning_rate": 8.663307523945038e-06,
+      "loss": 0.2185,
+      "step": 2825
+    },
+    {
+      "epoch": 0.943730171981967,
+      "grad_norm": 0.4506561928586293,
+      "learning_rate": 8.661984627816827e-06,
+      "loss": 0.2136,
+      "step": 2826
+    },
+    {
+      "epoch": 0.9440641175488396,
+      "grad_norm": 0.40555871123613024,
+      "learning_rate": 8.660661178502719e-06,
+      "loss": 0.2094,
+      "step": 2827
+    },
+    {
+      "epoch": 0.9443980631157122,
+      "grad_norm": 0.48122612843516865,
+      "learning_rate": 8.659337176202636e-06,
+      "loss": 0.2177,
+      "step": 2828
+    },
+    {
+      "epoch": 0.9447320086825848,
+      "grad_norm": 0.493730557455204,
+      "learning_rate": 8.658012621116585e-06,
+      "loss": 0.2033,
+      "step": 2829
+    },
+    {
+      "epoch": 0.9450659542494574,
+      "grad_norm": 0.39913730987982116,
+      "learning_rate": 8.656687513444656e-06,
+      "loss": 0.1957,
+      "step": 2830
+    },
+    {
+      "epoch": 0.94539989981633,
+      "grad_norm": 0.43231231966574224,
+      "learning_rate": 8.655361853387024e-06,
+      "loss": 0.2212,
+      "step": 2831
+    },
+    {
+      "epoch": 0.9457338453832025,
+      "grad_norm": 0.5034429330788713,
+      "learning_rate": 8.654035641143944e-06,
+      "loss": 0.2266,
+      "step": 2832
+    },
+    {
+      "epoch": 0.9460677909500751,
+      "grad_norm": 0.5248570196879081,
+      "learning_rate": 8.652708876915752e-06,
+      "loss": 0.2183,
+      "step": 2833
+    },
+    {
+      "epoch": 0.9464017365169477,
+      "grad_norm": 0.47638029426551504,
+      "learning_rate": 8.651381560902876e-06,
+      "loss": 0.2257,
+      "step": 2834
+    },
+    {
+      "epoch": 0.9467356820838203,
+      "grad_norm": 0.4872575546717003,
+      "learning_rate": 8.650053693305824e-06,
+      "loss": 0.2371,
+      "step": 2835
+    },
+    {
+      "epoch": 0.9470696276506929,
+      "grad_norm": 0.535075069115959,
+      "learning_rate": 8.648725274325182e-06,
+      "loss": 0.2206,
+      "step": 2836
+    },
+    {
+      "epoch": 0.9474035732175655,
+      "grad_norm": 0.48011848827739007,
+      "learning_rate": 8.647396304161625e-06,
+      "loss": 0.2204,
+      "step": 2837
+    },
+    {
+      "epoch": 0.9477375187844381,
+      "grad_norm": 0.43107044927809024,
+      "learning_rate": 8.64606678301591e-06,
+      "loss": 0.2143,
+      "step": 2838
+    },
+    {
+      "epoch": 0.9480714643513107,
+      "grad_norm": 0.48263476741763794,
+      "learning_rate": 8.644736711088874e-06,
+      "loss": 0.2128,
+      "step": 2839
+    },
+    {
+      "epoch": 0.9484054099181833,
+      "grad_norm": 0.4227729549713639,
+      "learning_rate": 8.643406088581446e-06,
+      "loss": 0.2181,
+      "step": 2840
+    },
+    {
+      "epoch": 0.948739355485056,
+      "grad_norm": 0.4423800504922258,
+      "learning_rate": 8.642074915694626e-06,
+      "loss": 0.2181,
+      "step": 2841
+    },
+    {
+      "epoch": 0.9490733010519286,
+      "grad_norm": 0.49729486477624335,
+      "learning_rate": 8.640743192629507e-06,
+      "loss": 0.2196,
+      "step": 2842
+    },
+    {
+      "epoch": 0.9494072466188012,
+      "grad_norm": 0.48595434515596525,
+      "learning_rate": 8.63941091958726e-06,
+      "loss": 0.2181,
+      "step": 2843
+    },
+    {
+      "epoch": 0.9497411921856738,
+      "grad_norm": 0.5001238407946756,
+      "learning_rate": 8.638078096769141e-06,
+      "loss": 0.2209,
+      "step": 2844
+    },
+    {
+      "epoch": 0.9500751377525464,
+      "grad_norm": 0.4720041136115112,
+      "learning_rate": 8.636744724376488e-06,
+      "loss": 0.2214,
+      "step": 2845
+    },
+    {
+      "epoch": 0.950409083319419,
+      "grad_norm": 0.48399406369761006,
+      "learning_rate": 8.635410802610724e-06,
+      "loss": 0.2292,
+      "step": 2846
+    },
+    {
+      "epoch": 0.9507430288862916,
+      "grad_norm": 0.4332190126394425,
+      "learning_rate": 8.634076331673354e-06,
+      "loss": 0.2182,
+      "step": 2847
+    },
+    {
+      "epoch": 0.9510769744531641,
+      "grad_norm": 0.47377715160398876,
+      "learning_rate": 8.632741311765962e-06,
+      "loss": 0.2217,
+      "step": 2848
+    },
+    {
+      "epoch": 0.9514109200200367,
+      "grad_norm": 0.4803537971441186,
+      "learning_rate": 8.631405743090223e-06,
+      "loss": 0.2224,
+      "step": 2849
+    },
+    {
+      "epoch": 0.9517448655869093,
+      "grad_norm": 0.44465165141275165,
+      "learning_rate": 8.630069625847885e-06,
+      "loss": 0.212,
+      "step": 2850
+    },
+    {
+      "epoch": 0.9520788111537819,
+      "grad_norm": 0.4189509028928389,
+      "learning_rate": 8.628732960240788e-06,
+      "loss": 0.2201,
+      "step": 2851
+    },
+    {
+      "epoch": 0.9524127567206545,
+      "grad_norm": 0.4072448497585389,
+      "learning_rate": 8.627395746470852e-06,
+      "loss": 0.1999,
+      "step": 2852
+    },
+    {
+      "epoch": 0.9527467022875271,
+      "grad_norm": 0.5302490534823262,
+      "learning_rate": 8.626057984740077e-06,
+      "loss": 0.2235,
+      "step": 2853
+    },
+    {
+      "epoch": 0.9530806478543997,
+      "grad_norm": 0.4770176390794186,
+      "learning_rate": 8.624719675250547e-06,
+      "loss": 0.2092,
+      "step": 2854
+    },
+    {
+      "epoch": 0.9534145934212723,
+      "grad_norm": 0.4649548910468379,
+      "learning_rate": 8.623380818204431e-06,
+      "loss": 0.2206,
+      "step": 2855
+    },
+    {
+      "epoch": 0.9537485389881449,
+      "grad_norm": 0.41794317440850726,
+      "learning_rate": 8.622041413803979e-06,
+      "loss": 0.2299,
+      "step": 2856
+    },
+    {
+      "epoch": 0.9540824845550175,
+      "grad_norm": 0.4270595046761071,
+      "learning_rate": 8.620701462251522e-06,
+      "loss": 0.2274,
+      "step": 2857
+    },
+    {
+      "epoch": 0.9544164301218901,
+      "grad_norm": 0.5605112243213927,
+      "learning_rate": 8.619360963749478e-06,
+      "loss": 0.2366,
+      "step": 2858
+    },
+    {
+      "epoch": 0.9547503756887628,
+      "grad_norm": 0.4315029266321909,
+      "learning_rate": 8.618019918500342e-06,
+      "loss": 0.2155,
+      "step": 2859
+    },
+    {
+      "epoch": 0.9550843212556354,
+      "grad_norm": 0.41819570091246855,
+      "learning_rate": 8.616678326706698e-06,
+      "loss": 0.2032,
+      "step": 2860
+    },
+    {
+      "epoch": 0.955418266822508,
+      "grad_norm": 0.4720800653986111,
+      "learning_rate": 8.615336188571208e-06,
+      "loss": 0.2106,
+      "step": 2861
+    },
+    {
+      "epoch": 0.9557522123893806,
+      "grad_norm": 0.4336533842521673,
+      "learning_rate": 8.613993504296617e-06,
+      "loss": 0.2106,
+      "step": 2862
+    },
+    {
+      "epoch": 0.9560861579562532,
+      "grad_norm": 0.4842098122111017,
+      "learning_rate": 8.612650274085755e-06,
+      "loss": 0.2153,
+      "step": 2863
+    },
+    {
+      "epoch": 0.9564201035231257,
+      "grad_norm": 0.46478104433885903,
+      "learning_rate": 8.61130649814153e-06,
+      "loss": 0.2172,
+      "step": 2864
+    },
+    {
+      "epoch": 0.9567540490899983,
+      "grad_norm": 0.3995110999710848,
+      "learning_rate": 8.609962176666936e-06,
+      "loss": 0.2008,
+      "step": 2865
+    },
+    {
+      "epoch": 0.9570879946568709,
+      "grad_norm": 0.4619066958368434,
+      "learning_rate": 8.608617309865051e-06,
+      "loss": 0.2247,
+      "step": 2866
+    },
+    {
+      "epoch": 0.9574219402237435,
+      "grad_norm": 0.4772269876791035,
+      "learning_rate": 8.60727189793903e-06,
+      "loss": 0.2238,
+      "step": 2867
+    },
+    {
+      "epoch": 0.9577558857906161,
+      "grad_norm": 0.45025507277873233,
+      "learning_rate": 8.605925941092114e-06,
+      "loss": 0.2321,
+      "step": 2868
+    },
+    {
+      "epoch": 0.9580898313574887,
+      "grad_norm": 0.4240250807538226,
+      "learning_rate": 8.604579439527627e-06,
+      "loss": 0.2221,
+      "step": 2869
+    },
+    {
+      "epoch": 0.9584237769243613,
+      "grad_norm": 0.4467352359560645,
+      "learning_rate": 8.603232393448974e-06,
+      "loss": 0.2115,
+      "step": 2870
+    },
+    {
+      "epoch": 0.9587577224912339,
+      "grad_norm": 0.43281845584862416,
+      "learning_rate": 8.601884803059641e-06,
+      "loss": 0.2084,
+      "step": 2871
+    },
+    {
+      "epoch": 0.9590916680581065,
+      "grad_norm": 0.43146109431192975,
+      "learning_rate": 8.600536668563197e-06,
+      "loss": 0.2185,
+      "step": 2872
+    },
+    {
+      "epoch": 0.9594256136249791,
+      "grad_norm": 0.47249516324982266,
+      "learning_rate": 8.599187990163296e-06,
+      "loss": 0.2221,
+      "step": 2873
+    },
+    {
+      "epoch": 0.9597595591918517,
+      "grad_norm": 0.4698332572735109,
+      "learning_rate": 8.597838768063667e-06,
+      "loss": 0.23,
+      "step": 2874
+    },
+    {
+      "epoch": 0.9600935047587243,
+      "grad_norm": 0.5412304324762713,
+      "learning_rate": 8.596489002468132e-06,
+      "loss": 0.2245,
+      "step": 2875
+    },
+    {
+      "epoch": 0.960427450325597,
+      "grad_norm": 0.44637589874465433,
+      "learning_rate": 8.595138693580583e-06,
+      "loss": 0.2233,
+      "step": 2876
+    },
+    {
+      "epoch": 0.9607613958924696,
+      "grad_norm": 0.43336083402117387,
+      "learning_rate": 8.593787841605004e-06,
+      "loss": 0.2048,
+      "step": 2877
+    },
+    {
+      "epoch": 0.9610953414593422,
+      "grad_norm": 0.4507857608921773,
+      "learning_rate": 8.592436446745457e-06,
+      "loss": 0.2187,
+      "step": 2878
+    },
+    {
+      "epoch": 0.9614292870262148,
+      "grad_norm": 0.4137810248936209,
+      "learning_rate": 8.591084509206085e-06,
+      "loss": 0.2288,
+      "step": 2879
+    },
+    {
+      "epoch": 0.9617632325930874,
+      "grad_norm": 0.4381991589239642,
+      "learning_rate": 8.589732029191113e-06,
+      "loss": 0.2156,
+      "step": 2880
+    },
+    {
+      "epoch": 0.9620971781599599,
+      "grad_norm": 0.46254816718933667,
+      "learning_rate": 8.588379006904852e-06,
+      "loss": 0.2345,
+      "step": 2881
+    },
+    {
+      "epoch": 0.9624311237268325,
+      "grad_norm": 0.447052657136721,
+      "learning_rate": 8.587025442551689e-06,
+      "loss": 0.211,
+      "step": 2882
+    },
+    {
+      "epoch": 0.9627650692937051,
+      "grad_norm": 0.4220565998823512,
+      "learning_rate": 8.585671336336096e-06,
+      "loss": 0.2142,
+      "step": 2883
+    },
+    {
+      "epoch": 0.9630990148605777,
+      "grad_norm": 0.46402238237283716,
+      "learning_rate": 8.58431668846263e-06,
+      "loss": 0.2124,
+      "step": 2884
+    },
+    {
+      "epoch": 0.9634329604274503,
+      "grad_norm": 0.8148299624701707,
+      "learning_rate": 8.582961499135925e-06,
+      "loss": 0.2138,
+      "step": 2885
+    },
+    {
+      "epoch": 0.9637669059943229,
+      "grad_norm": 0.4218423758746241,
+      "learning_rate": 8.581605768560694e-06,
+      "loss": 0.2237,
+      "step": 2886
+    },
+    {
+      "epoch": 0.9641008515611955,
+      "grad_norm": 0.42468992711883435,
+      "learning_rate": 8.580249496941742e-06,
+      "loss": 0.2222,
+      "step": 2887
+    },
+    {
+      "epoch": 0.9644347971280681,
+      "grad_norm": 0.45231316231310115,
+      "learning_rate": 8.578892684483947e-06,
+      "loss": 0.2304,
+      "step": 2888
+    },
+    {
+      "epoch": 0.9647687426949407,
+      "grad_norm": 0.39665581763729213,
+      "learning_rate": 8.577535331392272e-06,
+      "loss": 0.2061,
+      "step": 2889
+    },
+    {
+      "epoch": 0.9651026882618133,
+      "grad_norm": 0.42218622196372374,
+      "learning_rate": 8.57617743787176e-06,
+      "loss": 0.214,
+      "step": 2890
+    },
+    {
+      "epoch": 0.9654366338286859,
+      "grad_norm": 0.4327387439898741,
+      "learning_rate": 8.574819004127539e-06,
+      "loss": 0.2165,
+      "step": 2891
+    },
+    {
+      "epoch": 0.9657705793955585,
+      "grad_norm": 0.4368794278929328,
+      "learning_rate": 8.573460030364816e-06,
+      "loss": 0.2057,
+      "step": 2892
+    },
+    {
+      "epoch": 0.9661045249624312,
+      "grad_norm": 0.46821955032272594,
+      "learning_rate": 8.572100516788878e-06,
+      "loss": 0.2122,
+      "step": 2893
+    },
+    {
+      "epoch": 0.9664384705293038,
+      "grad_norm": 0.5024454235492738,
+      "learning_rate": 8.570740463605096e-06,
+      "loss": 0.2301,
+      "step": 2894
+    },
+    {
+      "epoch": 0.9667724160961764,
+      "grad_norm": 0.42826897322512725,
+      "learning_rate": 8.569379871018925e-06,
+      "loss": 0.2359,
+      "step": 2895
+    },
+    {
+      "epoch": 0.967106361663049,
+      "grad_norm": 0.447750666872138,
+      "learning_rate": 8.568018739235895e-06,
+      "loss": 0.1986,
+      "step": 2896
+    },
+    {
+      "epoch": 0.9674403072299215,
+      "grad_norm": 0.45508421449926406,
+      "learning_rate": 8.566657068461624e-06,
+      "loss": 0.2089,
+      "step": 2897
+    },
+    {
+      "epoch": 0.9677742527967941,
+      "grad_norm": 0.4895779825553509,
+      "learning_rate": 8.565294858901804e-06,
+      "loss": 0.2276,
+      "step": 2898
+    },
+    {
+      "epoch": 0.9681081983636667,
+      "grad_norm": 0.4114005627874931,
+      "learning_rate": 8.563932110762218e-06,
+      "loss": 0.2125,
+      "step": 2899
+    },
+    {
+      "epoch": 0.9684421439305393,
+      "grad_norm": 0.44863986840084646,
+      "learning_rate": 8.562568824248722e-06,
+      "loss": 0.2294,
+      "step": 2900
+    },
+    {
+      "epoch": 0.9687760894974119,
+      "grad_norm": 0.45066913390824037,
+      "learning_rate": 8.561204999567258e-06,
+      "loss": 0.2125,
+      "step": 2901
+    },
+    {
+      "epoch": 0.9691100350642845,
+      "grad_norm": 0.7353204759248914,
+      "learning_rate": 8.559840636923845e-06,
+      "loss": 0.2262,
+      "step": 2902
+    },
+    {
+      "epoch": 0.9694439806311571,
+      "grad_norm": 0.4509345787133013,
+      "learning_rate": 8.55847573652459e-06,
+      "loss": 0.2242,
+      "step": 2903
+    },
+    {
+      "epoch": 0.9697779261980297,
+      "grad_norm": 0.41027552893332203,
+      "learning_rate": 8.557110298575674e-06,
+      "loss": 0.1956,
+      "step": 2904
+    },
+    {
+      "epoch": 0.9701118717649023,
+      "grad_norm": 0.3886765860123989,
+      "learning_rate": 8.555744323283364e-06,
+      "loss": 0.2062,
+      "step": 2905
+    },
+    {
+      "epoch": 0.9704458173317749,
+      "grad_norm": 0.4099609507610317,
+      "learning_rate": 8.554377810854006e-06,
+      "loss": 0.2229,
+      "step": 2906
+    },
+    {
+      "epoch": 0.9707797628986475,
+      "grad_norm": 0.4514135763925288,
+      "learning_rate": 8.553010761494029e-06,
+      "loss": 0.205,
+      "step": 2907
+    },
+    {
+      "epoch": 0.9711137084655201,
+      "grad_norm": 0.6008850226168184,
+      "learning_rate": 8.551643175409941e-06,
+      "loss": 0.2156,
+      "step": 2908
+    },
+    {
+      "epoch": 0.9714476540323927,
+      "grad_norm": 0.4142063360961322,
+      "learning_rate": 8.550275052808332e-06,
+      "loss": 0.2333,
+      "step": 2909
+    },
+    {
+      "epoch": 0.9717815995992654,
+      "grad_norm": 0.3950140050648195,
+      "learning_rate": 8.548906393895876e-06,
+      "loss": 0.202,
+      "step": 2910
+    },
+    {
+      "epoch": 0.972115545166138,
+      "grad_norm": 0.4092389771262237,
+      "learning_rate": 8.547537198879318e-06,
+      "loss": 0.221,
+      "step": 2911
+    },
+    {
+      "epoch": 0.9724494907330106,
+      "grad_norm": 0.4256297246287956,
+      "learning_rate": 8.546167467965496e-06,
+      "loss": 0.2134,
+      "step": 2912
+    },
+    {
+      "epoch": 0.9727834362998831,
+      "grad_norm": 0.4442048031336666,
+      "learning_rate": 8.544797201361324e-06,
+      "loss": 0.2213,
+      "step": 2913
+    },
+    {
+      "epoch": 0.9731173818667557,
+      "grad_norm": 0.7029604309389493,
+      "learning_rate": 8.543426399273796e-06,
+      "loss": 0.2621,
+      "step": 2914
+    },
+    {
+      "epoch": 0.9734513274336283,
+      "grad_norm": 0.45508518355245203,
+      "learning_rate": 8.542055061909988e-06,
+      "loss": 0.2271,
+      "step": 2915
+    },
+    {
+      "epoch": 0.9737852730005009,
+      "grad_norm": 0.4386269165717581,
+      "learning_rate": 8.540683189477057e-06,
+      "loss": 0.223,
+      "step": 2916
+    },
+    {
+      "epoch": 0.9741192185673735,
+      "grad_norm": 0.392448221999816,
+      "learning_rate": 8.539310782182238e-06,
+      "loss": 0.2021,
+      "step": 2917
+    },
+    {
+      "epoch": 0.9744531641342461,
+      "grad_norm": 0.42094965559359726,
+      "learning_rate": 8.537937840232853e-06,
+      "loss": 0.2052,
+      "step": 2918
+    },
+    {
+      "epoch": 0.9747871097011187,
+      "grad_norm": 0.454216033457181,
+      "learning_rate": 8.5365643638363e-06,
+      "loss": 0.2272,
+      "step": 2919
+    },
+    {
+      "epoch": 0.9751210552679913,
+      "grad_norm": 0.4225256473955746,
+      "learning_rate": 8.535190353200056e-06,
+      "loss": 0.2199,
+      "step": 2920
+    },
+    {
+      "epoch": 0.9754550008348639,
+      "grad_norm": 0.39002247751830826,
+      "learning_rate": 8.533815808531685e-06,
+      "loss": 0.2065,
+      "step": 2921
+    },
+    {
+      "epoch": 0.9757889464017365,
+      "grad_norm": 0.4266599827086228,
+      "learning_rate": 8.532440730038826e-06,
+      "loss": 0.2187,
+      "step": 2922
+    },
+    {
+      "epoch": 0.9761228919686091,
+      "grad_norm": 0.4205083010157622,
+      "learning_rate": 8.531065117929202e-06,
+      "loss": 0.2215,
+      "step": 2923
+    },
+    {
+      "epoch": 0.9764568375354817,
+      "grad_norm": 0.39958558170677844,
+      "learning_rate": 8.529688972410616e-06,
+      "loss": 0.2024,
+      "step": 2924
+    },
+    {
+      "epoch": 0.9767907831023543,
+      "grad_norm": 0.4258603303840793,
+      "learning_rate": 8.52831229369095e-06,
+      "loss": 0.2219,
+      "step": 2925
+    },
+    {
+      "epoch": 0.977124728669227,
+      "grad_norm": 0.46196673066836863,
+      "learning_rate": 8.526935081978166e-06,
+      "loss": 0.2247,
+      "step": 2926
+    },
+    {
+      "epoch": 0.9774586742360996,
+      "grad_norm": 0.39273292233721263,
+      "learning_rate": 8.52555733748031e-06,
+      "loss": 0.2047,
+      "step": 2927
+    },
+    {
+      "epoch": 0.9777926198029722,
+      "grad_norm": 0.43992124078925776,
+      "learning_rate": 8.524179060405507e-06,
+      "loss": 0.2215,
+      "step": 2928
+    },
+    {
+      "epoch": 0.9781265653698448,
+      "grad_norm": 0.43852567181358093,
+      "learning_rate": 8.52280025096196e-06,
+      "loss": 0.2194,
+      "step": 2929
+    },
+    {
+      "epoch": 0.9784605109367173,
+      "grad_norm": 0.3978127024829244,
+      "learning_rate": 8.521420909357956e-06,
+      "loss": 0.2048,
+      "step": 2930
+    },
+    {
+      "epoch": 0.9787944565035899,
+      "grad_norm": 0.5547762340317196,
+      "learning_rate": 8.52004103580186e-06,
+      "loss": 0.2273,
+      "step": 2931
+    },
+    {
+      "epoch": 0.9791284020704625,
+      "grad_norm": 0.43347864645126716,
+      "learning_rate": 8.51866063050212e-06,
+      "loss": 0.2086,
+      "step": 2932
+    },
+    {
+      "epoch": 0.9794623476373351,
+      "grad_norm": 0.4228219103235732,
+      "learning_rate": 8.51727969366726e-06,
+      "loss": 0.2133,
+      "step": 2933
+    },
+    {
+      "epoch": 0.9797962932042077,
+      "grad_norm": 0.44917306834090986,
+      "learning_rate": 8.515898225505885e-06,
+      "loss": 0.2042,
+      "step": 2934
+    },
+    {
+      "epoch": 0.9801302387710803,
+      "grad_norm": 0.46303289407329606,
+      "learning_rate": 8.514516226226688e-06,
+      "loss": 0.2111,
+      "step": 2935
+    },
+    {
+      "epoch": 0.9804641843379529,
+      "grad_norm": 0.4636036916535876,
+      "learning_rate": 8.513133696038432e-06,
+      "loss": 0.232,
+      "step": 2936
+    },
+    {
+      "epoch": 0.9807981299048255,
+      "grad_norm": 0.3664746812558084,
+      "learning_rate": 8.511750635149965e-06,
+      "loss": 0.1995,
+      "step": 2937
+    },
+    {
+      "epoch": 0.9811320754716981,
+      "grad_norm": 0.3944802889752811,
+      "learning_rate": 8.510367043770213e-06,
+      "loss": 0.1984,
+      "step": 2938
+    },
+    {
+      "epoch": 0.9814660210385707,
+      "grad_norm": 0.444656078405521,
+      "learning_rate": 8.508982922108188e-06,
+      "loss": 0.2209,
+      "step": 2939
+    },
+    {
+      "epoch": 0.9817999666054433,
+      "grad_norm": 0.4562655623442214,
+      "learning_rate": 8.507598270372977e-06,
+      "loss": 0.2312,
+      "step": 2940
+    },
+    {
+      "epoch": 0.9821339121723159,
+      "grad_norm": 0.41457463444511544,
+      "learning_rate": 8.506213088773744e-06,
+      "loss": 0.2142,
+      "step": 2941
+    },
+    {
+      "epoch": 0.9824678577391885,
+      "grad_norm": 0.39229806393694383,
+      "learning_rate": 8.504827377519743e-06,
+      "loss": 0.2077,
+      "step": 2942
+    },
+    {
+      "epoch": 0.9828018033060612,
+      "grad_norm": 0.45229518000097885,
+      "learning_rate": 8.503441136820296e-06,
+      "loss": 0.2343,
+      "step": 2943
+    },
+    {
+      "epoch": 0.9831357488729338,
+      "grad_norm": 0.4069598220008855,
+      "learning_rate": 8.502054366884813e-06,
+      "loss": 0.2062,
+      "step": 2944
+    },
+    {
+      "epoch": 0.9834696944398064,
+      "grad_norm": 0.5008800220194679,
+      "learning_rate": 8.500667067922784e-06,
+      "loss": 0.2183,
+      "step": 2945
+    },
+    {
+      "epoch": 0.9838036400066789,
+      "grad_norm": 0.4430084078638324,
+      "learning_rate": 8.499279240143776e-06,
+      "loss": 0.2272,
+      "step": 2946
+    },
+    {
+      "epoch": 0.9841375855735515,
+      "grad_norm": 0.4201159436753723,
+      "learning_rate": 8.497890883757434e-06,
+      "loss": 0.2145,
+      "step": 2947
+    },
+    {
+      "epoch": 0.9844715311404241,
+      "grad_norm": 0.4087927730039263,
+      "learning_rate": 8.496501998973489e-06,
+      "loss": 0.2124,
+      "step": 2948
+    },
+    {
+      "epoch": 0.9848054767072967,
+      "grad_norm": 0.4203559208003164,
+      "learning_rate": 8.495112586001747e-06,
+      "loss": 0.2237,
+      "step": 2949
+    },
+    {
+      "epoch": 0.9851394222741693,
+      "grad_norm": 0.464817002877942,
+      "learning_rate": 8.493722645052093e-06,
+      "loss": 0.2126,
+      "step": 2950
+    },
+    {
+      "epoch": 0.9854733678410419,
+      "grad_norm": 0.46608580721174564,
+      "learning_rate": 8.4923321763345e-06,
+      "loss": 0.2061,
+      "step": 2951
+    },
+    {
+      "epoch": 0.9858073134079145,
+      "grad_norm": 0.4967627432556802,
+      "learning_rate": 8.490941180059009e-06,
+      "loss": 0.2227,
+      "step": 2952
+    },
+    {
+      "epoch": 0.9861412589747871,
+      "grad_norm": 0.37913666532255536,
+      "learning_rate": 8.489549656435748e-06,
+      "loss": 0.2015,
+      "step": 2953
+    },
+    {
+      "epoch": 0.9864752045416597,
+      "grad_norm": 0.4623920401155906,
+      "learning_rate": 8.488157605674924e-06,
+      "loss": 0.2315,
+      "step": 2954
+    },
+    {
+      "epoch": 0.9868091501085323,
+      "grad_norm": 0.4250267254101623,
+      "learning_rate": 8.486765027986821e-06,
+      "loss": 0.2106,
+      "step": 2955
+    },
+    {
+      "epoch": 0.9871430956754049,
+      "grad_norm": 0.4378715394618819,
+      "learning_rate": 8.485371923581807e-06,
+      "loss": 0.2234,
+      "step": 2956
+    },
+    {
+      "epoch": 0.9874770412422775,
+      "grad_norm": 0.39969124237667886,
+      "learning_rate": 8.483978292670324e-06,
+      "loss": 0.2013,
+      "step": 2957
+    },
+    {
+      "epoch": 0.9878109868091501,
+      "grad_norm": 0.4052628643444376,
+      "learning_rate": 8.482584135462896e-06,
+      "loss": 0.2049,
+      "step": 2958
+    },
+    {
+      "epoch": 0.9881449323760227,
+      "grad_norm": 0.5345263629856472,
+      "learning_rate": 8.48118945217013e-06,
+      "loss": 0.2483,
+      "step": 2959
+    },
+    {
+      "epoch": 0.9884788779428954,
+      "grad_norm": 0.4411343257938832,
+      "learning_rate": 8.479794243002707e-06,
+      "loss": 0.2185,
+      "step": 2960
+    },
+    {
+      "epoch": 0.988812823509768,
+      "grad_norm": 0.41540826133254866,
+      "learning_rate": 8.47839850817139e-06,
+      "loss": 0.2125,
+      "step": 2961
+    },
+    {
+      "epoch": 0.9891467690766405,
+      "grad_norm": 0.44157585681599704,
+      "learning_rate": 8.477002247887024e-06,
+      "loss": 0.1991,
+      "step": 2962
+    },
+    {
+      "epoch": 0.9894807146435131,
+      "grad_norm": 0.4391424439864253,
+      "learning_rate": 8.475605462360525e-06,
+      "loss": 0.2279,
+      "step": 2963
+    },
+    {
+      "epoch": 0.9898146602103857,
+      "grad_norm": 0.5046416290709523,
+      "learning_rate": 8.474208151802898e-06,
+      "loss": 0.2396,
+      "step": 2964
+    },
+    {
+      "epoch": 0.9901486057772583,
+      "grad_norm": 0.4567927616624758,
+      "learning_rate": 8.472810316425223e-06,
+      "loss": 0.2277,
+      "step": 2965
+    },
+    {
+      "epoch": 0.9904825513441309,
+      "grad_norm": 0.4022340376300813,
+      "learning_rate": 8.471411956438657e-06,
+      "loss": 0.2108,
+      "step": 2966
+    },
+    {
+      "epoch": 0.9908164969110035,
+      "grad_norm": 0.4798602783262579,
+      "learning_rate": 8.470013072054442e-06,
+      "loss": 0.2264,
+      "step": 2967
+    },
+    {
+      "epoch": 0.9911504424778761,
+      "grad_norm": 0.42605908508496787,
+      "learning_rate": 8.468613663483894e-06,
+      "loss": 0.2163,
+      "step": 2968
+    },
+    {
+      "epoch": 0.9914843880447487,
+      "grad_norm": 0.7310766075725814,
+      "learning_rate": 8.467213730938408e-06,
+      "loss": 0.2194,
+      "step": 2969
+    },
+    {
+      "epoch": 0.9918183336116213,
+      "grad_norm": 0.40667359914165013,
+      "learning_rate": 8.465813274629466e-06,
+      "loss": 0.214,
+      "step": 2970
+    },
+    {
+      "epoch": 0.9921522791784939,
+      "grad_norm": 0.43679647168307384,
+      "learning_rate": 8.46441229476862e-06,
+      "loss": 0.2407,
+      "step": 2971
+    },
+    {
+      "epoch": 0.9924862247453665,
+      "grad_norm": 0.41879553528039526,
+      "learning_rate": 8.463010791567503e-06,
+      "loss": 0.2183,
+      "step": 2972
+    },
+    {
+      "epoch": 0.9928201703122391,
+      "grad_norm": 0.43420445883979225,
+      "learning_rate": 8.461608765237832e-06,
+      "loss": 0.2221,
+      "step": 2973
+    },
+    {
+      "epoch": 0.9931541158791117,
+      "grad_norm": 0.42725936711171186,
+      "learning_rate": 8.460206215991398e-06,
+      "loss": 0.2236,
+      "step": 2974
+    },
+    {
+      "epoch": 0.9934880614459843,
+      "grad_norm": 0.45881095769887553,
+      "learning_rate": 8.458803144040071e-06,
+      "loss": 0.2374,
+      "step": 2975
+    },
+    {
+      "epoch": 0.993822007012857,
+      "grad_norm": 0.4298372232535485,
+      "learning_rate": 8.457399549595803e-06,
+      "loss": 0.2045,
+      "step": 2976
+    },
+    {
+      "epoch": 0.9941559525797296,
+      "grad_norm": 0.4446130630825048,
+      "learning_rate": 8.455995432870626e-06,
+      "loss": 0.2287,
+      "step": 2977
+    },
+    {
+      "epoch": 0.9944898981466022,
+      "grad_norm": 0.4227750823184624,
+      "learning_rate": 8.454590794076642e-06,
+      "loss": 0.2199,
+      "step": 2978
+    },
+    {
+      "epoch": 0.9948238437134747,
+      "grad_norm": 0.8347772521786165,
+      "learning_rate": 8.453185633426044e-06,
+      "loss": 0.2208,
+      "step": 2979
+    },
+    {
+      "epoch": 0.9951577892803473,
+      "grad_norm": 0.38705711704080775,
+      "learning_rate": 8.451779951131096e-06,
+      "loss": 0.2048,
+      "step": 2980
+    },
+    {
+      "epoch": 0.9954917348472199,
+      "grad_norm": 0.53915343201022,
+      "learning_rate": 8.450373747404143e-06,
+      "loss": 0.2052,
+      "step": 2981
+    },
+    {
+      "epoch": 0.9958256804140925,
+      "grad_norm": 0.42343951131201196,
+      "learning_rate": 8.448967022457611e-06,
+      "loss": 0.2123,
+      "step": 2982
+    },
+    {
+      "epoch": 0.9961596259809651,
+      "grad_norm": 0.39481374145182524,
+      "learning_rate": 8.447559776503998e-06,
+      "loss": 0.2132,
+      "step": 2983
+    },
+    {
+      "epoch": 0.9964935715478377,
+      "grad_norm": 0.42340444262938,
+      "learning_rate": 8.446152009755886e-06,
+      "loss": 0.2093,
+      "step": 2984
+    },
+    {
+      "epoch": 0.9968275171147103,
+      "grad_norm": 0.5923161067138794,
+      "learning_rate": 8.444743722425937e-06,
+      "loss": 0.2359,
+      "step": 2985
+    },
+    {
+      "epoch": 0.9971614626815829,
+      "grad_norm": 0.537393246167669,
+      "learning_rate": 8.443334914726886e-06,
+      "loss": 0.2035,
+      "step": 2986
+    },
+    {
+      "epoch": 0.9974954082484555,
+      "grad_norm": 0.47016253843865485,
+      "learning_rate": 8.441925586871556e-06,
+      "loss": 0.2039,
+      "step": 2987
+    },
+    {
+      "epoch": 0.9978293538153281,
+      "grad_norm": 0.3945282745483418,
+      "learning_rate": 8.440515739072836e-06,
+      "loss": 0.2017,
+      "step": 2988
+    },
+    {
+      "epoch": 0.9981632993822007,
+      "grad_norm": 0.42985799339322056,
+      "learning_rate": 8.439105371543703e-06,
+      "loss": 0.2134,
+      "step": 2989
+    },
+    {
+      "epoch": 0.9984972449490733,
+      "grad_norm": 0.43566435703151096,
+      "learning_rate": 8.43769448449721e-06,
+      "loss": 0.2139,
+      "step": 2990
+    },
+    {
+      "epoch": 0.9988311905159459,
+      "grad_norm": 0.43132891430456666,
+      "learning_rate": 8.436283078146488e-06,
+      "loss": 0.2223,
+      "step": 2991
+    },
+    {
+      "epoch": 0.9991651360828185,
+      "grad_norm": 0.546060681351875,
+      "learning_rate": 8.434871152704745e-06,
+      "loss": 0.219,
+      "step": 2992
+    },
+    {
+      "epoch": 0.9994990816496911,
+      "grad_norm": 0.45277563875621096,
+      "learning_rate": 8.433458708385272e-06,
+      "loss": 0.2125,
+      "step": 2993
+    },
+    {
+      "epoch": 0.9998330272165638,
+      "grad_norm": 0.4072238139150075,
+      "learning_rate": 8.432045745401431e-06,
+      "loss": 0.2231,
+      "step": 2994
+    },
+    {
+      "epoch": 0.9998330272165638,
+      "eval_loss": 0.21826396882534027,
+      "eval_runtime": 187.3402,
+      "eval_samples_per_second": 107.681,
+      "eval_steps_per_second": 1.687,
+      "step": 2994
+    },
+    {
+      "epoch": 1.0001669727834364,
+      "grad_norm": 0.37766855711266734,
+      "learning_rate": 8.430632263966672e-06,
+      "loss": 0.2026,
+      "step": 2995
+    },
+    {
+      "epoch": 1.0005009183503089,
+      "grad_norm": 0.42746862033576705,
+      "learning_rate": 8.429218264294512e-06,
+      "loss": 0.1963,
+      "step": 2996
+    },
+    {
+      "epoch": 1.0008348639171816,
+      "grad_norm": 0.42313117231942904,
+      "learning_rate": 8.427803746598557e-06,
+      "loss": 0.1963,
+      "step": 2997
+    },
+    {
+      "epoch": 1.001168809484054,
+      "grad_norm": 0.40410590824339193,
+      "learning_rate": 8.426388711092486e-06,
+      "loss": 0.1937,
+      "step": 2998
+    },
+    {
+      "epoch": 1.0015027550509268,
+      "grad_norm": 0.3865466171900227,
+      "learning_rate": 8.424973157990053e-06,
+      "loss": 0.1889,
+      "step": 2999
+    },
+    {
+      "epoch": 1.0018367006177993,
+      "grad_norm": 0.42882739122705976,
+      "learning_rate": 8.4235570875051e-06,
+      "loss": 0.2107,
+      "step": 3000
+    },
+    {
+      "epoch": 1.002170646184672,
+      "grad_norm": 0.45285004117271893,
+      "learning_rate": 8.422140499851536e-06,
+      "loss": 0.2176,
+      "step": 3001
+    },
+    {
+      "epoch": 1.0025045917515445,
+      "grad_norm": 0.411068092342574,
+      "learning_rate": 8.420723395243356e-06,
+      "loss": 0.1989,
+      "step": 3002
+    },
+    {
+      "epoch": 1.002838537318417,
+      "grad_norm": 0.4226831623463052,
+      "learning_rate": 8.419305773894628e-06,
+      "loss": 0.1881,
+      "step": 3003
+    },
+    {
+      "epoch": 1.0031724828852897,
+      "grad_norm": 0.4373537624300133,
+      "learning_rate": 8.417887636019504e-06,
+      "loss": 0.1967,
+      "step": 3004
+    },
+    {
+      "epoch": 1.0035064284521622,
+      "grad_norm": 0.4359816928757857,
+      "learning_rate": 8.416468981832207e-06,
+      "loss": 0.1959,
+      "step": 3005
+    },
+    {
+      "epoch": 1.003840374019035,
+      "grad_norm": 0.4397588855669126,
+      "learning_rate": 8.415049811547043e-06,
+      "loss": 0.1953,
+      "step": 3006
+    },
+    {
+      "epoch": 1.0041743195859074,
+      "grad_norm": 0.4361235547061362,
+      "learning_rate": 8.413630125378393e-06,
+      "loss": 0.1965,
+      "step": 3007
+    },
+    {
+      "epoch": 1.0045082651527801,
+      "grad_norm": 0.44202989818976113,
+      "learning_rate": 8.412209923540719e-06,
+      "loss": 0.2008,
+      "step": 3008
+    },
+    {
+      "epoch": 1.0048422107196526,
+      "grad_norm": 0.5020608356876495,
+      "learning_rate": 8.41078920624856e-06,
+      "loss": 0.2118,
+      "step": 3009
+    },
+    {
+      "epoch": 1.0051761562865253,
+      "grad_norm": 0.427867506708943,
+      "learning_rate": 8.409367973716527e-06,
+      "loss": 0.2082,
+      "step": 3010
+    },
+    {
+      "epoch": 1.0055101018533978,
+      "grad_norm": 0.3950096858664292,
+      "learning_rate": 8.40794622615932e-06,
+      "loss": 0.1906,
+      "step": 3011
+    },
+    {
+      "epoch": 1.0058440474202706,
+      "grad_norm": 0.46858017830302484,
+      "learning_rate": 8.406523963791709e-06,
+      "loss": 0.2002,
+      "step": 3012
+    },
+    {
+      "epoch": 1.006177992987143,
+      "grad_norm": 0.46112765843275033,
+      "learning_rate": 8.405101186828542e-06,
+      "loss": 0.2069,
+      "step": 3013
+    },
+    {
+      "epoch": 1.0065119385540158,
+      "grad_norm": 0.7898365532056968,
+      "learning_rate": 8.403677895484746e-06,
+      "loss": 0.2108,
+      "step": 3014
+    },
+    {
+      "epoch": 1.0068458841208883,
+      "grad_norm": 0.37587310991070744,
+      "learning_rate": 8.402254089975328e-06,
+      "loss": 0.188,
+      "step": 3015
+    },
+    {
+      "epoch": 1.007179829687761,
+      "grad_norm": 0.4706264177833519,
+      "learning_rate": 8.400829770515369e-06,
+      "loss": 0.1941,
+      "step": 3016
+    },
+    {
+      "epoch": 1.0075137752546335,
+      "grad_norm": 0.4025345426782933,
+      "learning_rate": 8.399404937320031e-06,
+      "loss": 0.187,
+      "step": 3017
+    },
+    {
+      "epoch": 1.0078477208215062,
+      "grad_norm": 0.4425795571729669,
+      "learning_rate": 8.397979590604548e-06,
+      "loss": 0.2139,
+      "step": 3018
+    },
+    {
+      "epoch": 1.0081816663883787,
+      "grad_norm": 0.41370338504034,
+      "learning_rate": 8.39655373058424e-06,
+      "loss": 0.1948,
+      "step": 3019
+    },
+    {
+      "epoch": 1.0085156119552512,
+      "grad_norm": 0.49065974447042454,
+      "learning_rate": 8.395127357474498e-06,
+      "loss": 0.2014,
+      "step": 3020
+    },
+    {
+      "epoch": 1.008849557522124,
+      "grad_norm": 0.42889156437278847,
+      "learning_rate": 8.39370047149079e-06,
+      "loss": 0.1955,
+      "step": 3021
+    },
+    {
+      "epoch": 1.0091835030889964,
+      "grad_norm": 0.4188098849683885,
+      "learning_rate": 8.39227307284867e-06,
+      "loss": 0.1857,
+      "step": 3022
+    },
+    {
+      "epoch": 1.0095174486558691,
+      "grad_norm": 0.4496220697484712,
+      "learning_rate": 8.390845161763756e-06,
+      "loss": 0.2073,
+      "step": 3023
+    },
+    {
+      "epoch": 1.0098513942227416,
+      "grad_norm": 0.4471112381315624,
+      "learning_rate": 8.389416738451755e-06,
+      "loss": 0.1859,
+      "step": 3024
+    },
+    {
+      "epoch": 1.0101853397896143,
+      "grad_norm": 0.5110693539703756,
+      "learning_rate": 8.387987803128447e-06,
+      "loss": 0.2178,
+      "step": 3025
+    },
+    {
+      "epoch": 1.0105192853564868,
+      "grad_norm": 0.5629538582211671,
+      "learning_rate": 8.386558356009691e-06,
+      "loss": 0.208,
+      "step": 3026
+    },
+    {
+      "epoch": 1.0108532309233595,
+      "grad_norm": 0.43042048600723654,
+      "learning_rate": 8.385128397311418e-06,
+      "loss": 0.191,
+      "step": 3027
+    },
+    {
+      "epoch": 1.011187176490232,
+      "grad_norm": 0.443595870073697,
+      "learning_rate": 8.383697927249641e-06,
+      "loss": 0.1974,
+      "step": 3028
+    },
+    {
+      "epoch": 1.0115211220571048,
+      "grad_norm": 0.5034619295424242,
+      "learning_rate": 8.382266946040453e-06,
+      "loss": 0.2031,
+      "step": 3029
+    },
+    {
+      "epoch": 1.0118550676239773,
+      "grad_norm": 0.48727180978971396,
+      "learning_rate": 8.380835453900017e-06,
+      "loss": 0.2093,
+      "step": 3030
+    },
+    {
+      "epoch": 1.01218901319085,
+      "grad_norm": 0.4396347345023138,
+      "learning_rate": 8.379403451044576e-06,
+      "loss": 0.197,
+      "step": 3031
+    },
+    {
+      "epoch": 1.0125229587577225,
+      "grad_norm": 0.44433522767753325,
+      "learning_rate": 8.377970937690455e-06,
+      "loss": 0.2064,
+      "step": 3032
+    },
+    {
+      "epoch": 1.0128569043245952,
+      "grad_norm": 0.42254899723175227,
+      "learning_rate": 8.376537914054048e-06,
+      "loss": 0.1891,
+      "step": 3033
+    },
+    {
+      "epoch": 1.0131908498914677,
+      "grad_norm": 0.4347839679518912,
+      "learning_rate": 8.37510438035183e-06,
+      "loss": 0.193,
+      "step": 3034
+    },
+    {
+      "epoch": 1.0135247954583404,
+      "grad_norm": 0.4350862330519345,
+      "learning_rate": 8.373670336800358e-06,
+      "loss": 0.2063,
+      "step": 3035
+    },
+    {
+      "epoch": 1.013858741025213,
+      "grad_norm": 0.5486004506039731,
+      "learning_rate": 8.372235783616258e-06,
+      "loss": 0.2109,
+      "step": 3036
+    },
+    {
+      "epoch": 1.0141926865920854,
+      "grad_norm": 0.41129617581244454,
+      "learning_rate": 8.370800721016232e-06,
+      "loss": 0.1962,
+      "step": 3037
+    },
+    {
+      "epoch": 1.014526632158958,
+      "grad_norm": 0.4257282696992318,
+      "learning_rate": 8.369365149217072e-06,
+      "loss": 0.1982,
+      "step": 3038
+    },
+    {
+      "epoch": 1.0148605777258306,
+      "grad_norm": 0.46728603791520196,
+      "learning_rate": 8.36792906843563e-06,
+      "loss": 0.1908,
+      "step": 3039
+    },
+    {
+      "epoch": 1.0151945232927033,
+      "grad_norm": 0.4749390533604057,
+      "learning_rate": 8.366492478888849e-06,
+      "loss": 0.2052,
+      "step": 3040
+    },
+    {
+      "epoch": 1.0155284688595758,
+      "grad_norm": 0.46810287444765153,
+      "learning_rate": 8.365055380793737e-06,
+      "loss": 0.2041,
+      "step": 3041
+    },
+    {
+      "epoch": 1.0158624144264485,
+      "grad_norm": 0.42069357580760686,
+      "learning_rate": 8.363617774367389e-06,
+      "loss": 0.2063,
+      "step": 3042
+    },
+    {
+      "epoch": 1.016196359993321,
+      "grad_norm": 0.38632263608832246,
+      "learning_rate": 8.36217965982697e-06,
+      "loss": 0.1904,
+      "step": 3043
+    },
+    {
+      "epoch": 1.0165303055601937,
+      "grad_norm": 0.43057927948036445,
+      "learning_rate": 8.360741037389727e-06,
+      "loss": 0.216,
+      "step": 3044
+    },
+    {
+      "epoch": 1.0168642511270662,
+      "grad_norm": 0.4222757576848028,
+      "learning_rate": 8.359301907272976e-06,
+      "loss": 0.1915,
+      "step": 3045
+    },
+    {
+      "epoch": 1.017198196693939,
+      "grad_norm": 0.3786692087465665,
+      "learning_rate": 8.35786226969412e-06,
+      "loss": 0.1788,
+      "step": 3046
+    },
+    {
+      "epoch": 1.0175321422608115,
+      "grad_norm": 0.45325514122512583,
+      "learning_rate": 8.356422124870629e-06,
+      "loss": 0.2057,
+      "step": 3047
+    },
+    {
+      "epoch": 1.0178660878276842,
+      "grad_norm": 0.40238918835240045,
+      "learning_rate": 8.354981473020056e-06,
+      "loss": 0.1918,
+      "step": 3048
+    },
+    {
+      "epoch": 1.0182000333945567,
+      "grad_norm": 0.4393213758556735,
+      "learning_rate": 8.353540314360027e-06,
+      "loss": 0.1956,
+      "step": 3049
+    },
+    {
+      "epoch": 1.0185339789614294,
+      "grad_norm": 0.41890810895805786,
+      "learning_rate": 8.352098649108246e-06,
+      "loss": 0.1966,
+      "step": 3050
+    },
+    {
+      "epoch": 1.0188679245283019,
+      "grad_norm": 0.4322860515599835,
+      "learning_rate": 8.350656477482497e-06,
+      "loss": 0.1988,
+      "step": 3051
+    },
+    {
+      "epoch": 1.0192018700951744,
+      "grad_norm": 0.4546480187544326,
+      "learning_rate": 8.349213799700635e-06,
+      "loss": 0.2049,
+      "step": 3052
+    },
+    {
+      "epoch": 1.019535815662047,
+      "grad_norm": 0.4717472602597745,
+      "learning_rate": 8.34777061598059e-06,
+      "loss": 0.2074,
+      "step": 3053
+    },
+    {
+      "epoch": 1.0198697612289196,
+      "grad_norm": 0.4401320901117122,
+      "learning_rate": 8.346326926540377e-06,
+      "loss": 0.2015,
+      "step": 3054
+    },
+    {
+      "epoch": 1.0202037067957923,
+      "grad_norm": 0.43023737821846486,
+      "learning_rate": 8.344882731598079e-06,
+      "loss": 0.2026,
+      "step": 3055
+    },
+    {
+      "epoch": 1.0205376523626648,
+      "grad_norm": 0.41879723678450614,
+      "learning_rate": 8.343438031371858e-06,
+      "loss": 0.1928,
+      "step": 3056
+    },
+    {
+      "epoch": 1.0208715979295375,
+      "grad_norm": 0.4906147418309719,
+      "learning_rate": 8.341992826079956e-06,
+      "loss": 0.2218,
+      "step": 3057
+    },
+    {
+      "epoch": 1.02120554349641,
+      "grad_norm": 0.475504726884439,
+      "learning_rate": 8.340547115940688e-06,
+      "loss": 0.1776,
+      "step": 3058
+    },
+    {
+      "epoch": 1.0215394890632827,
+      "grad_norm": 0.49917711995665354,
+      "learning_rate": 8.339100901172443e-06,
+      "loss": 0.2136,
+      "step": 3059
+    },
+    {
+      "epoch": 1.0218734346301552,
+      "grad_norm": 0.6420536276666094,
+      "learning_rate": 8.337654181993691e-06,
+      "loss": 0.1996,
+      "step": 3060
+    },
+    {
+      "epoch": 1.022207380197028,
+      "grad_norm": 0.586797523605855,
+      "learning_rate": 8.336206958622975e-06,
+      "loss": 0.2052,
+      "step": 3061
+    },
+    {
+      "epoch": 1.0225413257639004,
+      "grad_norm": 0.42846778032094296,
+      "learning_rate": 8.334759231278915e-06,
+      "loss": 0.1913,
+      "step": 3062
+    },
+    {
+      "epoch": 1.0228752713307732,
+      "grad_norm": 0.43006716318612825,
+      "learning_rate": 8.333311000180208e-06,
+      "loss": 0.1964,
+      "step": 3063
+    },
+    {
+      "epoch": 1.0232092168976457,
+      "grad_norm": 0.415041357651441,
+      "learning_rate": 8.331862265545627e-06,
+      "loss": 0.1903,
+      "step": 3064
+    },
+    {
+      "epoch": 1.0235431624645184,
+      "grad_norm": 0.4809451997950013,
+      "learning_rate": 8.330413027594019e-06,
+      "loss": 0.2193,
+      "step": 3065
+    },
+    {
+      "epoch": 1.0238771080313909,
+      "grad_norm": 0.4849800241783216,
+      "learning_rate": 8.328963286544309e-06,
+      "loss": 0.2022,
+      "step": 3066
+    },
+    {
+      "epoch": 1.0242110535982636,
+      "grad_norm": 0.4358252609774767,
+      "learning_rate": 8.327513042615496e-06,
+      "loss": 0.2153,
+      "step": 3067
+    },
+    {
+      "epoch": 1.024544999165136,
+      "grad_norm": 0.42611601243396807,
+      "learning_rate": 8.326062296026657e-06,
+      "loss": 0.1973,
+      "step": 3068
+    },
+    {
+      "epoch": 1.0248789447320086,
+      "grad_norm": 0.5112017276974804,
+      "learning_rate": 8.324611046996947e-06,
+      "loss": 0.219,
+      "step": 3069
+    },
+    {
+      "epoch": 1.0252128902988813,
+      "grad_norm": 0.4373035623997926,
+      "learning_rate": 8.32315929574559e-06,
+      "loss": 0.1959,
+      "step": 3070
+    },
+    {
+      "epoch": 1.0255468358657538,
+      "grad_norm": 0.3837563091069947,
+      "learning_rate": 8.321707042491895e-06,
+      "loss": 0.1844,
+      "step": 3071
+    },
+    {
+      "epoch": 1.0258807814326265,
+      "grad_norm": 0.45421219166684185,
+      "learning_rate": 8.320254287455238e-06,
+      "loss": 0.2056,
+      "step": 3072
+    },
+    {
+      "epoch": 1.026214726999499,
+      "grad_norm": 0.40781482071872827,
+      "learning_rate": 8.318801030855078e-06,
+      "loss": 0.1864,
+      "step": 3073
+    },
+    {
+      "epoch": 1.0265486725663717,
+      "grad_norm": 0.4706705010349072,
+      "learning_rate": 8.317347272910944e-06,
+      "loss": 0.2033,
+      "step": 3074
+    },
+    {
+      "epoch": 1.0268826181332442,
+      "grad_norm": 0.45550347872407415,
+      "learning_rate": 8.315893013842441e-06,
+      "loss": 0.2192,
+      "step": 3075
+    },
+    {
+      "epoch": 1.027216563700117,
+      "grad_norm": 0.4563220551157251,
+      "learning_rate": 8.31443825386926e-06,
+      "loss": 0.2164,
+      "step": 3076
+    },
+    {
+      "epoch": 1.0275505092669894,
+      "grad_norm": 0.4866578429632529,
+      "learning_rate": 8.312982993211151e-06,
+      "loss": 0.2222,
+      "step": 3077
+    },
+    {
+      "epoch": 1.0278844548338621,
+      "grad_norm": 0.4244291790269176,
+      "learning_rate": 8.311527232087951e-06,
+      "loss": 0.2031,
+      "step": 3078
+    },
+    {
+      "epoch": 1.0282184004007346,
+      "grad_norm": 0.4243539342498891,
+      "learning_rate": 8.310070970719573e-06,
+      "loss": 0.2022,
+      "step": 3079
+    },
+    {
+      "epoch": 1.0285523459676074,
+      "grad_norm": 0.4218336141097441,
+      "learning_rate": 8.308614209325997e-06,
+      "loss": 0.2044,
+      "step": 3080
+    },
+    {
+      "epoch": 1.0288862915344799,
+      "grad_norm": 0.4661210673858716,
+      "learning_rate": 8.30715694812729e-06,
+      "loss": 0.202,
+      "step": 3081
+    },
+    {
+      "epoch": 1.0292202371013526,
+      "grad_norm": 0.4768217692487075,
+      "learning_rate": 8.305699187343586e-06,
+      "loss": 0.2128,
+      "step": 3082
+    },
+    {
+      "epoch": 1.029554182668225,
+      "grad_norm": 0.46166941719548166,
+      "learning_rate": 8.304240927195094e-06,
+      "loss": 0.2108,
+      "step": 3083
+    },
+    {
+      "epoch": 1.0298881282350978,
+      "grad_norm": 0.43191760215598024,
+      "learning_rate": 8.302782167902103e-06,
+      "loss": 0.2062,
+      "step": 3084
+    },
+    {
+      "epoch": 1.0302220738019703,
+      "grad_norm": 0.41257731682620635,
+      "learning_rate": 8.30132290968498e-06,
+      "loss": 0.1818,
+      "step": 3085
+    },
+    {
+      "epoch": 1.0305560193688428,
+      "grad_norm": 0.43840781003063306,
+      "learning_rate": 8.299863152764158e-06,
+      "loss": 0.2066,
+      "step": 3086
+    },
+    {
+      "epoch": 1.0308899649357155,
+      "grad_norm": 0.4919766494174308,
+      "learning_rate": 8.298402897360152e-06,
+      "loss": 0.1962,
+      "step": 3087
+    },
+    {
+      "epoch": 1.031223910502588,
+      "grad_norm": 0.4370771922590659,
+      "learning_rate": 8.29694214369355e-06,
+      "loss": 0.1989,
+      "step": 3088
+    },
+    {
+      "epoch": 1.0315578560694607,
+      "grad_norm": 0.4522687494366065,
+      "learning_rate": 8.295480891985019e-06,
+      "loss": 0.197,
+      "step": 3089
+    },
+    {
+      "epoch": 1.0318918016363332,
+      "grad_norm": 0.459698970205727,
+      "learning_rate": 8.294019142455295e-06,
+      "loss": 0.215,
+      "step": 3090
+    },
+    {
+      "epoch": 1.032225747203206,
+      "grad_norm": 0.45768357278304855,
+      "learning_rate": 8.292556895325195e-06,
+      "loss": 0.1834,
+      "step": 3091
+    },
+    {
+      "epoch": 1.0325596927700784,
+      "grad_norm": 0.4779953772645454,
+      "learning_rate": 8.291094150815607e-06,
+      "loss": 0.2046,
+      "step": 3092
+    },
+    {
+      "epoch": 1.0328936383369511,
+      "grad_norm": 0.4110970900415633,
+      "learning_rate": 8.289630909147494e-06,
+      "loss": 0.1926,
+      "step": 3093
+    },
+    {
+      "epoch": 1.0332275839038236,
+      "grad_norm": 0.4263110732845005,
+      "learning_rate": 8.2881671705419e-06,
+      "loss": 0.1961,
+      "step": 3094
+    },
+    {
+      "epoch": 1.0335615294706963,
+      "grad_norm": 0.42351189564424013,
+      "learning_rate": 8.286702935219936e-06,
+      "loss": 0.1924,
+      "step": 3095
+    },
+    {
+      "epoch": 1.0338954750375688,
+      "grad_norm": 0.5111127526263297,
+      "learning_rate": 8.285238203402796e-06,
+      "loss": 0.2099,
+      "step": 3096
+    },
+    {
+      "epoch": 1.0342294206044416,
+      "grad_norm": 0.4395888062180706,
+      "learning_rate": 8.283772975311742e-06,
+      "loss": 0.2091,
+      "step": 3097
+    },
+    {
+      "epoch": 1.034563366171314,
+      "grad_norm": 0.40597673314935695,
+      "learning_rate": 8.282307251168116e-06,
+      "loss": 0.1948,
+      "step": 3098
+    },
+    {
+      "epoch": 1.0348973117381868,
+      "grad_norm": 0.41337566836535816,
+      "learning_rate": 8.28084103119333e-06,
+      "loss": 0.1956,
+      "step": 3099
+    },
+    {
+      "epoch": 1.0352312573050593,
+      "grad_norm": 0.4440489461102574,
+      "learning_rate": 8.279374315608877e-06,
+      "loss": 0.1849,
+      "step": 3100
+    },
+    {
+      "epoch": 1.0355652028719318,
+      "grad_norm": 0.43649748059427285,
+      "learning_rate": 8.27790710463632e-06,
+      "loss": 0.202,
+      "step": 3101
+    },
+    {
+      "epoch": 1.0358991484388045,
+      "grad_norm": 0.45527974348537775,
+      "learning_rate": 8.276439398497298e-06,
+      "loss": 0.2069,
+      "step": 3102
+    },
+    {
+      "epoch": 1.036233094005677,
+      "grad_norm": 0.4411849130395665,
+      "learning_rate": 8.274971197413527e-06,
+      "loss": 0.2018,
+      "step": 3103
+    },
+    {
+      "epoch": 1.0365670395725497,
+      "grad_norm": 0.432927432928707,
+      "learning_rate": 8.273502501606794e-06,
+      "loss": 0.1921,
+      "step": 3104
+    },
+    {
+      "epoch": 1.0369009851394222,
+      "grad_norm": 0.48645160430520945,
+      "learning_rate": 8.272033311298965e-06,
+      "loss": 0.2062,
+      "step": 3105
+    },
+    {
+      "epoch": 1.037234930706295,
+      "grad_norm": 0.4723298015246345,
+      "learning_rate": 8.270563626711979e-06,
+      "loss": 0.2191,
+      "step": 3106
+    },
+    {
+      "epoch": 1.0375688762731674,
+      "grad_norm": 0.4286934741750141,
+      "learning_rate": 8.269093448067845e-06,
+      "loss": 0.1974,
+      "step": 3107
+    },
+    {
+      "epoch": 1.0379028218400401,
+      "grad_norm": 0.4898091980655584,
+      "learning_rate": 8.267622775588653e-06,
+      "loss": 0.2071,
+      "step": 3108
+    },
+    {
+      "epoch": 1.0382367674069126,
+      "grad_norm": 0.42263697887339474,
+      "learning_rate": 8.266151609496567e-06,
+      "loss": 0.1943,
+      "step": 3109
+    },
+    {
+      "epoch": 1.0385707129737853,
+      "grad_norm": 0.4409916018139914,
+      "learning_rate": 8.26467995001382e-06,
+      "loss": 0.1845,
+      "step": 3110
+    },
+    {
+      "epoch": 1.0389046585406578,
+      "grad_norm": 0.46867210638599427,
+      "learning_rate": 8.26320779736273e-06,
+      "loss": 0.2111,
+      "step": 3111
+    },
+    {
+      "epoch": 1.0392386041075306,
+      "grad_norm": 1.1833877800166233,
+      "learning_rate": 8.261735151765678e-06,
+      "loss": 0.2006,
+      "step": 3112
+    },
+    {
+      "epoch": 1.039572549674403,
+      "grad_norm": 0.4392933903047106,
+      "learning_rate": 8.260262013445126e-06,
+      "loss": 0.1959,
+      "step": 3113
+    },
+    {
+      "epoch": 1.0399064952412758,
+      "grad_norm": 0.43251596850546686,
+      "learning_rate": 8.258788382623607e-06,
+      "loss": 0.1921,
+      "step": 3114
+    },
+    {
+      "epoch": 1.0402404408081483,
+      "grad_norm": 0.5317197119185111,
+      "learning_rate": 8.257314259523732e-06,
+      "loss": 0.2015,
+      "step": 3115
+    },
+    {
+      "epoch": 1.040574386375021,
+      "grad_norm": 0.4656601804952947,
+      "learning_rate": 8.255839644368185e-06,
+      "loss": 0.2033,
+      "step": 3116
+    },
+    {
+      "epoch": 1.0409083319418935,
+      "grad_norm": 0.48325787940709314,
+      "learning_rate": 8.254364537379725e-06,
+      "loss": 0.1973,
+      "step": 3117
+    },
+    {
+      "epoch": 1.041242277508766,
+      "grad_norm": 0.4615708346164602,
+      "learning_rate": 8.25288893878118e-06,
+      "loss": 0.2173,
+      "step": 3118
+    },
+    {
+      "epoch": 1.0415762230756387,
+      "grad_norm": 0.4910469321775321,
+      "learning_rate": 8.251412848795462e-06,
+      "loss": 0.2112,
+      "step": 3119
+    },
+    {
+      "epoch": 1.0419101686425112,
+      "grad_norm": 0.42906476123756765,
+      "learning_rate": 8.249936267645546e-06,
+      "loss": 0.2085,
+      "step": 3120
+    },
+    {
+      "epoch": 1.042244114209384,
+      "grad_norm": 0.4468814505175119,
+      "learning_rate": 8.248459195554492e-06,
+      "loss": 0.2027,
+      "step": 3121
+    },
+    {
+      "epoch": 1.0425780597762564,
+      "grad_norm": 0.4510701292444035,
+      "learning_rate": 8.246981632745428e-06,
+      "loss": 0.216,
+      "step": 3122
+    },
+    {
+      "epoch": 1.0429120053431291,
+      "grad_norm": 0.5012622614358289,
+      "learning_rate": 8.245503579441554e-06,
+      "loss": 0.2041,
+      "step": 3123
+    },
+    {
+      "epoch": 1.0432459509100016,
+      "grad_norm": 0.44604941183333585,
+      "learning_rate": 8.244025035866151e-06,
+      "loss": 0.2107,
+      "step": 3124
+    },
+    {
+      "epoch": 1.0435798964768743,
+      "grad_norm": 0.43182075677881904,
+      "learning_rate": 8.242546002242569e-06,
+      "loss": 0.2012,
+      "step": 3125
+    },
+    {
+      "epoch": 1.0439138420437468,
+      "grad_norm": 0.7065204162623006,
+      "learning_rate": 8.241066478794233e-06,
+      "loss": 0.2149,
+      "step": 3126
+    },
+    {
+      "epoch": 1.0442477876106195,
+      "grad_norm": 0.43548371799456587,
+      "learning_rate": 8.239586465744644e-06,
+      "loss": 0.1964,
+      "step": 3127
+    },
+    {
+      "epoch": 1.044581733177492,
+      "grad_norm": 0.45976736574310023,
+      "learning_rate": 8.238105963317376e-06,
+      "loss": 0.1977,
+      "step": 3128
+    },
+    {
+      "epoch": 1.0449156787443648,
+      "grad_norm": 0.45610032095662145,
+      "learning_rate": 8.236624971736071e-06,
+      "loss": 0.2006,
+      "step": 3129
+    },
+    {
+      "epoch": 1.0452496243112372,
+      "grad_norm": 0.4178927483884343,
+      "learning_rate": 8.235143491224458e-06,
+      "loss": 0.1941,
+      "step": 3130
+    },
+    {
+      "epoch": 1.04558356987811,
+      "grad_norm": 0.4450014435903185,
+      "learning_rate": 8.233661522006324e-06,
+      "loss": 0.2011,
+      "step": 3131
+    },
+    {
+      "epoch": 1.0459175154449825,
+      "grad_norm": 0.4678822230492103,
+      "learning_rate": 8.232179064305545e-06,
+      "loss": 0.2141,
+      "step": 3132
+    },
+    {
+      "epoch": 1.0462514610118552,
+      "grad_norm": 0.4659113518770199,
+      "learning_rate": 8.230696118346059e-06,
+      "loss": 0.2029,
+      "step": 3133
+    },
+    {
+      "epoch": 1.0465854065787277,
+      "grad_norm": 0.42572466204955733,
+      "learning_rate": 8.229212684351886e-06,
+      "loss": 0.2092,
+      "step": 3134
+    },
+    {
+      "epoch": 1.0469193521456002,
+      "grad_norm": 0.42976487899817956,
+      "learning_rate": 8.227728762547112e-06,
+      "loss": 0.2049,
+      "step": 3135
+    },
+    {
+      "epoch": 1.0472532977124729,
+      "grad_norm": 0.4285031948928849,
+      "learning_rate": 8.226244353155906e-06,
+      "loss": 0.1895,
+      "step": 3136
+    },
+    {
+      "epoch": 1.0475872432793454,
+      "grad_norm": 0.41985602222315516,
+      "learning_rate": 8.2247594564025e-06,
+      "loss": 0.1976,
+      "step": 3137
+    },
+    {
+      "epoch": 1.047921188846218,
+      "grad_norm": 0.44464462060650317,
+      "learning_rate": 8.22327407251121e-06,
+      "loss": 0.1917,
+      "step": 3138
+    },
+    {
+      "epoch": 1.0482551344130906,
+      "grad_norm": 0.4273282426042788,
+      "learning_rate": 8.221788201706416e-06,
+      "loss": 0.202,
+      "step": 3139
+    },
+    {
+      "epoch": 1.0485890799799633,
+      "grad_norm": 0.41653410454050144,
+      "learning_rate": 8.22030184421258e-06,
+      "loss": 0.2017,
+      "step": 3140
+    },
+    {
+      "epoch": 1.0489230255468358,
+      "grad_norm": 0.43858551157838355,
+      "learning_rate": 8.218815000254233e-06,
+      "loss": 0.2022,
+      "step": 3141
+    },
+    {
+      "epoch": 1.0492569711137085,
+      "grad_norm": 0.4908298917701791,
+      "learning_rate": 8.21732767005598e-06,
+      "loss": 0.2132,
+      "step": 3142
+    },
+    {
+      "epoch": 1.049590916680581,
+      "grad_norm": 0.4749605912490831,
+      "learning_rate": 8.215839853842498e-06,
+      "loss": 0.2158,
+      "step": 3143
+    },
+    {
+      "epoch": 1.0499248622474537,
+      "grad_norm": 0.4483578369643801,
+      "learning_rate": 8.214351551838541e-06,
+      "loss": 0.2043,
+      "step": 3144
+    },
+    {
+      "epoch": 1.0502588078143262,
+      "grad_norm": 0.42665670967430674,
+      "learning_rate": 8.212862764268936e-06,
+      "loss": 0.1972,
+      "step": 3145
+    },
+    {
+      "epoch": 1.050592753381199,
+      "grad_norm": 0.44815585260638,
+      "learning_rate": 8.21137349135858e-06,
+      "loss": 0.2022,
+      "step": 3146
+    },
+    {
+      "epoch": 1.0509266989480714,
+      "grad_norm": 0.42978041785750576,
+      "learning_rate": 8.209883733332444e-06,
+      "loss": 0.193,
+      "step": 3147
+    },
+    {
+      "epoch": 1.0512606445149442,
+      "grad_norm": 0.40783792703270516,
+      "learning_rate": 8.208393490415576e-06,
+      "loss": 0.2001,
+      "step": 3148
+    },
+    {
+      "epoch": 1.0515945900818167,
+      "grad_norm": 0.5034157622557892,
+      "learning_rate": 8.206902762833095e-06,
+      "loss": 0.1908,
+      "step": 3149
+    },
+    {
+      "epoch": 1.0519285356486892,
+      "grad_norm": 0.41596953468049164,
+      "learning_rate": 8.205411550810189e-06,
+      "loss": 0.2026,
+      "step": 3150
+    },
+    {
+      "epoch": 1.0522624812155619,
+      "grad_norm": 0.4387109004909959,
+      "learning_rate": 8.203919854572126e-06,
+      "loss": 0.1955,
+      "step": 3151
+    },
+    {
+      "epoch": 1.0525964267824344,
+      "grad_norm": 0.4400267808923557,
+      "learning_rate": 8.202427674344246e-06,
+      "loss": 0.1956,
+      "step": 3152
+    },
+    {
+      "epoch": 1.052930372349307,
+      "grad_norm": 0.42088665905677247,
+      "learning_rate": 8.200935010351958e-06,
+      "loss": 0.2036,
+      "step": 3153
+    },
+    {
+      "epoch": 1.0532643179161796,
+      "grad_norm": 0.42093977058051263,
+      "learning_rate": 8.199441862820746e-06,
+      "loss": 0.2094,
+      "step": 3154
+    },
+    {
+      "epoch": 1.0535982634830523,
+      "grad_norm": 0.41817125248997694,
+      "learning_rate": 8.197948231976169e-06,
+      "loss": 0.1905,
+      "step": 3155
+    },
+    {
+      "epoch": 1.0539322090499248,
+      "grad_norm": 0.42802200051982714,
+      "learning_rate": 8.196454118043856e-06,
+      "loss": 0.2004,
+      "step": 3156
+    },
+    {
+      "epoch": 1.0542661546167975,
+      "grad_norm": 0.4750486331903779,
+      "learning_rate": 8.194959521249512e-06,
+      "loss": 0.2019,
+      "step": 3157
+    },
+    {
+      "epoch": 1.05460010018367,
+      "grad_norm": 0.5160559948271373,
+      "learning_rate": 8.193464441818913e-06,
+      "loss": 0.209,
+      "step": 3158
+    },
+    {
+      "epoch": 1.0549340457505427,
+      "grad_norm": 0.420087588687286,
+      "learning_rate": 8.191968879977907e-06,
+      "loss": 0.2003,
+      "step": 3159
+    },
+    {
+      "epoch": 1.0552679913174152,
+      "grad_norm": 0.4037748789668453,
+      "learning_rate": 8.190472835952419e-06,
+      "loss": 0.1872,
+      "step": 3160
+    },
+    {
+      "epoch": 1.055601936884288,
+      "grad_norm": 0.38711125361133414,
+      "learning_rate": 8.188976309968443e-06,
+      "loss": 0.1858,
+      "step": 3161
+    },
+    {
+      "epoch": 1.0559358824511604,
+      "grad_norm": 0.47560444912511207,
+      "learning_rate": 8.187479302252045e-06,
+      "loss": 0.2034,
+      "step": 3162
+    },
+    {
+      "epoch": 1.0562698280180332,
+      "grad_norm": 0.3964934919375712,
+      "learning_rate": 8.185981813029368e-06,
+      "loss": 0.1892,
+      "step": 3163
+    },
+    {
+      "epoch": 1.0566037735849056,
+      "grad_norm": 0.45616318954926843,
+      "learning_rate": 8.184483842526623e-06,
+      "loss": 0.1934,
+      "step": 3164
+    },
+    {
+      "epoch": 1.0569377191517784,
+      "grad_norm": 0.4221881863168421,
+      "learning_rate": 8.1829853909701e-06,
+      "loss": 0.1882,
+      "step": 3165
+    },
+    {
+      "epoch": 1.0572716647186509,
+      "grad_norm": 0.4946404078548695,
+      "learning_rate": 8.181486458586153e-06,
+      "loss": 0.2197,
+      "step": 3166
+    },
+    {
+      "epoch": 1.0576056102855234,
+      "grad_norm": 0.41759418433544915,
+      "learning_rate": 8.179987045601217e-06,
+      "loss": 0.1905,
+      "step": 3167
+    },
+    {
+      "epoch": 1.057939555852396,
+      "grad_norm": 0.4680014004671277,
+      "learning_rate": 8.178487152241795e-06,
+      "loss": 0.2116,
+      "step": 3168
+    },
+    {
+      "epoch": 1.0582735014192686,
+      "grad_norm": 0.4679411328019381,
+      "learning_rate": 8.17698677873446e-06,
+      "loss": 0.2281,
+      "step": 3169
+    },
+    {
+      "epoch": 1.0586074469861413,
+      "grad_norm": 0.4306592891134695,
+      "learning_rate": 8.175485925305867e-06,
+      "loss": 0.1981,
+      "step": 3170
+    },
+    {
+      "epoch": 1.0589413925530138,
+      "grad_norm": 0.4279511404583339,
+      "learning_rate": 8.173984592182736e-06,
+      "loss": 0.2152,
+      "step": 3171
+    },
+    {
+      "epoch": 1.0592753381198865,
+      "grad_norm": 0.4256581923170128,
+      "learning_rate": 8.172482779591858e-06,
+      "loss": 0.1965,
+      "step": 3172
+    },
+    {
+      "epoch": 1.059609283686759,
+      "grad_norm": 0.42259600236442063,
+      "learning_rate": 8.170980487760101e-06,
+      "loss": 0.1959,
+      "step": 3173
+    },
+    {
+      "epoch": 1.0599432292536317,
+      "grad_norm": 0.3919759523982466,
+      "learning_rate": 8.169477716914405e-06,
+      "loss": 0.1835,
+      "step": 3174
+    },
+    {
+      "epoch": 1.0602771748205042,
+      "grad_norm": 0.3965257390740146,
+      "learning_rate": 8.16797446728178e-06,
+      "loss": 0.192,
+      "step": 3175
+    },
+    {
+      "epoch": 1.060611120387377,
+      "grad_norm": 0.44522593096004287,
+      "learning_rate": 8.16647073908931e-06,
+      "loss": 0.2098,
+      "step": 3176
+    },
+    {
+      "epoch": 1.0609450659542494,
+      "grad_norm": 0.4581897421406975,
+      "learning_rate": 8.164966532564152e-06,
+      "loss": 0.2049,
+      "step": 3177
+    },
+    {
+      "epoch": 1.0612790115211221,
+      "grad_norm": 0.4995506758881796,
+      "learning_rate": 8.163461847933532e-06,
+      "loss": 0.1989,
+      "step": 3178
+    },
+    {
+      "epoch": 1.0616129570879946,
+      "grad_norm": 0.40476982409428536,
+      "learning_rate": 8.161956685424752e-06,
+      "loss": 0.1843,
+      "step": 3179
+    },
+    {
+      "epoch": 1.0619469026548674,
+      "grad_norm": 0.42655470604742796,
+      "learning_rate": 8.160451045265183e-06,
+      "loss": 0.2032,
+      "step": 3180
+    },
+    {
+      "epoch": 1.0622808482217398,
+      "grad_norm": 0.42018746091677206,
+      "learning_rate": 8.158944927682269e-06,
+      "loss": 0.1965,
+      "step": 3181
+    },
+    {
+      "epoch": 1.0626147937886126,
+      "grad_norm": 0.4268157415568209,
+      "learning_rate": 8.157438332903531e-06,
+      "loss": 0.2006,
+      "step": 3182
+    },
+    {
+      "epoch": 1.062948739355485,
+      "grad_norm": 0.4033846341010823,
+      "learning_rate": 8.155931261156555e-06,
+      "loss": 0.1924,
+      "step": 3183
+    },
+    {
+      "epoch": 1.0632826849223576,
+      "grad_norm": 0.43889640040181027,
+      "learning_rate": 8.154423712669003e-06,
+      "loss": 0.1986,
+      "step": 3184
+    },
+    {
+      "epoch": 1.0636166304892303,
+      "grad_norm": 0.4001342148930981,
+      "learning_rate": 8.152915687668603e-06,
+      "loss": 0.1862,
+      "step": 3185
+    },
+    {
+      "epoch": 1.0639505760561028,
+      "grad_norm": 0.4105761999473694,
+      "learning_rate": 8.151407186383166e-06,
+      "loss": 0.191,
+      "step": 3186
+    },
+    {
+      "epoch": 1.0642845216229755,
+      "grad_norm": 0.4319067425333874,
+      "learning_rate": 8.149898209040568e-06,
+      "loss": 0.1997,
+      "step": 3187
+    },
+    {
+      "epoch": 1.064618467189848,
+      "grad_norm": 0.4560783659343897,
+      "learning_rate": 8.148388755868757e-06,
+      "loss": 0.2077,
+      "step": 3188
+    },
+    {
+      "epoch": 1.0649524127567207,
+      "grad_norm": 0.43798081875672745,
+      "learning_rate": 8.146878827095751e-06,
+      "loss": 0.1913,
+      "step": 3189
+    },
+    {
+      "epoch": 1.0652863583235932,
+      "grad_norm": 0.4372702184821335,
+      "learning_rate": 8.145368422949647e-06,
+      "loss": 0.2034,
+      "step": 3190
+    },
+    {
+      "epoch": 1.065620303890466,
+      "grad_norm": 0.41986516394757967,
+      "learning_rate": 8.143857543658606e-06,
+      "loss": 0.199,
+      "step": 3191
+    },
+    {
+      "epoch": 1.0659542494573384,
+      "grad_norm": 0.41907063437667164,
+      "learning_rate": 8.142346189450866e-06,
+      "loss": 0.2159,
+      "step": 3192
+    },
+    {
+      "epoch": 1.0662881950242111,
+      "grad_norm": 0.4094081046487141,
+      "learning_rate": 8.140834360554734e-06,
+      "loss": 0.1948,
+      "step": 3193
+    },
+    {
+      "epoch": 1.0666221405910836,
+      "grad_norm": 0.4371273342597166,
+      "learning_rate": 8.13932205719859e-06,
+      "loss": 0.2082,
+      "step": 3194
+    },
+    {
+      "epoch": 1.0669560861579563,
+      "grad_norm": 0.4135820465008076,
+      "learning_rate": 8.137809279610885e-06,
+      "loss": 0.1929,
+      "step": 3195
+    },
+    {
+      "epoch": 1.0672900317248288,
+      "grad_norm": 0.4399567478630838,
+      "learning_rate": 8.13629602802014e-06,
+      "loss": 0.1995,
+      "step": 3196
+    },
+    {
+      "epoch": 1.0676239772917016,
+      "grad_norm": 0.5092407448174172,
+      "learning_rate": 8.134782302654953e-06,
+      "loss": 0.2134,
+      "step": 3197
+    },
+    {
+      "epoch": 1.067957922858574,
+      "grad_norm": 0.5597000261870062,
+      "learning_rate": 8.133268103743989e-06,
+      "loss": 0.2111,
+      "step": 3198
+    },
+    {
+      "epoch": 1.0682918684254465,
+      "grad_norm": 0.4341592549068151,
+      "learning_rate": 8.131753431515984e-06,
+      "loss": 0.2052,
+      "step": 3199
+    },
+    {
+      "epoch": 1.0686258139923193,
+      "grad_norm": 0.5131722560356433,
+      "learning_rate": 8.130238286199747e-06,
+      "loss": 0.1998,
+      "step": 3200
+    },
+    {
+      "epoch": 1.0689597595591918,
+      "grad_norm": 0.4204116895958866,
+      "learning_rate": 8.128722668024161e-06,
+      "loss": 0.1961,
+      "step": 3201
+    },
+    {
+      "epoch": 1.0692937051260645,
+      "grad_norm": 0.46505953263469596,
+      "learning_rate": 8.127206577218177e-06,
+      "loss": 0.2106,
+      "step": 3202
+    },
+    {
+      "epoch": 1.069627650692937,
+      "grad_norm": 0.44604847632939687,
+      "learning_rate": 8.125690014010814e-06,
+      "loss": 0.2052,
+      "step": 3203
+    },
+    {
+      "epoch": 1.0699615962598097,
+      "grad_norm": 0.43756933565755846,
+      "learning_rate": 8.124172978631173e-06,
+      "loss": 0.1994,
+      "step": 3204
+    },
+    {
+      "epoch": 1.0702955418266822,
+      "grad_norm": 0.461832138118902,
+      "learning_rate": 8.12265547130842e-06,
+      "loss": 0.2158,
+      "step": 3205
+    },
+    {
+      "epoch": 1.070629487393555,
+      "grad_norm": 0.5180640624879851,
+      "learning_rate": 8.121137492271787e-06,
+      "loss": 0.206,
+      "step": 3206
+    },
+    {
+      "epoch": 1.0709634329604274,
+      "grad_norm": 0.44492318961314764,
+      "learning_rate": 8.119619041750586e-06,
+      "loss": 0.2123,
+      "step": 3207
+    },
+    {
+      "epoch": 1.0712973785273001,
+      "grad_norm": 0.44935278367080866,
+      "learning_rate": 8.118100119974197e-06,
+      "loss": 0.1938,
+      "step": 3208
+    },
+    {
+      "epoch": 1.0716313240941726,
+      "grad_norm": 0.5594811829388091,
+      "learning_rate": 8.116580727172071e-06,
+      "loss": 0.1879,
+      "step": 3209
+    },
+    {
+      "epoch": 1.0719652696610453,
+      "grad_norm": 0.4261468140314612,
+      "learning_rate": 8.115060863573729e-06,
+      "loss": 0.194,
+      "step": 3210
+    },
+    {
+      "epoch": 1.0722992152279178,
+      "grad_norm": 0.5033805368157565,
+      "learning_rate": 8.113540529408766e-06,
+      "loss": 0.201,
+      "step": 3211
+    },
+    {
+      "epoch": 1.0726331607947905,
+      "grad_norm": 0.41711740898266775,
+      "learning_rate": 8.112019724906844e-06,
+      "loss": 0.1898,
+      "step": 3212
+    },
+    {
+      "epoch": 1.072967106361663,
+      "grad_norm": 0.4474697888666802,
+      "learning_rate": 8.1104984502977e-06,
+      "loss": 0.2019,
+      "step": 3213
+    },
+    {
+      "epoch": 1.0733010519285355,
+      "grad_norm": 0.4425877508940847,
+      "learning_rate": 8.108976705811138e-06,
+      "loss": 0.204,
+      "step": 3214
+    },
+    {
+      "epoch": 1.0736349974954082,
+      "grad_norm": 0.40031817179384527,
+      "learning_rate": 8.107454491677041e-06,
+      "loss": 0.191,
+      "step": 3215
+    },
+    {
+      "epoch": 1.0739689430622807,
+      "grad_norm": 0.4374992487231899,
+      "learning_rate": 8.10593180812535e-06,
+      "loss": 0.1825,
+      "step": 3216
+    },
+    {
+      "epoch": 1.0743028886291535,
+      "grad_norm": 0.4304869205108331,
+      "learning_rate": 8.104408655386092e-06,
+      "loss": 0.1964,
+      "step": 3217
+    },
+    {
+      "epoch": 1.074636834196026,
+      "grad_norm": 0.4273397458580168,
+      "learning_rate": 8.102885033689352e-06,
+      "loss": 0.1987,
+      "step": 3218
+    },
+    {
+      "epoch": 1.0749707797628987,
+      "grad_norm": 0.44715648855381623,
+      "learning_rate": 8.101360943265293e-06,
+      "loss": 0.2094,
+      "step": 3219
+    },
+    {
+      "epoch": 1.0753047253297712,
+      "grad_norm": 0.433584901382971,
+      "learning_rate": 8.099836384344146e-06,
+      "loss": 0.192,
+      "step": 3220
+    },
+    {
+      "epoch": 1.075638670896644,
+      "grad_norm": 0.45246811449875446,
+      "learning_rate": 8.098311357156213e-06,
+      "loss": 0.2063,
+      "step": 3221
+    },
+    {
+      "epoch": 1.0759726164635164,
+      "grad_norm": 0.3831669548869065,
+      "learning_rate": 8.096785861931868e-06,
+      "loss": 0.1873,
+      "step": 3222
+    },
+    {
+      "epoch": 1.076306562030389,
+      "grad_norm": 0.39589825338369394,
+      "learning_rate": 8.095259898901557e-06,
+      "loss": 0.1884,
+      "step": 3223
+    },
+    {
+      "epoch": 1.0766405075972616,
+      "grad_norm": 0.42988110937230956,
+      "learning_rate": 8.09373346829579e-06,
+      "loss": 0.1989,
+      "step": 3224
+    },
+    {
+      "epoch": 1.0769744531641343,
+      "grad_norm": 0.3861062119239817,
+      "learning_rate": 8.092206570345158e-06,
+      "loss": 0.1876,
+      "step": 3225
+    },
+    {
+      "epoch": 1.0773083987310068,
+      "grad_norm": 0.39497963164084615,
+      "learning_rate": 8.090679205280311e-06,
+      "loss": 0.1875,
+      "step": 3226
+    },
+    {
+      "epoch": 1.0776423442978795,
+      "grad_norm": 0.4204241046414486,
+      "learning_rate": 8.08915137333198e-06,
+      "loss": 0.1985,
+      "step": 3227
+    },
+    {
+      "epoch": 1.077976289864752,
+      "grad_norm": 0.512738178352209,
+      "learning_rate": 8.08762307473096e-06,
+      "loss": 0.2269,
+      "step": 3228
+    },
+    {
+      "epoch": 1.0783102354316247,
+      "grad_norm": 0.44715999802900963,
+      "learning_rate": 8.08609430970812e-06,
+      "loss": 0.2031,
+      "step": 3229
+    },
+    {
+      "epoch": 1.0786441809984972,
+      "grad_norm": 0.3995392929976696,
+      "learning_rate": 8.084565078494396e-06,
+      "loss": 0.1908,
+      "step": 3230
+    },
+    {
+      "epoch": 1.07897812656537,
+      "grad_norm": 0.4132071042034365,
+      "learning_rate": 8.083035381320798e-06,
+      "loss": 0.1984,
+      "step": 3231
+    },
+    {
+      "epoch": 1.0793120721322425,
+      "grad_norm": 0.4507397205246835,
+      "learning_rate": 8.081505218418403e-06,
+      "loss": 0.2041,
+      "step": 3232
+    },
+    {
+      "epoch": 1.079646017699115,
+      "grad_norm": 0.39389622349185976,
+      "learning_rate": 8.079974590018363e-06,
+      "loss": 0.1941,
+      "step": 3233
+    },
+    {
+      "epoch": 1.0799799632659877,
+      "grad_norm": 0.418373210132237,
+      "learning_rate": 8.078443496351893e-06,
+      "loss": 0.1896,
+      "step": 3234
+    },
+    {
+      "epoch": 1.0803139088328602,
+      "grad_norm": 0.44554108624356625,
+      "learning_rate": 8.076911937650288e-06,
+      "loss": 0.1959,
+      "step": 3235
+    },
+    {
+      "epoch": 1.0806478543997329,
+      "grad_norm": 0.3895348321590768,
+      "learning_rate": 8.075379914144902e-06,
+      "loss": 0.1834,
+      "step": 3236
+    },
+    {
+      "epoch": 1.0809817999666054,
+      "grad_norm": 0.41066865301421246,
+      "learning_rate": 8.073847426067172e-06,
+      "loss": 0.1926,
+      "step": 3237
+    },
+    {
+      "epoch": 1.081315745533478,
+      "grad_norm": 0.4979728510699529,
+      "learning_rate": 8.072314473648595e-06,
+      "loss": 0.2136,
+      "step": 3238
+    },
+    {
+      "epoch": 1.0816496911003506,
+      "grad_norm": 0.4147497389444579,
+      "learning_rate": 8.07078105712074e-06,
+      "loss": 0.2014,
+      "step": 3239
+    },
+    {
+      "epoch": 1.0819836366672233,
+      "grad_norm": 0.44728845985893656,
+      "learning_rate": 8.06924717671525e-06,
+      "loss": 0.1971,
+      "step": 3240
+    },
+    {
+      "epoch": 1.0823175822340958,
+      "grad_norm": 0.4179130657469007,
+      "learning_rate": 8.067712832663831e-06,
+      "loss": 0.1949,
+      "step": 3241
+    },
+    {
+      "epoch": 1.0826515278009685,
+      "grad_norm": 0.4107644564937398,
+      "learning_rate": 8.066178025198272e-06,
+      "loss": 0.1877,
+      "step": 3242
+    },
+    {
+      "epoch": 1.082985473367841,
+      "grad_norm": 0.46230738921006503,
+      "learning_rate": 8.064642754550418e-06,
+      "loss": 0.2038,
+      "step": 3243
+    },
+    {
+      "epoch": 1.0833194189347137,
+      "grad_norm": 0.44196147805322944,
+      "learning_rate": 8.06310702095219e-06,
+      "loss": 0.2034,
+      "step": 3244
+    },
+    {
+      "epoch": 1.0836533645015862,
+      "grad_norm": 0.4247083802737399,
+      "learning_rate": 8.06157082463558e-06,
+      "loss": 0.2016,
+      "step": 3245
+    },
+    {
+      "epoch": 1.083987310068459,
+      "grad_norm": 0.4871767192301708,
+      "learning_rate": 8.060034165832648e-06,
+      "loss": 0.2036,
+      "step": 3246
+    },
+    {
+      "epoch": 1.0843212556353314,
+      "grad_norm": 0.44229334769971135,
+      "learning_rate": 8.058497044775526e-06,
+      "loss": 0.2002,
+      "step": 3247
+    },
+    {
+      "epoch": 1.084655201202204,
+      "grad_norm": 0.47425800895867326,
+      "learning_rate": 8.05695946169641e-06,
+      "loss": 0.2031,
+      "step": 3248
+    },
+    {
+      "epoch": 1.0849891467690767,
+      "grad_norm": 0.4145195665483283,
+      "learning_rate": 8.055421416827575e-06,
+      "loss": 0.1965,
+      "step": 3249
+    },
+    {
+      "epoch": 1.0853230923359491,
+      "grad_norm": 0.5037905106010119,
+      "learning_rate": 8.053882910401359e-06,
+      "loss": 0.2002,
+      "step": 3250
+    },
+    {
+      "epoch": 1.0856570379028219,
+      "grad_norm": 0.47633932245924165,
+      "learning_rate": 8.052343942650168e-06,
+      "loss": 0.2055,
+      "step": 3251
+    },
+    {
+      "epoch": 1.0859909834696944,
+      "grad_norm": 0.4534882323600842,
+      "learning_rate": 8.050804513806488e-06,
+      "loss": 0.2039,
+      "step": 3252
+    },
+    {
+      "epoch": 1.086324929036567,
+      "grad_norm": 0.4399351104535666,
+      "learning_rate": 8.049264624102862e-06,
+      "loss": 0.1916,
+      "step": 3253
+    },
+    {
+      "epoch": 1.0866588746034396,
+      "grad_norm": 0.42972776831463017,
+      "learning_rate": 8.047724273771909e-06,
+      "loss": 0.2051,
+      "step": 3254
+    },
+    {
+      "epoch": 1.0869928201703123,
+      "grad_norm": 0.4387648183372129,
+      "learning_rate": 8.046183463046322e-06,
+      "loss": 0.2131,
+      "step": 3255
+    },
+    {
+      "epoch": 1.0873267657371848,
+      "grad_norm": 0.4531319223093814,
+      "learning_rate": 8.044642192158854e-06,
+      "loss": 0.2045,
+      "step": 3256
+    },
+    {
+      "epoch": 1.0876607113040575,
+      "grad_norm": 0.43051407362446054,
+      "learning_rate": 8.043100461342332e-06,
+      "loss": 0.2063,
+      "step": 3257
+    },
+    {
+      "epoch": 1.08799465687093,
+      "grad_norm": 0.4543744781210135,
+      "learning_rate": 8.041558270829655e-06,
+      "loss": 0.1944,
+      "step": 3258
+    },
+    {
+      "epoch": 1.0883286024378027,
+      "grad_norm": 0.5087181521948336,
+      "learning_rate": 8.04001562085379e-06,
+      "loss": 0.2223,
+      "step": 3259
+    },
+    {
+      "epoch": 1.0886625480046752,
+      "grad_norm": 0.4310436393030242,
+      "learning_rate": 8.038472511647768e-06,
+      "loss": 0.1884,
+      "step": 3260
+    },
+    {
+      "epoch": 1.088996493571548,
+      "grad_norm": 0.4297632017698525,
+      "learning_rate": 8.036928943444698e-06,
+      "loss": 0.2041,
+      "step": 3261
+    },
+    {
+      "epoch": 1.0893304391384204,
+      "grad_norm": 0.4827595445998538,
+      "learning_rate": 8.03538491647775e-06,
+      "loss": 0.2051,
+      "step": 3262
+    },
+    {
+      "epoch": 1.089664384705293,
+      "grad_norm": 0.46301803588539736,
+      "learning_rate": 8.03384043098017e-06,
+      "loss": 0.1946,
+      "step": 3263
+    },
+    {
+      "epoch": 1.0899983302721656,
+      "grad_norm": 0.5171554254848715,
+      "learning_rate": 8.032295487185273e-06,
+      "loss": 0.2109,
+      "step": 3264
+    },
+    {
+      "epoch": 1.0903322758390381,
+      "grad_norm": 0.38795308444933635,
+      "learning_rate": 8.030750085326438e-06,
+      "loss": 0.1782,
+      "step": 3265
+    },
+    {
+      "epoch": 1.0906662214059109,
+      "grad_norm": 0.47435267419979654,
+      "learning_rate": 8.029204225637114e-06,
+      "loss": 0.2044,
+      "step": 3266
+    },
+    {
+      "epoch": 1.0910001669727833,
+      "grad_norm": 0.4123685515803338,
+      "learning_rate": 8.027657908350826e-06,
+      "loss": 0.1808,
+      "step": 3267
+    },
+    {
+      "epoch": 1.091334112539656,
+      "grad_norm": 0.4947662241675235,
+      "learning_rate": 8.026111133701162e-06,
+      "loss": 0.2165,
+      "step": 3268
+    },
+    {
+      "epoch": 1.0916680581065286,
+      "grad_norm": 0.5047171583821233,
+      "learning_rate": 8.02456390192178e-06,
+      "loss": 0.2062,
+      "step": 3269
+    },
+    {
+      "epoch": 1.0920020036734013,
+      "grad_norm": 0.41658647064399756,
+      "learning_rate": 8.023016213246406e-06,
+      "loss": 0.1888,
+      "step": 3270
+    },
+    {
+      "epoch": 1.0923359492402738,
+      "grad_norm": 0.44506089965581364,
+      "learning_rate": 8.021468067908839e-06,
+      "loss": 0.2019,
+      "step": 3271
+    },
+    {
+      "epoch": 1.0926698948071465,
+      "grad_norm": 0.4885924662885089,
+      "learning_rate": 8.019919466142945e-06,
+      "loss": 0.1967,
+      "step": 3272
+    },
+    {
+      "epoch": 1.093003840374019,
+      "grad_norm": 0.4729823871926014,
+      "learning_rate": 8.018370408182655e-06,
+      "loss": 0.1982,
+      "step": 3273
+    },
+    {
+      "epoch": 1.0933377859408917,
+      "grad_norm": 0.4531774083310163,
+      "learning_rate": 8.016820894261975e-06,
+      "loss": 0.2072,
+      "step": 3274
+    },
+    {
+      "epoch": 1.0936717315077642,
+      "grad_norm": 0.43365265276494597,
+      "learning_rate": 8.015270924614977e-06,
+      "loss": 0.2068,
+      "step": 3275
+    },
+    {
+      "epoch": 1.094005677074637,
+      "grad_norm": 0.48378575313322314,
+      "learning_rate": 8.013720499475804e-06,
+      "loss": 0.2127,
+      "step": 3276
+    },
+    {
+      "epoch": 1.0943396226415094,
+      "grad_norm": 0.4462044392760605,
+      "learning_rate": 8.012169619078662e-06,
+      "loss": 0.205,
+      "step": 3277
+    },
+    {
+      "epoch": 1.0946735682083821,
+      "grad_norm": 0.4410639604021698,
+      "learning_rate": 8.010618283657834e-06,
+      "loss": 0.2022,
+      "step": 3278
+    },
+    {
+      "epoch": 1.0950075137752546,
+      "grad_norm": 0.41187747484604836,
+      "learning_rate": 8.009066493447664e-06,
+      "loss": 0.1917,
+      "step": 3279
+    },
+    {
+      "epoch": 1.0953414593421273,
+      "grad_norm": 0.4505023943246841,
+      "learning_rate": 8.00751424868257e-06,
+      "loss": 0.2128,
+      "step": 3280
+    },
+    {
+      "epoch": 1.0956754049089998,
+      "grad_norm": 0.44407650340933,
+      "learning_rate": 8.005961549597037e-06,
+      "loss": 0.2023,
+      "step": 3281
+    },
+    {
+      "epoch": 1.0960093504758723,
+      "grad_norm": 0.49103707615481035,
+      "learning_rate": 8.004408396425617e-06,
+      "loss": 0.1892,
+      "step": 3282
+    },
+    {
+      "epoch": 1.096343296042745,
+      "grad_norm": 0.41128125303211827,
+      "learning_rate": 8.002854789402931e-06,
+      "loss": 0.1902,
+      "step": 3283
+    },
+    {
+      "epoch": 1.0966772416096175,
+      "grad_norm": 0.42074924001477537,
+      "learning_rate": 8.001300728763674e-06,
+      "loss": 0.1842,
+      "step": 3284
+    },
+    {
+      "epoch": 1.0970111871764903,
+      "grad_norm": 0.44415785033200067,
+      "learning_rate": 7.999746214742603e-06,
+      "loss": 0.2089,
+      "step": 3285
+    },
+    {
+      "epoch": 1.0973451327433628,
+      "grad_norm": 0.3976674538896059,
+      "learning_rate": 7.998191247574545e-06,
+      "loss": 0.1876,
+      "step": 3286
+    },
+    {
+      "epoch": 1.0976790783102355,
+      "grad_norm": 0.4602221499860677,
+      "learning_rate": 7.996635827494397e-06,
+      "loss": 0.1924,
+      "step": 3287
+    },
+    {
+      "epoch": 1.098013023877108,
+      "grad_norm": 0.4569109408695937,
+      "learning_rate": 7.995079954737122e-06,
+      "loss": 0.2004,
+      "step": 3288
+    },
+    {
+      "epoch": 1.0983469694439807,
+      "grad_norm": 0.46440228075117673,
+      "learning_rate": 7.993523629537753e-06,
+      "loss": 0.201,
+      "step": 3289
+    },
+    {
+      "epoch": 1.0986809150108532,
+      "grad_norm": 0.4210453426591052,
+      "learning_rate": 7.991966852131394e-06,
+      "loss": 0.2009,
+      "step": 3290
+    },
+    {
+      "epoch": 1.099014860577726,
+      "grad_norm": 0.4197516901784847,
+      "learning_rate": 7.990409622753212e-06,
+      "loss": 0.1993,
+      "step": 3291
+    },
+    {
+      "epoch": 1.0993488061445984,
+      "grad_norm": 0.40831891802316234,
+      "learning_rate": 7.988851941638445e-06,
+      "loss": 0.1941,
+      "step": 3292
+    },
+    {
+      "epoch": 1.0996827517114711,
+      "grad_norm": 0.4286089460485976,
+      "learning_rate": 7.987293809022401e-06,
+      "loss": 0.2052,
+      "step": 3293
+    },
+    {
+      "epoch": 1.1000166972783436,
+      "grad_norm": 0.4800654204970883,
+      "learning_rate": 7.985735225140452e-06,
+      "loss": 0.2093,
+      "step": 3294
+    },
+    {
+      "epoch": 1.1003506428452163,
+      "grad_norm": 0.4753445481388503,
+      "learning_rate": 7.984176190228042e-06,
+      "loss": 0.2095,
+      "step": 3295
+    },
+    {
+      "epoch": 1.1006845884120888,
+      "grad_norm": 0.4414696105218096,
+      "learning_rate": 7.98261670452068e-06,
+      "loss": 0.2049,
+      "step": 3296
+    },
+    {
+      "epoch": 1.1010185339789613,
+      "grad_norm": 0.46108006143762986,
+      "learning_rate": 7.981056768253945e-06,
+      "loss": 0.2192,
+      "step": 3297
+    },
+    {
+      "epoch": 1.101352479545834,
+      "grad_norm": 0.40832219511245,
+      "learning_rate": 7.979496381663486e-06,
+      "loss": 0.1899,
+      "step": 3298
+    },
+    {
+      "epoch": 1.1016864251127065,
+      "grad_norm": 0.43698727164514994,
+      "learning_rate": 7.977935544985016e-06,
+      "loss": 0.2032,
+      "step": 3299
+    },
+    {
+      "epoch": 1.1020203706795793,
+      "grad_norm": 0.4212410386989296,
+      "learning_rate": 7.976374258454317e-06,
+      "loss": 0.1885,
+      "step": 3300
+    },
+    {
+      "epoch": 1.1023543162464517,
+      "grad_norm": 0.41194340216926234,
+      "learning_rate": 7.97481252230724e-06,
+      "loss": 0.1988,
+      "step": 3301
+    },
+    {
+      "epoch": 1.1026882618133245,
+      "grad_norm": 0.4276995668094239,
+      "learning_rate": 7.973250336779705e-06,
+      "loss": 0.2041,
+      "step": 3302
+    },
+    {
+      "epoch": 1.103022207380197,
+      "grad_norm": 0.4857427454118398,
+      "learning_rate": 7.971687702107698e-06,
+      "loss": 0.194,
+      "step": 3303
+    },
+    {
+      "epoch": 1.1033561529470697,
+      "grad_norm": 0.4220206495507896,
+      "learning_rate": 7.970124618527274e-06,
+      "loss": 0.1857,
+      "step": 3304
+    },
+    {
+      "epoch": 1.1036900985139422,
+      "grad_norm": 0.4270813488896286,
+      "learning_rate": 7.968561086274553e-06,
+      "loss": 0.1964,
+      "step": 3305
+    },
+    {
+      "epoch": 1.104024044080815,
+      "grad_norm": 0.4252133936392048,
+      "learning_rate": 7.966997105585727e-06,
+      "loss": 0.2049,
+      "step": 3306
+    },
+    {
+      "epoch": 1.1043579896476874,
+      "grad_norm": 0.43526141115105044,
+      "learning_rate": 7.965432676697052e-06,
+      "loss": 0.2081,
+      "step": 3307
+    },
+    {
+      "epoch": 1.10469193521456,
+      "grad_norm": 0.39251872349305744,
+      "learning_rate": 7.963867799844855e-06,
+      "loss": 0.1874,
+      "step": 3308
+    },
+    {
+      "epoch": 1.1050258807814326,
+      "grad_norm": 0.3944749883657532,
+      "learning_rate": 7.962302475265527e-06,
+      "loss": 0.1925,
+      "step": 3309
+    },
+    {
+      "epoch": 1.1053598263483053,
+      "grad_norm": 0.42273055292457995,
+      "learning_rate": 7.960736703195533e-06,
+      "loss": 0.1909,
+      "step": 3310
+    },
+    {
+      "epoch": 1.1056937719151778,
+      "grad_norm": 0.4328304273348919,
+      "learning_rate": 7.959170483871398e-06,
+      "loss": 0.2037,
+      "step": 3311
+    },
+    {
+      "epoch": 1.1060277174820503,
+      "grad_norm": 0.4745377240310588,
+      "learning_rate": 7.957603817529715e-06,
+      "loss": 0.2082,
+      "step": 3312
+    },
+    {
+      "epoch": 1.106361663048923,
+      "grad_norm": 0.45131928288623807,
+      "learning_rate": 7.956036704407153e-06,
+      "loss": 0.1966,
+      "step": 3313
+    },
+    {
+      "epoch": 1.1066956086157955,
+      "grad_norm": 0.41211914551956946,
+      "learning_rate": 7.954469144740441e-06,
+      "loss": 0.194,
+      "step": 3314
+    },
+    {
+      "epoch": 1.1070295541826682,
+      "grad_norm": 0.45784621244927204,
+      "learning_rate": 7.952901138766376e-06,
+      "loss": 0.2016,
+      "step": 3315
+    },
+    {
+      "epoch": 1.1073634997495407,
+      "grad_norm": 0.4387965133352464,
+      "learning_rate": 7.951332686721825e-06,
+      "loss": 0.2084,
+      "step": 3316
+    },
+    {
+      "epoch": 1.1076974453164135,
+      "grad_norm": 0.41301207819634184,
+      "learning_rate": 7.94976378884372e-06,
+      "loss": 0.208,
+      "step": 3317
+    },
+    {
+      "epoch": 1.108031390883286,
+      "grad_norm": 0.4522258424463022,
+      "learning_rate": 7.948194445369065e-06,
+      "loss": 0.2003,
+      "step": 3318
+    },
+    {
+      "epoch": 1.1083653364501587,
+      "grad_norm": 0.45104153739353287,
+      "learning_rate": 7.946624656534922e-06,
+      "loss": 0.2051,
+      "step": 3319
+    },
+    {
+      "epoch": 1.1086992820170312,
+      "grad_norm": 0.47403483171073474,
+      "learning_rate": 7.945054422578432e-06,
+      "loss": 0.2062,
+      "step": 3320
+    },
+    {
+      "epoch": 1.1090332275839039,
+      "grad_norm": 0.44902598801173543,
+      "learning_rate": 7.943483743736793e-06,
+      "loss": 0.2075,
+      "step": 3321
+    },
+    {
+      "epoch": 1.1093671731507764,
+      "grad_norm": 0.4303479990019636,
+      "learning_rate": 7.941912620247276e-06,
+      "loss": 0.1969,
+      "step": 3322
+    },
+    {
+      "epoch": 1.109701118717649,
+      "grad_norm": 0.4960675003356271,
+      "learning_rate": 7.940341052347219e-06,
+      "loss": 0.2157,
+      "step": 3323
+    },
+    {
+      "epoch": 1.1100350642845216,
+      "grad_norm": 0.42395659697291693,
+      "learning_rate": 7.938769040274022e-06,
+      "loss": 0.2068,
+      "step": 3324
+    },
+    {
+      "epoch": 1.1103690098513943,
+      "grad_norm": 0.4340552153546476,
+      "learning_rate": 7.937196584265161e-06,
+      "loss": 0.2016,
+      "step": 3325
+    },
+    {
+      "epoch": 1.1107029554182668,
+      "grad_norm": 0.4117355175562917,
+      "learning_rate": 7.93562368455817e-06,
+      "loss": 0.2022,
+      "step": 3326
+    },
+    {
+      "epoch": 1.1110369009851395,
+      "grad_norm": 0.5466416870551023,
+      "learning_rate": 7.934050341390659e-06,
+      "loss": 0.1812,
+      "step": 3327
+    },
+    {
+      "epoch": 1.111370846552012,
+      "grad_norm": 0.4095963376065435,
+      "learning_rate": 7.932476555000294e-06,
+      "loss": 0.1928,
+      "step": 3328
+    },
+    {
+      "epoch": 1.1117047921188847,
+      "grad_norm": 0.45614123456720185,
+      "learning_rate": 7.930902325624816e-06,
+      "loss": 0.1947,
+      "step": 3329
+    },
+    {
+      "epoch": 1.1120387376857572,
+      "grad_norm": 0.43679448876694665,
+      "learning_rate": 7.929327653502032e-06,
+      "loss": 0.1928,
+      "step": 3330
+    },
+    {
+      "epoch": 1.1123726832526297,
+      "grad_norm": 0.4824447177607817,
+      "learning_rate": 7.927752538869816e-06,
+      "loss": 0.208,
+      "step": 3331
+    },
+    {
+      "epoch": 1.1127066288195024,
+      "grad_norm": 0.4681733368568868,
+      "learning_rate": 7.926176981966102e-06,
+      "loss": 0.2127,
+      "step": 3332
+    },
+    {
+      "epoch": 1.113040574386375,
+      "grad_norm": 0.5039341589751777,
+      "learning_rate": 7.924600983028903e-06,
+      "loss": 0.2128,
+      "step": 3333
+    },
+    {
+      "epoch": 1.1133745199532477,
+      "grad_norm": 0.41043359902653814,
+      "learning_rate": 7.92302454229629e-06,
+      "loss": 0.1901,
+      "step": 3334
+    },
+    {
+      "epoch": 1.1137084655201201,
+      "grad_norm": 0.42173077749314025,
+      "learning_rate": 7.9214476600064e-06,
+      "loss": 0.1946,
+      "step": 3335
+    },
+    {
+      "epoch": 1.1140424110869929,
+      "grad_norm": 0.5686198735094352,
+      "learning_rate": 7.919870336397444e-06,
+      "loss": 0.2045,
+      "step": 3336
+    },
+    {
+      "epoch": 1.1143763566538654,
+      "grad_norm": 0.4263229791855886,
+      "learning_rate": 7.918292571707693e-06,
+      "loss": 0.1998,
+      "step": 3337
+    },
+    {
+      "epoch": 1.114710302220738,
+      "grad_norm": 0.4538265435685427,
+      "learning_rate": 7.916714366175487e-06,
+      "loss": 0.1977,
+      "step": 3338
+    },
+    {
+      "epoch": 1.1150442477876106,
+      "grad_norm": 0.4145724982137613,
+      "learning_rate": 7.915135720039233e-06,
+      "loss": 0.1908,
+      "step": 3339
+    },
+    {
+      "epoch": 1.1153781933544833,
+      "grad_norm": 0.4018593488275227,
+      "learning_rate": 7.913556633537403e-06,
+      "loss": 0.2041,
+      "step": 3340
+    },
+    {
+      "epoch": 1.1157121389213558,
+      "grad_norm": 0.4354152187879273,
+      "learning_rate": 7.91197710690854e-06,
+      "loss": 0.2153,
+      "step": 3341
+    },
+    {
+      "epoch": 1.1160460844882285,
+      "grad_norm": 0.43252317066527857,
+      "learning_rate": 7.910397140391244e-06,
+      "loss": 0.2014,
+      "step": 3342
+    },
+    {
+      "epoch": 1.116380030055101,
+      "grad_norm": 0.4474161800654269,
+      "learning_rate": 7.908816734224195e-06,
+      "loss": 0.2027,
+      "step": 3343
+    },
+    {
+      "epoch": 1.1167139756219737,
+      "grad_norm": 0.4972888829593969,
+      "learning_rate": 7.907235888646126e-06,
+      "loss": 0.22,
+      "step": 3344
+    },
+    {
+      "epoch": 1.1170479211888462,
+      "grad_norm": 0.4505839982249951,
+      "learning_rate": 7.905654603895843e-06,
+      "loss": 0.1939,
+      "step": 3345
+    },
+    {
+      "epoch": 1.1173818667557187,
+      "grad_norm": 0.4374753952465327,
+      "learning_rate": 7.90407288021222e-06,
+      "loss": 0.2152,
+      "step": 3346
+    },
+    {
+      "epoch": 1.1177158123225914,
+      "grad_norm": 0.47875310572864327,
+      "learning_rate": 7.902490717834196e-06,
+      "loss": 0.2035,
+      "step": 3347
+    },
+    {
+      "epoch": 1.118049757889464,
+      "grad_norm": 0.4382931827428161,
+      "learning_rate": 7.90090811700077e-06,
+      "loss": 0.2026,
+      "step": 3348
+    },
+    {
+      "epoch": 1.1183837034563366,
+      "grad_norm": 0.4385411786075883,
+      "learning_rate": 7.899325077951018e-06,
+      "loss": 0.196,
+      "step": 3349
+    },
+    {
+      "epoch": 1.1187176490232091,
+      "grad_norm": 0.38871833531460476,
+      "learning_rate": 7.897741600924073e-06,
+      "loss": 0.1769,
+      "step": 3350
+    },
+    {
+      "epoch": 1.1190515945900819,
+      "grad_norm": 0.4519038045952756,
+      "learning_rate": 7.896157686159142e-06,
+      "loss": 0.2071,
+      "step": 3351
+    },
+    {
+      "epoch": 1.1193855401569544,
+      "grad_norm": 0.47224347629867824,
+      "learning_rate": 7.89457333389549e-06,
+      "loss": 0.2029,
+      "step": 3352
+    },
+    {
+      "epoch": 1.119719485723827,
+      "grad_norm": 0.4201454685987697,
+      "learning_rate": 7.892988544372454e-06,
+      "loss": 0.1991,
+      "step": 3353
+    },
+    {
+      "epoch": 1.1200534312906996,
+      "grad_norm": 0.4289082154934355,
+      "learning_rate": 7.891403317829434e-06,
+      "loss": 0.1932,
+      "step": 3354
+    },
+    {
+      "epoch": 1.1203873768575723,
+      "grad_norm": 0.4758235863184793,
+      "learning_rate": 7.889817654505897e-06,
+      "loss": 0.2088,
+      "step": 3355
+    },
+    {
+      "epoch": 1.1207213224244448,
+      "grad_norm": 0.4198002688938867,
+      "learning_rate": 7.888231554641377e-06,
+      "loss": 0.1923,
+      "step": 3356
+    },
+    {
+      "epoch": 1.1210552679913175,
+      "grad_norm": 0.43029326862973805,
+      "learning_rate": 7.886645018475474e-06,
+      "loss": 0.1915,
+      "step": 3357
+    },
+    {
+      "epoch": 1.12138921355819,
+      "grad_norm": 0.6278196412133785,
+      "learning_rate": 7.885058046247852e-06,
+      "loss": 0.2142,
+      "step": 3358
+    },
+    {
+      "epoch": 1.1217231591250627,
+      "grad_norm": 0.44070049869120914,
+      "learning_rate": 7.88347063819824e-06,
+      "loss": 0.2005,
+      "step": 3359
+    },
+    {
+      "epoch": 1.1220571046919352,
+      "grad_norm": 0.4247481458151347,
+      "learning_rate": 7.881882794566438e-06,
+      "loss": 0.1995,
+      "step": 3360
+    },
+    {
+      "epoch": 1.1223910502588077,
+      "grad_norm": 0.4145120275554937,
+      "learning_rate": 7.880294515592304e-06,
+      "loss": 0.1997,
+      "step": 3361
+    },
+    {
+      "epoch": 1.1227249958256804,
+      "grad_norm": 0.42952334403896586,
+      "learning_rate": 7.878705801515772e-06,
+      "loss": 0.1919,
+      "step": 3362
+    },
+    {
+      "epoch": 1.123058941392553,
+      "grad_norm": 0.43518221391978096,
+      "learning_rate": 7.877116652576832e-06,
+      "loss": 0.1963,
+      "step": 3363
+    },
+    {
+      "epoch": 1.1233928869594256,
+      "grad_norm": 0.4414778182611209,
+      "learning_rate": 7.875527069015545e-06,
+      "loss": 0.2023,
+      "step": 3364
+    },
+    {
+      "epoch": 1.1237268325262981,
+      "grad_norm": 0.48255357811517446,
+      "learning_rate": 7.873937051072037e-06,
+      "loss": 0.1912,
+      "step": 3365
+    },
+    {
+      "epoch": 1.1240607780931708,
+      "grad_norm": 0.4400050750711768,
+      "learning_rate": 7.872346598986496e-06,
+      "loss": 0.1995,
+      "step": 3366
+    },
+    {
+      "epoch": 1.1243947236600433,
+      "grad_norm": 0.4700279659952803,
+      "learning_rate": 7.87075571299918e-06,
+      "loss": 0.2191,
+      "step": 3367
+    },
+    {
+      "epoch": 1.124728669226916,
+      "grad_norm": 0.4577800514967404,
+      "learning_rate": 7.869164393350412e-06,
+      "loss": 0.1821,
+      "step": 3368
+    },
+    {
+      "epoch": 1.1250626147937886,
+      "grad_norm": 0.4447548135499851,
+      "learning_rate": 7.86757264028058e-06,
+      "loss": 0.1968,
+      "step": 3369
+    },
+    {
+      "epoch": 1.1253965603606613,
+      "grad_norm": 0.4270542850533781,
+      "learning_rate": 7.865980454030135e-06,
+      "loss": 0.199,
+      "step": 3370
+    },
+    {
+      "epoch": 1.1257305059275338,
+      "grad_norm": 0.3999805646644059,
+      "learning_rate": 7.864387834839598e-06,
+      "loss": 0.1932,
+      "step": 3371
+    },
+    {
+      "epoch": 1.1260644514944065,
+      "grad_norm": 0.4350361424348412,
+      "learning_rate": 7.86279478294955e-06,
+      "loss": 0.1992,
+      "step": 3372
+    },
+    {
+      "epoch": 1.126398397061279,
+      "grad_norm": 0.4781455786710474,
+      "learning_rate": 7.861201298600642e-06,
+      "loss": 0.2009,
+      "step": 3373
+    },
+    {
+      "epoch": 1.1267323426281517,
+      "grad_norm": 0.4167397575968442,
+      "learning_rate": 7.85960738203359e-06,
+      "loss": 0.1914,
+      "step": 3374
+    },
+    {
+      "epoch": 1.1270662881950242,
+      "grad_norm": 0.5743493123189489,
+      "learning_rate": 7.858013033489171e-06,
+      "loss": 0.2252,
+      "step": 3375
+    },
+    {
+      "epoch": 1.1274002337618967,
+      "grad_norm": 0.4175739966743874,
+      "learning_rate": 7.856418253208232e-06,
+      "loss": 0.2073,
+      "step": 3376
+    },
+    {
+      "epoch": 1.1277341793287694,
+      "grad_norm": 0.4040680795350153,
+      "learning_rate": 7.85482304143168e-06,
+      "loss": 0.1941,
+      "step": 3377
+    },
+    {
+      "epoch": 1.1280681248956421,
+      "grad_norm": 0.42600690189169366,
+      "learning_rate": 7.853227398400495e-06,
+      "loss": 0.1942,
+      "step": 3378
+    },
+    {
+      "epoch": 1.1284020704625146,
+      "grad_norm": 0.3704468500759086,
+      "learning_rate": 7.851631324355717e-06,
+      "loss": 0.1838,
+      "step": 3379
+    },
+    {
+      "epoch": 1.1287360160293871,
+      "grad_norm": 0.4178851626097371,
+      "learning_rate": 7.850034819538448e-06,
+      "loss": 0.2019,
+      "step": 3380
+    },
+    {
+      "epoch": 1.1290699615962598,
+      "grad_norm": 0.41850958354938694,
+      "learning_rate": 7.848437884189864e-06,
+      "loss": 0.1908,
+      "step": 3381
+    },
+    {
+      "epoch": 1.1294039071631323,
+      "grad_norm": 0.46566524574700385,
+      "learning_rate": 7.846840518551197e-06,
+      "loss": 0.2139,
+      "step": 3382
+    },
+    {
+      "epoch": 1.129737852730005,
+      "grad_norm": 0.43426626236451754,
+      "learning_rate": 7.845242722863749e-06,
+      "loss": 0.2025,
+      "step": 3383
+    },
+    {
+      "epoch": 1.1300717982968775,
+      "grad_norm": 0.43143461237116987,
+      "learning_rate": 7.843644497368886e-06,
+      "loss": 0.2002,
+      "step": 3384
+    },
+    {
+      "epoch": 1.1304057438637503,
+      "grad_norm": 0.5135500919403986,
+      "learning_rate": 7.842045842308038e-06,
+      "loss": 0.2073,
+      "step": 3385
+    },
+    {
+      "epoch": 1.1307396894306228,
+      "grad_norm": 0.47400554849908777,
+      "learning_rate": 7.840446757922704e-06,
+      "loss": 0.2046,
+      "step": 3386
+    },
+    {
+      "epoch": 1.1310736349974955,
+      "grad_norm": 0.43000385084533,
+      "learning_rate": 7.838847244454441e-06,
+      "loss": 0.2004,
+      "step": 3387
+    },
+    {
+      "epoch": 1.131407580564368,
+      "grad_norm": 0.44538414882916644,
+      "learning_rate": 7.837247302144874e-06,
+      "loss": 0.1996,
+      "step": 3388
+    },
+    {
+      "epoch": 1.1317415261312407,
+      "grad_norm": 0.4611482843216874,
+      "learning_rate": 7.835646931235697e-06,
+      "loss": 0.2063,
+      "step": 3389
+    },
+    {
+      "epoch": 1.1320754716981132,
+      "grad_norm": 0.4155829516189207,
+      "learning_rate": 7.83404613196866e-06,
+      "loss": 0.1903,
+      "step": 3390
+    },
+    {
+      "epoch": 1.132409417264986,
+      "grad_norm": 0.4401772944803297,
+      "learning_rate": 7.832444904585587e-06,
+      "loss": 0.1983,
+      "step": 3391
+    },
+    {
+      "epoch": 1.1327433628318584,
+      "grad_norm": 0.550134028287982,
+      "learning_rate": 7.83084324932836e-06,
+      "loss": 0.2153,
+      "step": 3392
+    },
+    {
+      "epoch": 1.133077308398731,
+      "grad_norm": 0.438562739435005,
+      "learning_rate": 7.829241166438925e-06,
+      "loss": 0.1966,
+      "step": 3393
+    },
+    {
+      "epoch": 1.1334112539656036,
+      "grad_norm": 0.4134789296728246,
+      "learning_rate": 7.827638656159302e-06,
+      "loss": 0.195,
+      "step": 3394
+    },
+    {
+      "epoch": 1.133745199532476,
+      "grad_norm": 0.4230251736651399,
+      "learning_rate": 7.826035718731564e-06,
+      "loss": 0.2065,
+      "step": 3395
+    },
+    {
+      "epoch": 1.1340791450993488,
+      "grad_norm": 0.4080363331592151,
+      "learning_rate": 7.824432354397857e-06,
+      "loss": 0.1896,
+      "step": 3396
+    },
+    {
+      "epoch": 1.1344130906662213,
+      "grad_norm": 0.5090023027343824,
+      "learning_rate": 7.822828563400384e-06,
+      "loss": 0.1999,
+      "step": 3397
+    },
+    {
+      "epoch": 1.134747036233094,
+      "grad_norm": 0.4361556768676404,
+      "learning_rate": 7.82122434598142e-06,
+      "loss": 0.2119,
+      "step": 3398
+    },
+    {
+      "epoch": 1.1350809817999665,
+      "grad_norm": 0.42843947744351496,
+      "learning_rate": 7.819619702383299e-06,
+      "loss": 0.2032,
+      "step": 3399
+    },
+    {
+      "epoch": 1.1354149273668392,
+      "grad_norm": 0.4308637813200796,
+      "learning_rate": 7.818014632848422e-06,
+      "loss": 0.2194,
+      "step": 3400
+    },
+    {
+      "epoch": 1.1357488729337117,
+      "grad_norm": 0.41419066838076307,
+      "learning_rate": 7.816409137619254e-06,
+      "loss": 0.2013,
+      "step": 3401
+    },
+    {
+      "epoch": 1.1360828185005845,
+      "grad_norm": 0.4978926371404705,
+      "learning_rate": 7.814803216938324e-06,
+      "loss": 0.2085,
+      "step": 3402
+    },
+    {
+      "epoch": 1.136416764067457,
+      "grad_norm": 0.4466119651320874,
+      "learning_rate": 7.813196871048226e-06,
+      "loss": 0.1905,
+      "step": 3403
+    },
+    {
+      "epoch": 1.1367507096343297,
+      "grad_norm": 0.46519231603409,
+      "learning_rate": 7.811590100191613e-06,
+      "loss": 0.2014,
+      "step": 3404
+    },
+    {
+      "epoch": 1.1370846552012022,
+      "grad_norm": 0.45057828935363947,
+      "learning_rate": 7.809982904611213e-06,
+      "loss": 0.2068,
+      "step": 3405
+    },
+    {
+      "epoch": 1.1374186007680749,
+      "grad_norm": 0.4500947824438429,
+      "learning_rate": 7.808375284549807e-06,
+      "loss": 0.2019,
+      "step": 3406
+    },
+    {
+      "epoch": 1.1377525463349474,
+      "grad_norm": 0.4538311873250359,
+      "learning_rate": 7.806767240250248e-06,
+      "loss": 0.2124,
+      "step": 3407
+    },
+    {
+      "epoch": 1.13808649190182,
+      "grad_norm": 0.49784144787499535,
+      "learning_rate": 7.805158771955448e-06,
+      "loss": 0.2027,
+      "step": 3408
+    },
+    {
+      "epoch": 1.1384204374686926,
+      "grad_norm": 0.4619216519515817,
+      "learning_rate": 7.803549879908385e-06,
+      "loss": 0.2047,
+      "step": 3409
+    },
+    {
+      "epoch": 1.138754383035565,
+      "grad_norm": 0.4107807403289291,
+      "learning_rate": 7.801940564352103e-06,
+      "loss": 0.1956,
+      "step": 3410
+    },
+    {
+      "epoch": 1.1390883286024378,
+      "grad_norm": 0.42471264599679803,
+      "learning_rate": 7.800330825529707e-06,
+      "loss": 0.1964,
+      "step": 3411
+    },
+    {
+      "epoch": 1.1394222741693105,
+      "grad_norm": 0.40572323274981525,
+      "learning_rate": 7.798720663684367e-06,
+      "loss": 0.1928,
+      "step": 3412
+    },
+    {
+      "epoch": 1.139756219736183,
+      "grad_norm": 0.4101254365791967,
+      "learning_rate": 7.797110079059315e-06,
+      "loss": 0.2079,
+      "step": 3413
+    },
+    {
+      "epoch": 1.1400901653030555,
+      "grad_norm": 0.4411300893319862,
+      "learning_rate": 7.795499071897855e-06,
+      "loss": 0.2028,
+      "step": 3414
+    },
+    {
+      "epoch": 1.1404241108699282,
+      "grad_norm": 0.4332006753538404,
+      "learning_rate": 7.79388764244334e-06,
+      "loss": 0.2156,
+      "step": 3415
+    },
+    {
+      "epoch": 1.1407580564368007,
+      "grad_norm": 0.41551192652535196,
+      "learning_rate": 7.792275790939202e-06,
+      "loss": 0.2108,
+      "step": 3416
+    },
+    {
+      "epoch": 1.1410920020036734,
+      "grad_norm": 0.4225371412103681,
+      "learning_rate": 7.790663517628927e-06,
+      "loss": 0.1969,
+      "step": 3417
+    },
+    {
+      "epoch": 1.141425947570546,
+      "grad_norm": 0.4391114729864135,
+      "learning_rate": 7.789050822756068e-06,
+      "loss": 0.2064,
+      "step": 3418
+    },
+    {
+      "epoch": 1.1417598931374187,
+      "grad_norm": 0.40251758639113394,
+      "learning_rate": 7.787437706564243e-06,
+      "loss": 0.1882,
+      "step": 3419
+    },
+    {
+      "epoch": 1.1420938387042912,
+      "grad_norm": 0.41418603101346974,
+      "learning_rate": 7.78582416929713e-06,
+      "loss": 0.1919,
+      "step": 3420
+    },
+    {
+      "epoch": 1.1424277842711639,
+      "grad_norm": 0.41423945325619144,
+      "learning_rate": 7.784210211198475e-06,
+      "loss": 0.2046,
+      "step": 3421
+    },
+    {
+      "epoch": 1.1427617298380364,
+      "grad_norm": 0.45911999518986674,
+      "learning_rate": 7.782595832512086e-06,
+      "loss": 0.2102,
+      "step": 3422
+    },
+    {
+      "epoch": 1.143095675404909,
+      "grad_norm": 0.446295856392653,
+      "learning_rate": 7.780981033481832e-06,
+      "loss": 0.2115,
+      "step": 3423
+    },
+    {
+      "epoch": 1.1434296209717816,
+      "grad_norm": 0.47779947300607567,
+      "learning_rate": 7.779365814351648e-06,
+      "loss": 0.212,
+      "step": 3424
+    },
+    {
+      "epoch": 1.143763566538654,
+      "grad_norm": 0.5041795675046281,
+      "learning_rate": 7.77775017536553e-06,
+      "loss": 0.2201,
+      "step": 3425
+    },
+    {
+      "epoch": 1.1440975121055268,
+      "grad_norm": 0.4626546155813629,
+      "learning_rate": 7.776134116767544e-06,
+      "loss": 0.2161,
+      "step": 3426
+    },
+    {
+      "epoch": 1.1444314576723995,
+      "grad_norm": 0.43618214464201066,
+      "learning_rate": 7.774517638801808e-06,
+      "loss": 0.2068,
+      "step": 3427
+    },
+    {
+      "epoch": 1.144765403239272,
+      "grad_norm": 0.4254363039124097,
+      "learning_rate": 7.772900741712516e-06,
+      "loss": 0.2032,
+      "step": 3428
+    },
+    {
+      "epoch": 1.1450993488061445,
+      "grad_norm": 0.45525908259895725,
+      "learning_rate": 7.771283425743916e-06,
+      "loss": 0.2043,
+      "step": 3429
+    },
+    {
+      "epoch": 1.1454332943730172,
+      "grad_norm": 0.41941227927769686,
+      "learning_rate": 7.769665691140325e-06,
+      "loss": 0.2057,
+      "step": 3430
+    },
+    {
+      "epoch": 1.1457672399398897,
+      "grad_norm": 0.4616850221911409,
+      "learning_rate": 7.76804753814612e-06,
+      "loss": 0.1962,
+      "step": 3431
+    },
+    {
+      "epoch": 1.1461011855067624,
+      "grad_norm": 0.44619639348887946,
+      "learning_rate": 7.76642896700574e-06,
+      "loss": 0.1999,
+      "step": 3432
+    },
+    {
+      "epoch": 1.146435131073635,
+      "grad_norm": 0.42643033930129975,
+      "learning_rate": 7.764809977963692e-06,
+      "loss": 0.1966,
+      "step": 3433
+    },
+    {
+      "epoch": 1.1467690766405076,
+      "grad_norm": 0.41159559705421556,
+      "learning_rate": 7.763190571264542e-06,
+      "loss": 0.1919,
+      "step": 3434
+    },
+    {
+      "epoch": 1.1471030222073801,
+      "grad_norm": 0.4303534281592218,
+      "learning_rate": 7.761570747152923e-06,
+      "loss": 0.1985,
+      "step": 3435
+    },
+    {
+      "epoch": 1.1474369677742529,
+      "grad_norm": 0.42350721038806954,
+      "learning_rate": 7.759950505873523e-06,
+      "loss": 0.193,
+      "step": 3436
+    },
+    {
+      "epoch": 1.1477709133411254,
+      "grad_norm": 0.4596329657408379,
+      "learning_rate": 7.758329847671103e-06,
+      "loss": 0.2052,
+      "step": 3437
+    },
+    {
+      "epoch": 1.148104858907998,
+      "grad_norm": 0.4489699249368635,
+      "learning_rate": 7.75670877279048e-06,
+      "loss": 0.2104,
+      "step": 3438
+    },
+    {
+      "epoch": 1.1484388044748706,
+      "grad_norm": 0.43084296877947265,
+      "learning_rate": 7.755087281476539e-06,
+      "loss": 0.198,
+      "step": 3439
+    },
+    {
+      "epoch": 1.1487727500417433,
+      "grad_norm": 0.4057179440867843,
+      "learning_rate": 7.753465373974223e-06,
+      "loss": 0.1999,
+      "step": 3440
+    },
+    {
+      "epoch": 1.1491066956086158,
+      "grad_norm": 0.46218899540177816,
+      "learning_rate": 7.751843050528543e-06,
+      "loss": 0.199,
+      "step": 3441
+    },
+    {
+      "epoch": 1.1494406411754885,
+      "grad_norm": 0.4471325368000403,
+      "learning_rate": 7.750220311384567e-06,
+      "loss": 0.2127,
+      "step": 3442
+    },
+    {
+      "epoch": 1.149774586742361,
+      "grad_norm": 0.4508117782583261,
+      "learning_rate": 7.748597156787429e-06,
+      "loss": 0.2021,
+      "step": 3443
+    },
+    {
+      "epoch": 1.1501085323092335,
+      "grad_norm": 0.42529589186934397,
+      "learning_rate": 7.746973586982328e-06,
+      "loss": 0.2038,
+      "step": 3444
+    },
+    {
+      "epoch": 1.1504424778761062,
+      "grad_norm": 0.4501523886838213,
+      "learning_rate": 7.745349602214522e-06,
+      "loss": 0.2114,
+      "step": 3445
+    },
+    {
+      "epoch": 1.1507764234429787,
+      "grad_norm": 0.44047520372496385,
+      "learning_rate": 7.743725202729335e-06,
+      "loss": 0.1958,
+      "step": 3446
+    },
+    {
+      "epoch": 1.1511103690098514,
+      "grad_norm": 0.4644912732615579,
+      "learning_rate": 7.742100388772148e-06,
+      "loss": 0.2024,
+      "step": 3447
+    },
+    {
+      "epoch": 1.151444314576724,
+      "grad_norm": 0.41485035644284507,
+      "learning_rate": 7.74047516058841e-06,
+      "loss": 0.19,
+      "step": 3448
+    },
+    {
+      "epoch": 1.1517782601435966,
+      "grad_norm": 0.44397761503160493,
+      "learning_rate": 7.73884951842363e-06,
+      "loss": 0.2103,
+      "step": 3449
+    },
+    {
+      "epoch": 1.1521122057104691,
+      "grad_norm": 0.4329722118379211,
+      "learning_rate": 7.737223462523383e-06,
+      "loss": 0.2093,
+      "step": 3450
+    },
+    {
+      "epoch": 1.1524461512773418,
+      "grad_norm": 0.4536966006310295,
+      "learning_rate": 7.735596993133303e-06,
+      "loss": 0.2016,
+      "step": 3451
+    },
+    {
+      "epoch": 1.1527800968442143,
+      "grad_norm": 0.450365844932539,
+      "learning_rate": 7.733970110499086e-06,
+      "loss": 0.2061,
+      "step": 3452
+    },
+    {
+      "epoch": 1.153114042411087,
+      "grad_norm": 0.4556419046795417,
+      "learning_rate": 7.732342814866489e-06,
+      "loss": 0.2023,
+      "step": 3453
+    },
+    {
+      "epoch": 1.1534479879779596,
+      "grad_norm": 0.49031666549692304,
+      "learning_rate": 7.730715106481342e-06,
+      "loss": 0.2149,
+      "step": 3454
+    },
+    {
+      "epoch": 1.1537819335448323,
+      "grad_norm": 0.3880997191032563,
+      "learning_rate": 7.729086985589523e-06,
+      "loss": 0.1861,
+      "step": 3455
+    },
+    {
+      "epoch": 1.1541158791117048,
+      "grad_norm": 0.37937141823654447,
+      "learning_rate": 7.72745845243698e-06,
+      "loss": 0.1863,
+      "step": 3456
+    },
+    {
+      "epoch": 1.1544498246785775,
+      "grad_norm": 0.49966955106951333,
+      "learning_rate": 7.725829507269723e-06,
+      "loss": 0.1955,
+      "step": 3457
+    },
+    {
+      "epoch": 1.15478377024545,
+      "grad_norm": 0.4960476776887278,
+      "learning_rate": 7.724200150333826e-06,
+      "loss": 0.2025,
+      "step": 3458
+    },
+    {
+      "epoch": 1.1551177158123225,
+      "grad_norm": 0.4037417613399222,
+      "learning_rate": 7.722570381875418e-06,
+      "loss": 0.1939,
+      "step": 3459
+    },
+    {
+      "epoch": 1.1554516613791952,
+      "grad_norm": 0.4646572771220113,
+      "learning_rate": 7.720940202140698e-06,
+      "loss": 0.2096,
+      "step": 3460
+    },
+    {
+      "epoch": 1.155785606946068,
+      "grad_norm": 0.40519117416885664,
+      "learning_rate": 7.71930961137592e-06,
+      "loss": 0.1901,
+      "step": 3461
+    },
+    {
+      "epoch": 1.1561195525129404,
+      "grad_norm": 0.45808904281917306,
+      "learning_rate": 7.717678609827409e-06,
+      "loss": 0.2198,
+      "step": 3462
+    },
+    {
+      "epoch": 1.156453498079813,
+      "grad_norm": 0.49491562964628316,
+      "learning_rate": 7.716047197741543e-06,
+      "loss": 0.21,
+      "step": 3463
+    },
+    {
+      "epoch": 1.1567874436466856,
+      "grad_norm": 0.4560900514019919,
+      "learning_rate": 7.714415375364768e-06,
+      "loss": 0.2124,
+      "step": 3464
+    },
+    {
+      "epoch": 1.1571213892135581,
+      "grad_norm": 0.4567473518841226,
+      "learning_rate": 7.712783142943588e-06,
+      "loss": 0.197,
+      "step": 3465
+    },
+    {
+      "epoch": 1.1574553347804308,
+      "grad_norm": 0.45021820733197226,
+      "learning_rate": 7.711150500724574e-06,
+      "loss": 0.2032,
+      "step": 3466
+    },
+    {
+      "epoch": 1.1577892803473033,
+      "grad_norm": 0.43204683278363776,
+      "learning_rate": 7.709517448954353e-06,
+      "loss": 0.1972,
+      "step": 3467
+    },
+    {
+      "epoch": 1.158123225914176,
+      "grad_norm": 0.43096997145471116,
+      "learning_rate": 7.707883987879617e-06,
+      "loss": 0.1837,
+      "step": 3468
+    },
+    {
+      "epoch": 1.1584571714810485,
+      "grad_norm": 0.4107749143716421,
+      "learning_rate": 7.70625011774712e-06,
+      "loss": 0.193,
+      "step": 3469
+    },
+    {
+      "epoch": 1.1587911170479213,
+      "grad_norm": 0.42555377932051347,
+      "learning_rate": 7.70461583880368e-06,
+      "loss": 0.2027,
+      "step": 3470
+    },
+    {
+      "epoch": 1.1591250626147938,
+      "grad_norm": 0.537801799999701,
+      "learning_rate": 7.70298115129617e-06,
+      "loss": 0.2128,
+      "step": 3471
+    },
+    {
+      "epoch": 1.1594590081816665,
+      "grad_norm": 0.4701917219067799,
+      "learning_rate": 7.701346055471533e-06,
+      "loss": 0.1973,
+      "step": 3472
+    },
+    {
+      "epoch": 1.159792953748539,
+      "grad_norm": 0.4336001405965153,
+      "learning_rate": 7.699710551576763e-06,
+      "loss": 0.1959,
+      "step": 3473
+    },
+    {
+      "epoch": 1.1601268993154115,
+      "grad_norm": 0.5109622071728623,
+      "learning_rate": 7.69807463985893e-06,
+      "loss": 0.2068,
+      "step": 3474
+    },
+    {
+      "epoch": 1.1604608448822842,
+      "grad_norm": 0.4399649168868142,
+      "learning_rate": 7.696438320565152e-06,
+      "loss": 0.2018,
+      "step": 3475
+    },
+    {
+      "epoch": 1.160794790449157,
+      "grad_norm": 0.434807916850808,
+      "learning_rate": 7.694801593942615e-06,
+      "loss": 0.1922,
+      "step": 3476
+    },
+    {
+      "epoch": 1.1611287360160294,
+      "grad_norm": 0.4307111475753005,
+      "learning_rate": 7.69316446023857e-06,
+      "loss": 0.1963,
+      "step": 3477
+    },
+    {
+      "epoch": 1.161462681582902,
+      "grad_norm": 0.39136888781514806,
+      "learning_rate": 7.691526919700319e-06,
+      "loss": 0.1892,
+      "step": 3478
+    },
+    {
+      "epoch": 1.1617966271497746,
+      "grad_norm": 0.5072994571524962,
+      "learning_rate": 7.689888972575237e-06,
+      "loss": 0.2129,
+      "step": 3479
+    },
+    {
+      "epoch": 1.162130572716647,
+      "grad_norm": 0.45165856698956547,
+      "learning_rate": 7.688250619110752e-06,
+      "loss": 0.199,
+      "step": 3480
+    },
+    {
+      "epoch": 1.1624645182835198,
+      "grad_norm": 0.47159071816202563,
+      "learning_rate": 7.686611859554361e-06,
+      "loss": 0.2176,
+      "step": 3481
+    },
+    {
+      "epoch": 1.1627984638503923,
+      "grad_norm": 0.42425403478938783,
+      "learning_rate": 7.684972694153612e-06,
+      "loss": 0.2011,
+      "step": 3482
+    },
+    {
+      "epoch": 1.163132409417265,
+      "grad_norm": 0.46430876367122836,
+      "learning_rate": 7.683333123156122e-06,
+      "loss": 0.2024,
+      "step": 3483
+    },
+    {
+      "epoch": 1.1634663549841375,
+      "grad_norm": 0.4252770376389329,
+      "learning_rate": 7.681693146809572e-06,
+      "loss": 0.1989,
+      "step": 3484
+    },
+    {
+      "epoch": 1.1638003005510102,
+      "grad_norm": 0.38138227868488445,
+      "learning_rate": 7.680052765361693e-06,
+      "loss": 0.1877,
+      "step": 3485
+    },
+    {
+      "epoch": 1.1641342461178827,
+      "grad_norm": 0.4621436050450307,
+      "learning_rate": 7.678411979060289e-06,
+      "loss": 0.2138,
+      "step": 3486
+    },
+    {
+      "epoch": 1.1644681916847555,
+      "grad_norm": 0.4910521500257586,
+      "learning_rate": 7.676770788153218e-06,
+      "loss": 0.2029,
+      "step": 3487
+    },
+    {
+      "epoch": 1.164802137251628,
+      "grad_norm": 0.45326505140278767,
+      "learning_rate": 7.6751291928884e-06,
+      "loss": 0.2071,
+      "step": 3488
+    },
+    {
+      "epoch": 1.1651360828185007,
+      "grad_norm": 0.43879315421992415,
+      "learning_rate": 7.673487193513821e-06,
+      "loss": 0.2008,
+      "step": 3489
+    },
+    {
+      "epoch": 1.1654700283853732,
+      "grad_norm": 0.4220316459234618,
+      "learning_rate": 7.671844790277522e-06,
+      "loss": 0.201,
+      "step": 3490
+    },
+    {
+      "epoch": 1.1658039739522459,
+      "grad_norm": 0.4440932041786498,
+      "learning_rate": 7.670201983427606e-06,
+      "loss": 0.2145,
+      "step": 3491
+    },
+    {
+      "epoch": 1.1661379195191184,
+      "grad_norm": 0.42691236313301645,
+      "learning_rate": 7.66855877321224e-06,
+      "loss": 0.2015,
+      "step": 3492
+    },
+    {
+      "epoch": 1.1664718650859909,
+      "grad_norm": 0.501936815427682,
+      "learning_rate": 7.666915159879651e-06,
+      "loss": 0.2119,
+      "step": 3493
+    },
+    {
+      "epoch": 1.1668058106528636,
+      "grad_norm": 0.44911542211817373,
+      "learning_rate": 7.665271143678125e-06,
+      "loss": 0.2143,
+      "step": 3494
+    },
+    {
+      "epoch": 1.167139756219736,
+      "grad_norm": 0.4624678777626138,
+      "learning_rate": 7.66362672485601e-06,
+      "loss": 0.2079,
+      "step": 3495
+    },
+    {
+      "epoch": 1.1674737017866088,
+      "grad_norm": 0.40846524379588184,
+      "learning_rate": 7.661981903661715e-06,
+      "loss": 0.1976,
+      "step": 3496
+    },
+    {
+      "epoch": 1.1678076473534813,
+      "grad_norm": 0.4103460531002227,
+      "learning_rate": 7.66033668034371e-06,
+      "loss": 0.1968,
+      "step": 3497
+    },
+    {
+      "epoch": 1.168141592920354,
+      "grad_norm": 0.4630302942431863,
+      "learning_rate": 7.658691055150524e-06,
+      "loss": 0.214,
+      "step": 3498
+    },
+    {
+      "epoch": 1.1684755384872265,
+      "grad_norm": 0.4578551544166337,
+      "learning_rate": 7.65704502833075e-06,
+      "loss": 0.2043,
+      "step": 3499
+    },
+    {
+      "epoch": 1.1688094840540992,
+      "grad_norm": 0.4524932747516926,
+      "learning_rate": 7.655398600133037e-06,
+      "loss": 0.2279,
+      "step": 3500
+    },
+    {
+      "epoch": 1.1691434296209717,
+      "grad_norm": 0.881428383101394,
+      "learning_rate": 7.653751770806101e-06,
+      "loss": 0.1923,
+      "step": 3501
+    },
+    {
+      "epoch": 1.1694773751878444,
+      "grad_norm": 0.41607736093404346,
+      "learning_rate": 7.652104540598712e-06,
+      "loss": 0.2049,
+      "step": 3502
+    },
+    {
+      "epoch": 1.169811320754717,
+      "grad_norm": 0.40214250677631747,
+      "learning_rate": 7.650456909759707e-06,
+      "loss": 0.1925,
+      "step": 3503
+    },
+    {
+      "epoch": 1.1701452663215897,
+      "grad_norm": 0.4504111125904756,
+      "learning_rate": 7.648808878537976e-06,
+      "loss": 0.192,
+      "step": 3504
+    },
+    {
+      "epoch": 1.1704792118884622,
+      "grad_norm": 0.4602724838722225,
+      "learning_rate": 7.647160447182475e-06,
+      "loss": 0.2087,
+      "step": 3505
+    },
+    {
+      "epoch": 1.1708131574553349,
+      "grad_norm": 0.4104405110762258,
+      "learning_rate": 7.645511615942218e-06,
+      "loss": 0.192,
+      "step": 3506
+    },
+    {
+      "epoch": 1.1711471030222074,
+      "grad_norm": 0.36937151858707484,
+      "learning_rate": 7.643862385066285e-06,
+      "loss": 0.1778,
+      "step": 3507
+    },
+    {
+      "epoch": 1.1714810485890799,
+      "grad_norm": 0.4572947626174961,
+      "learning_rate": 7.642212754803804e-06,
+      "loss": 0.1994,
+      "step": 3508
+    },
+    {
+      "epoch": 1.1718149941559526,
+      "grad_norm": 0.4816706066033821,
+      "learning_rate": 7.640562725403978e-06,
+      "loss": 0.2051,
+      "step": 3509
+    },
+    {
+      "epoch": 1.1721489397228253,
+      "grad_norm": 0.4382076984794106,
+      "learning_rate": 7.638912297116061e-06,
+      "loss": 0.2049,
+      "step": 3510
+    },
+    {
+      "epoch": 1.1724828852896978,
+      "grad_norm": 0.40178546966678536,
+      "learning_rate": 7.637261470189369e-06,
+      "loss": 0.1978,
+      "step": 3511
+    },
+    {
+      "epoch": 1.1728168308565703,
+      "grad_norm": 0.39280539287687943,
+      "learning_rate": 7.635610244873277e-06,
+      "loss": 0.1912,
+      "step": 3512
+    },
+    {
+      "epoch": 1.173150776423443,
+      "grad_norm": 0.46472165624676787,
+      "learning_rate": 7.633958621417226e-06,
+      "loss": 0.215,
+      "step": 3513
+    },
+    {
+      "epoch": 1.1734847219903155,
+      "grad_norm": 0.4297956230414589,
+      "learning_rate": 7.632306600070711e-06,
+      "loss": 0.1893,
+      "step": 3514
+    },
+    {
+      "epoch": 1.1738186675571882,
+      "grad_norm": 0.43762007892015997,
+      "learning_rate": 7.63065418108329e-06,
+      "loss": 0.1916,
+      "step": 3515
+    },
+    {
+      "epoch": 1.1741526131240607,
+      "grad_norm": 0.4897234537598655,
+      "learning_rate": 7.62900136470458e-06,
+      "loss": 0.1981,
+      "step": 3516
+    },
+    {
+      "epoch": 1.1744865586909334,
+      "grad_norm": 0.4051411254394592,
+      "learning_rate": 7.627348151184257e-06,
+      "loss": 0.1918,
+      "step": 3517
+    },
+    {
+      "epoch": 1.174820504257806,
+      "grad_norm": 0.4902705727555018,
+      "learning_rate": 7.625694540772062e-06,
+      "loss": 0.215,
+      "step": 3518
+    },
+    {
+      "epoch": 1.1751544498246786,
+      "grad_norm": 0.459335122777554,
+      "learning_rate": 7.624040533717789e-06,
+      "loss": 0.1931,
+      "step": 3519
+    },
+    {
+      "epoch": 1.1754883953915511,
+      "grad_norm": 0.43076363942331786,
+      "learning_rate": 7.622386130271296e-06,
+      "loss": 0.1907,
+      "step": 3520
+    },
+    {
+      "epoch": 1.1758223409584239,
+      "grad_norm": 0.4783616479824569,
+      "learning_rate": 7.620731330682501e-06,
+      "loss": 0.212,
+      "step": 3521
+    },
+    {
+      "epoch": 1.1761562865252964,
+      "grad_norm": 0.4356060638682757,
+      "learning_rate": 7.6190761352013795e-06,
+      "loss": 0.1951,
+      "step": 3522
+    },
+    {
+      "epoch": 1.1764902320921689,
+      "grad_norm": 0.4594433502156529,
+      "learning_rate": 7.61742054407797e-06,
+      "loss": 0.1982,
+      "step": 3523
+    },
+    {
+      "epoch": 1.1768241776590416,
+      "grad_norm": 0.45490492503768754,
+      "learning_rate": 7.615764557562368e-06,
+      "loss": 0.198,
+      "step": 3524
+    },
+    {
+      "epoch": 1.1771581232259143,
+      "grad_norm": 0.46105614813698614,
+      "learning_rate": 7.6141081759047305e-06,
+      "loss": 0.2014,
+      "step": 3525
+    },
+    {
+      "epoch": 1.1774920687927868,
+      "grad_norm": 0.46026783313250486,
+      "learning_rate": 7.612451399355273e-06,
+      "loss": 0.2189,
+      "step": 3526
+    },
+    {
+      "epoch": 1.1778260143596593,
+      "grad_norm": 0.4356200212691802,
+      "learning_rate": 7.610794228164271e-06,
+      "loss": 0.2029,
+      "step": 3527
+    },
+    {
+      "epoch": 1.178159959926532,
+      "grad_norm": 0.4297337094056724,
+      "learning_rate": 7.60913666258206e-06,
+      "loss": 0.2142,
+      "step": 3528
+    },
+    {
+      "epoch": 1.1784939054934045,
+      "grad_norm": 0.40876834016172153,
+      "learning_rate": 7.6074787028590325e-06,
+      "loss": 0.1937,
+      "step": 3529
+    },
+    {
+      "epoch": 1.1788278510602772,
+      "grad_norm": 0.3885624115347344,
+      "learning_rate": 7.605820349245645e-06,
+      "loss": 0.1974,
+      "step": 3530
+    },
+    {
+      "epoch": 1.1791617966271497,
+      "grad_norm": 0.4338874777406309,
+      "learning_rate": 7.6041616019924125e-06,
+      "loss": 0.1979,
+      "step": 3531
+    },
+    {
+      "epoch": 1.1794957421940224,
+      "grad_norm": 0.4408923442251734,
+      "learning_rate": 7.602502461349907e-06,
+      "loss": 0.2044,
+      "step": 3532
+    },
+    {
+      "epoch": 1.179829687760895,
+      "grad_norm": 0.41283196570865327,
+      "learning_rate": 7.600842927568761e-06,
+      "loss": 0.208,
+      "step": 3533
+    },
+    {
+      "epoch": 1.1801636333277676,
+      "grad_norm": 0.42555424705496336,
+      "learning_rate": 7.599183000899667e-06,
+      "loss": 0.2055,
+      "step": 3534
+    },
+    {
+      "epoch": 1.1804975788946401,
+      "grad_norm": 0.43912943678038574,
+      "learning_rate": 7.597522681593375e-06,
+      "loss": 0.2036,
+      "step": 3535
+    },
+    {
+      "epoch": 1.1808315244615128,
+      "grad_norm": 0.45635857674949853,
+      "learning_rate": 7.595861969900698e-06,
+      "loss": 0.2137,
+      "step": 3536
+    },
+    {
+      "epoch": 1.1811654700283853,
+      "grad_norm": 0.4757314323116078,
+      "learning_rate": 7.5942008660725065e-06,
+      "loss": 0.2046,
+      "step": 3537
+    },
+    {
+      "epoch": 1.181499415595258,
+      "grad_norm": 0.3847032911424942,
+      "learning_rate": 7.5925393703597265e-06,
+      "loss": 0.1854,
+      "step": 3538
+    },
+    {
+      "epoch": 1.1818333611621306,
+      "grad_norm": 0.40804861098671225,
+      "learning_rate": 7.59087748301335e-06,
+      "loss": 0.1961,
+      "step": 3539
+    },
+    {
+      "epoch": 1.1821673067290033,
+      "grad_norm": 0.4048208000344706,
+      "learning_rate": 7.5892152042844224e-06,
+      "loss": 0.201,
+      "step": 3540
+    },
+    {
+      "epoch": 1.1825012522958758,
+      "grad_norm": 0.41873793093657335,
+      "learning_rate": 7.58755253442405e-06,
+      "loss": 0.1955,
+      "step": 3541
+    },
+    {
+      "epoch": 1.1828351978627483,
+      "grad_norm": 0.45334205386003373,
+      "learning_rate": 7.585889473683401e-06,
+      "loss": 0.2101,
+      "step": 3542
+    },
+    {
+      "epoch": 1.183169143429621,
+      "grad_norm": 0.40478766876394096,
+      "learning_rate": 7.5842260223137e-06,
+      "loss": 0.1894,
+      "step": 3543
+    },
+    {
+      "epoch": 1.1835030889964935,
+      "grad_norm": 0.4299142249734497,
+      "learning_rate": 7.5825621805662285e-06,
+      "loss": 0.1943,
+      "step": 3544
+    },
+    {
+      "epoch": 1.1838370345633662,
+      "grad_norm": 0.3919396554368549,
+      "learning_rate": 7.580897948692332e-06,
+      "loss": 0.176,
+      "step": 3545
+    },
+    {
+      "epoch": 1.1841709801302387,
+      "grad_norm": 0.4507907583061385,
+      "learning_rate": 7.579233326943412e-06,
+      "loss": 0.2061,
+      "step": 3546
+    },
+    {
+      "epoch": 1.1845049256971114,
+      "grad_norm": 0.43486901958401214,
+      "learning_rate": 7.577568315570925e-06,
+      "loss": 0.2101,
+      "step": 3547
+    },
+    {
+      "epoch": 1.184838871263984,
+      "grad_norm": 0.611895109414638,
+      "learning_rate": 7.5759029148263975e-06,
+      "loss": 0.2105,
+      "step": 3548
+    },
+    {
+      "epoch": 1.1851728168308566,
+      "grad_norm": 0.46180151479561693,
+      "learning_rate": 7.574237124961403e-06,
+      "loss": 0.1991,
+      "step": 3549
+    },
+    {
+      "epoch": 1.1855067623977291,
+      "grad_norm": 0.40920033333286243,
+      "learning_rate": 7.572570946227582e-06,
+      "loss": 0.1985,
+      "step": 3550
+    },
+    {
+      "epoch": 1.1858407079646018,
+      "grad_norm": 0.5521228822921554,
+      "learning_rate": 7.570904378876627e-06,
+      "loss": 0.2064,
+      "step": 3551
+    },
+    {
+      "epoch": 1.1861746535314743,
+      "grad_norm": 0.4360368544101082,
+      "learning_rate": 7.569237423160294e-06,
+      "loss": 0.1977,
+      "step": 3552
+    },
+    {
+      "epoch": 1.186508599098347,
+      "grad_norm": 0.5055283028151625,
+      "learning_rate": 7.567570079330395e-06,
+      "loss": 0.2105,
+      "step": 3553
+    },
+    {
+      "epoch": 1.1868425446652195,
+      "grad_norm": 0.47535148943957384,
+      "learning_rate": 7.565902347638806e-06,
+      "loss": 0.2198,
+      "step": 3554
+    },
+    {
+      "epoch": 1.1871764902320923,
+      "grad_norm": 0.4567763936303909,
+      "learning_rate": 7.564234228337452e-06,
+      "loss": 0.2059,
+      "step": 3555
+    },
+    {
+      "epoch": 1.1875104357989648,
+      "grad_norm": 0.43216007765801734,
+      "learning_rate": 7.5625657216783276e-06,
+      "loss": 0.2042,
+      "step": 3556
+    },
+    {
+      "epoch": 1.1878443813658373,
+      "grad_norm": 0.4437876561197619,
+      "learning_rate": 7.560896827913478e-06,
+      "loss": 0.2085,
+      "step": 3557
+    },
+    {
+      "epoch": 1.18817832693271,
+      "grad_norm": 0.4136617092910776,
+      "learning_rate": 7.559227547295007e-06,
+      "loss": 0.1918,
+      "step": 3558
+    },
+    {
+      "epoch": 1.1885122724995827,
+      "grad_norm": 0.3950033730804212,
+      "learning_rate": 7.557557880075082e-06,
+      "loss": 0.191,
+      "step": 3559
+    },
+    {
+      "epoch": 1.1888462180664552,
+      "grad_norm": 0.44611070160298427,
+      "learning_rate": 7.555887826505926e-06,
+      "loss": 0.1998,
+      "step": 3560
+    },
+    {
+      "epoch": 1.1891801636333277,
+      "grad_norm": 0.49252383091679314,
+      "learning_rate": 7.554217386839817e-06,
+      "loss": 0.2101,
+      "step": 3561
+    },
+    {
+      "epoch": 1.1895141092002004,
+      "grad_norm": 0.43600290528031405,
+      "learning_rate": 7.552546561329097e-06,
+      "loss": 0.2078,
+      "step": 3562
+    },
+    {
+      "epoch": 1.189848054767073,
+      "grad_norm": 0.4641562946328928,
+      "learning_rate": 7.550875350226166e-06,
+      "loss": 0.2018,
+      "step": 3563
+    },
+    {
+      "epoch": 1.1901820003339456,
+      "grad_norm": 0.4107029251971812,
+      "learning_rate": 7.549203753783475e-06,
+      "loss": 0.1873,
+      "step": 3564
+    },
+    {
+      "epoch": 1.190515945900818,
+      "grad_norm": 0.8638079712764937,
+      "learning_rate": 7.547531772253542e-06,
+      "loss": 0.2183,
+      "step": 3565
+    },
+    {
+      "epoch": 1.1908498914676908,
+      "grad_norm": 0.46739237740310874,
+      "learning_rate": 7.54585940588894e-06,
+      "loss": 0.2061,
+      "step": 3566
+    },
+    {
+      "epoch": 1.1911838370345633,
+      "grad_norm": 0.44207366884569477,
+      "learning_rate": 7.544186654942296e-06,
+      "loss": 0.1888,
+      "step": 3567
+    },
+    {
+      "epoch": 1.191517782601436,
+      "grad_norm": 0.40029726288340856,
+      "learning_rate": 7.542513519666302e-06,
+      "loss": 0.1842,
+      "step": 3568
+    },
+    {
+      "epoch": 1.1918517281683085,
+      "grad_norm": 0.45850744724415704,
+      "learning_rate": 7.540840000313705e-06,
+      "loss": 0.2071,
+      "step": 3569
+    },
+    {
+      "epoch": 1.1921856737351813,
+      "grad_norm": 0.4217447661377353,
+      "learning_rate": 7.539166097137306e-06,
+      "loss": 0.1989,
+      "step": 3570
+    },
+    {
+      "epoch": 1.1925196193020537,
+      "grad_norm": 0.47099508009727703,
+      "learning_rate": 7.537491810389972e-06,
+      "loss": 0.2064,
+      "step": 3571
+    },
+    {
+      "epoch": 1.1928535648689262,
+      "grad_norm": 0.49006517879961325,
+      "learning_rate": 7.535817140324622e-06,
+      "loss": 0.2089,
+      "step": 3572
+    },
+    {
+      "epoch": 1.193187510435799,
+      "grad_norm": 0.45108690927294987,
+      "learning_rate": 7.534142087194234e-06,
+      "loss": 0.2034,
+      "step": 3573
+    },
+    {
+      "epoch": 1.1935214560026717,
+      "grad_norm": 0.40647564236000283,
+      "learning_rate": 7.532466651251846e-06,
+      "loss": 0.1976,
+      "step": 3574
+    },
+    {
+      "epoch": 1.1938554015695442,
+      "grad_norm": 0.4361047207843498,
+      "learning_rate": 7.5307908327505506e-06,
+      "loss": 0.186,
+      "step": 3575
+    },
+    {
+      "epoch": 1.1941893471364167,
+      "grad_norm": 0.41159680206823224,
+      "learning_rate": 7.529114631943501e-06,
+      "loss": 0.18,
+      "step": 3576
+    },
+    {
+      "epoch": 1.1945232927032894,
+      "grad_norm": 0.5154574701726614,
+      "learning_rate": 7.527438049083908e-06,
+      "loss": 0.2203,
+      "step": 3577
+    },
+    {
+      "epoch": 1.1948572382701619,
+      "grad_norm": 0.48966209480112627,
+      "learning_rate": 7.5257610844250385e-06,
+      "loss": 0.2164,
+      "step": 3578
+    },
+    {
+      "epoch": 1.1951911838370346,
+      "grad_norm": 0.484286723912471,
+      "learning_rate": 7.524083738220214e-06,
+      "loss": 0.2147,
+      "step": 3579
+    },
+    {
+      "epoch": 1.195525129403907,
+      "grad_norm": 0.4568042601993156,
+      "learning_rate": 7.522406010722824e-06,
+      "loss": 0.2115,
+      "step": 3580
+    },
+    {
+      "epoch": 1.1958590749707798,
+      "grad_norm": 0.4637639130204421,
+      "learning_rate": 7.5207279021863045e-06,
+      "loss": 0.2011,
+      "step": 3581
+    },
+    {
+      "epoch": 1.1961930205376523,
+      "grad_norm": 0.4503469517122692,
+      "learning_rate": 7.5190494128641545e-06,
+      "loss": 0.2017,
+      "step": 3582
+    },
+    {
+      "epoch": 1.196526966104525,
+      "grad_norm": 0.4115714466848455,
+      "learning_rate": 7.5173705430099295e-06,
+      "loss": 0.1894,
+      "step": 3583
+    },
+    {
+      "epoch": 1.1968609116713975,
+      "grad_norm": 0.43313860284242234,
+      "learning_rate": 7.515691292877243e-06,
+      "loss": 0.2049,
+      "step": 3584
+    },
+    {
+      "epoch": 1.1971948572382702,
+      "grad_norm": 0.44242325330952814,
+      "learning_rate": 7.514011662719766e-06,
+      "loss": 0.2134,
+      "step": 3585
+    },
+    {
+      "epoch": 1.1975288028051427,
+      "grad_norm": 0.49641971012087477,
+      "learning_rate": 7.512331652791226e-06,
+      "loss": 0.2194,
+      "step": 3586
+    },
+    {
+      "epoch": 1.1978627483720155,
+      "grad_norm": 0.4505528169848719,
+      "learning_rate": 7.510651263345408e-06,
+      "loss": 0.2128,
+      "step": 3587
+    },
+    {
+      "epoch": 1.198196693938888,
+      "grad_norm": 0.4207298979058369,
+      "learning_rate": 7.508970494636154e-06,
+      "loss": 0.203,
+      "step": 3588
+    },
+    {
+      "epoch": 1.1985306395057607,
+      "grad_norm": 0.41439317908897016,
+      "learning_rate": 7.507289346917366e-06,
+      "loss": 0.2028,
+      "step": 3589
+    },
+    {
+      "epoch": 1.1988645850726332,
+      "grad_norm": 0.3898376423037865,
+      "learning_rate": 7.505607820442997e-06,
+      "loss": 0.1838,
+      "step": 3590
+    },
+    {
+      "epoch": 1.1991985306395057,
+      "grad_norm": 0.44797911557528114,
+      "learning_rate": 7.503925915467066e-06,
+      "loss": 0.2067,
+      "step": 3591
+    },
+    {
+      "epoch": 1.1995324762063784,
+      "grad_norm": 0.4517517097789117,
+      "learning_rate": 7.502243632243645e-06,
+      "loss": 0.2047,
+      "step": 3592
+    },
+    {
+      "epoch": 1.1998664217732509,
+      "grad_norm": 0.43869540600955803,
+      "learning_rate": 7.500560971026856e-06,
+      "loss": 0.2081,
+      "step": 3593
+    },
+    {
+      "epoch": 1.2002003673401236,
+      "grad_norm": 0.46044286409999025,
+      "learning_rate": 7.498877932070892e-06,
+      "loss": 0.2008,
+      "step": 3594
+    },
+    {
+      "epoch": 1.200534312906996,
+      "grad_norm": 0.46702326825878965,
+      "learning_rate": 7.497194515629992e-06,
+      "loss": 0.1912,
+      "step": 3595
+    },
+    {
+      "epoch": 1.2008682584738688,
+      "grad_norm": 0.42290587241087874,
+      "learning_rate": 7.4955107219584575e-06,
+      "loss": 0.1955,
+      "step": 3596
+    },
+    {
+      "epoch": 1.2012022040407413,
+      "grad_norm": 0.42810888590783125,
+      "learning_rate": 7.493826551310645e-06,
+      "loss": 0.2,
+      "step": 3597
+    },
+    {
+      "epoch": 1.201536149607614,
+      "grad_norm": 0.44645503536762926,
+      "learning_rate": 7.492142003940966e-06,
+      "loss": 0.2086,
+      "step": 3598
+    },
+    {
+      "epoch": 1.2018700951744865,
+      "grad_norm": 0.3919758646583704,
+      "learning_rate": 7.490457080103895e-06,
+      "loss": 0.1833,
+      "step": 3599
+    },
+    {
+      "epoch": 1.2022040407413592,
+      "grad_norm": 0.440006706453243,
+      "learning_rate": 7.4887717800539584e-06,
+      "loss": 0.2122,
+      "step": 3600
+    },
+    {
+      "epoch": 1.2025379863082317,
+      "grad_norm": 0.41535578129544654,
+      "learning_rate": 7.48708610404574e-06,
+      "loss": 0.1889,
+      "step": 3601
+    },
+    {
+      "epoch": 1.2028719318751044,
+      "grad_norm": 0.4473257740232535,
+      "learning_rate": 7.48540005233388e-06,
+      "loss": 0.1918,
+      "step": 3602
+    },
+    {
+      "epoch": 1.203205877441977,
+      "grad_norm": 0.4341216912769296,
+      "learning_rate": 7.483713625173078e-06,
+      "loss": 0.2096,
+      "step": 3603
+    },
+    {
+      "epoch": 1.2035398230088497,
+      "grad_norm": 0.40777171525617034,
+      "learning_rate": 7.482026822818088e-06,
+      "loss": 0.1919,
+      "step": 3604
+    },
+    {
+      "epoch": 1.2038737685757221,
+      "grad_norm": 0.48363102294243154,
+      "learning_rate": 7.480339645523721e-06,
+      "loss": 0.2269,
+      "step": 3605
+    },
+    {
+      "epoch": 1.2042077141425946,
+      "grad_norm": 0.40851845309559937,
+      "learning_rate": 7.478652093544846e-06,
+      "loss": 0.1787,
+      "step": 3606
+    },
+    {
+      "epoch": 1.2045416597094674,
+      "grad_norm": 0.3933853695150573,
+      "learning_rate": 7.476964167136388e-06,
+      "loss": 0.1943,
+      "step": 3607
+    },
+    {
+      "epoch": 1.20487560527634,
+      "grad_norm": 0.4244304751713145,
+      "learning_rate": 7.475275866553326e-06,
+      "loss": 0.2014,
+      "step": 3608
+    },
+    {
+      "epoch": 1.2052095508432126,
+      "grad_norm": 0.6049087445094025,
+      "learning_rate": 7.473587192050698e-06,
+      "loss": 0.2119,
+      "step": 3609
+    },
+    {
+      "epoch": 1.205543496410085,
+      "grad_norm": 0.4302000569946228,
+      "learning_rate": 7.471898143883601e-06,
+      "loss": 0.1866,
+      "step": 3610
+    },
+    {
+      "epoch": 1.2058774419769578,
+      "grad_norm": 0.4329034497056259,
+      "learning_rate": 7.470208722307183e-06,
+      "loss": 0.2069,
+      "step": 3611
+    },
+    {
+      "epoch": 1.2062113875438303,
+      "grad_norm": 0.4175942295887134,
+      "learning_rate": 7.468518927576653e-06,
+      "loss": 0.2029,
+      "step": 3612
+    },
+    {
+      "epoch": 1.206545333110703,
+      "grad_norm": 0.40484000172415735,
+      "learning_rate": 7.466828759947271e-06,
+      "loss": 0.1942,
+      "step": 3613
+    },
+    {
+      "epoch": 1.2068792786775755,
+      "grad_norm": 0.412447371421282,
+      "learning_rate": 7.465138219674359e-06,
+      "loss": 0.2098,
+      "step": 3614
+    },
+    {
+      "epoch": 1.2072132242444482,
+      "grad_norm": 0.4136319006653538,
+      "learning_rate": 7.463447307013294e-06,
+      "loss": 0.1944,
+      "step": 3615
+    },
+    {
+      "epoch": 1.2075471698113207,
+      "grad_norm": 0.4652498470308249,
+      "learning_rate": 7.461756022219507e-06,
+      "loss": 0.2017,
+      "step": 3616
+    },
+    {
+      "epoch": 1.2078811153781934,
+      "grad_norm": 0.451017605411755,
+      "learning_rate": 7.460064365548486e-06,
+      "loss": 0.1925,
+      "step": 3617
+    },
+    {
+      "epoch": 1.208215060945066,
+      "grad_norm": 0.44487372326102165,
+      "learning_rate": 7.458372337255777e-06,
+      "loss": 0.2075,
+      "step": 3618
+    },
+    {
+      "epoch": 1.2085490065119386,
+      "grad_norm": 0.4495806275102939,
+      "learning_rate": 7.45667993759698e-06,
+      "loss": 0.2081,
+      "step": 3619
+    },
+    {
+      "epoch": 1.2088829520788111,
+      "grad_norm": 0.4430843200910173,
+      "learning_rate": 7.454987166827751e-06,
+      "loss": 0.206,
+      "step": 3620
+    },
+    {
+      "epoch": 1.2092168976456836,
+      "grad_norm": 0.4233773457511084,
+      "learning_rate": 7.4532940252038055e-06,
+      "loss": 0.2099,
+      "step": 3621
+    },
+    {
+      "epoch": 1.2095508432125563,
+      "grad_norm": 0.4321928280916543,
+      "learning_rate": 7.45160051298091e-06,
+      "loss": 0.1958,
+      "step": 3622
+    },
+    {
+      "epoch": 1.209884788779429,
+      "grad_norm": 0.5607630429277455,
+      "learning_rate": 7.4499066304148904e-06,
+      "loss": 0.2037,
+      "step": 3623
+    },
+    {
+      "epoch": 1.2102187343463016,
+      "grad_norm": 0.4184204477068433,
+      "learning_rate": 7.448212377761628e-06,
+      "loss": 0.2049,
+      "step": 3624
+    },
+    {
+      "epoch": 1.210552679913174,
+      "grad_norm": 0.4286820509175345,
+      "learning_rate": 7.4465177552770585e-06,
+      "loss": 0.2065,
+      "step": 3625
+    },
+    {
+      "epoch": 1.2108866254800468,
+      "grad_norm": 0.4204690969276894,
+      "learning_rate": 7.444822763217174e-06,
+      "loss": 0.1969,
+      "step": 3626
+    },
+    {
+      "epoch": 1.2112205710469193,
+      "grad_norm": 0.45133162737457627,
+      "learning_rate": 7.443127401838026e-06,
+      "loss": 0.2121,
+      "step": 3627
+    },
+    {
+      "epoch": 1.211554516613792,
+      "grad_norm": 0.6038373572960923,
+      "learning_rate": 7.441431671395717e-06,
+      "loss": 0.2025,
+      "step": 3628
+    },
+    {
+      "epoch": 1.2118884621806645,
+      "grad_norm": 0.46725093116741995,
+      "learning_rate": 7.439735572146407e-06,
+      "loss": 0.2076,
+      "step": 3629
+    },
+    {
+      "epoch": 1.2122224077475372,
+      "grad_norm": 0.4469655909074122,
+      "learning_rate": 7.438039104346312e-06,
+      "loss": 0.1991,
+      "step": 3630
+    },
+    {
+      "epoch": 1.2125563533144097,
+      "grad_norm": 0.4339059574842912,
+      "learning_rate": 7.436342268251702e-06,
+      "loss": 0.1954,
+      "step": 3631
+    },
+    {
+      "epoch": 1.2128902988812824,
+      "grad_norm": 0.47586137866658257,
+      "learning_rate": 7.434645064118906e-06,
+      "loss": 0.2117,
+      "step": 3632
+    },
+    {
+      "epoch": 1.213224244448155,
+      "grad_norm": 0.39511341244546483,
+      "learning_rate": 7.432947492204308e-06,
+      "loss": 0.1867,
+      "step": 3633
+    },
+    {
+      "epoch": 1.2135581900150276,
+      "grad_norm": 0.42677190119791664,
+      "learning_rate": 7.431249552764342e-06,
+      "loss": 0.2021,
+      "step": 3634
+    },
+    {
+      "epoch": 1.2138921355819001,
+      "grad_norm": 0.4669699424010275,
+      "learning_rate": 7.429551246055504e-06,
+      "loss": 0.1998,
+      "step": 3635
+    },
+    {
+      "epoch": 1.2142260811487728,
+      "grad_norm": 0.4174572295361658,
+      "learning_rate": 7.427852572334344e-06,
+      "loss": 0.2047,
+      "step": 3636
+    },
+    {
+      "epoch": 1.2145600267156453,
+      "grad_norm": 0.49130052281682524,
+      "learning_rate": 7.426153531857466e-06,
+      "loss": 0.2126,
+      "step": 3637
+    },
+    {
+      "epoch": 1.214893972282518,
+      "grad_norm": 0.39466608599092157,
+      "learning_rate": 7.424454124881531e-06,
+      "loss": 0.1912,
+      "step": 3638
+    },
+    {
+      "epoch": 1.2152279178493905,
+      "grad_norm": 0.5280861811057292,
+      "learning_rate": 7.422754351663252e-06,
+      "loss": 0.2041,
+      "step": 3639
+    },
+    {
+      "epoch": 1.215561863416263,
+      "grad_norm": 0.42326684754756433,
+      "learning_rate": 7.4210542124594e-06,
+      "loss": 0.1907,
+      "step": 3640
+    },
+    {
+      "epoch": 1.2158958089831358,
+      "grad_norm": 0.44549725989101097,
+      "learning_rate": 7.419353707526804e-06,
+      "loss": 0.2083,
+      "step": 3641
+    },
+    {
+      "epoch": 1.2162297545500083,
+      "grad_norm": 0.43094208142967383,
+      "learning_rate": 7.417652837122345e-06,
+      "loss": 0.2167,
+      "step": 3642
+    },
+    {
+      "epoch": 1.216563700116881,
+      "grad_norm": 0.5041095084876982,
+      "learning_rate": 7.4159516015029545e-06,
+      "loss": 0.2003,
+      "step": 3643
+    },
+    {
+      "epoch": 1.2168976456837535,
+      "grad_norm": 0.415265396940736,
+      "learning_rate": 7.414250000925629e-06,
+      "loss": 0.1979,
+      "step": 3644
+    },
+    {
+      "epoch": 1.2172315912506262,
+      "grad_norm": 0.4362697950301838,
+      "learning_rate": 7.412548035647416e-06,
+      "loss": 0.2035,
+      "step": 3645
+    },
+    {
+      "epoch": 1.2175655368174987,
+      "grad_norm": 0.3976738370024975,
+      "learning_rate": 7.4108457059254135e-06,
+      "loss": 0.1947,
+      "step": 3646
+    },
+    {
+      "epoch": 1.2178994823843714,
+      "grad_norm": 0.44831231308322833,
+      "learning_rate": 7.40914301201678e-06,
+      "loss": 0.2095,
+      "step": 3647
+    },
+    {
+      "epoch": 1.218233427951244,
+      "grad_norm": 0.38688375643877443,
+      "learning_rate": 7.407439954178729e-06,
+      "loss": 0.1935,
+      "step": 3648
+    },
+    {
+      "epoch": 1.2185673735181166,
+      "grad_norm": 0.42809394482498125,
+      "learning_rate": 7.405736532668525e-06,
+      "loss": 0.1997,
+      "step": 3649
+    },
+    {
+      "epoch": 1.218901319084989,
+      "grad_norm": 0.4272147459706146,
+      "learning_rate": 7.4040327477434926e-06,
+      "loss": 0.1933,
+      "step": 3650
+    },
+    {
+      "epoch": 1.2192352646518618,
+      "grad_norm": 0.4233262886986238,
+      "learning_rate": 7.402328599661006e-06,
+      "loss": 0.1961,
+      "step": 3651
+    },
+    {
+      "epoch": 1.2195692102187343,
+      "grad_norm": 0.4380339243339666,
+      "learning_rate": 7.400624088678497e-06,
+      "loss": 0.202,
+      "step": 3652
+    },
+    {
+      "epoch": 1.219903155785607,
+      "grad_norm": 0.4327575082842074,
+      "learning_rate": 7.398919215053455e-06,
+      "loss": 0.205,
+      "step": 3653
+    },
+    {
+      "epoch": 1.2202371013524795,
+      "grad_norm": 0.4371121995452199,
+      "learning_rate": 7.397213979043418e-06,
+      "loss": 0.2089,
+      "step": 3654
+    },
+    {
+      "epoch": 1.220571046919352,
+      "grad_norm": 0.4384003879413628,
+      "learning_rate": 7.395508380905983e-06,
+      "loss": 0.2007,
+      "step": 3655
+    },
+    {
+      "epoch": 1.2209049924862247,
+      "grad_norm": 0.42353068846381453,
+      "learning_rate": 7.393802420898801e-06,
+      "loss": 0.1949,
+      "step": 3656
+    },
+    {
+      "epoch": 1.2212389380530975,
+      "grad_norm": 0.388881556989798,
+      "learning_rate": 7.392096099279579e-06,
+      "loss": 0.1945,
+      "step": 3657
+    },
+    {
+      "epoch": 1.22157288361997,
+      "grad_norm": 0.4694668160111598,
+      "learning_rate": 7.390389416306073e-06,
+      "loss": 0.1988,
+      "step": 3658
+    },
+    {
+      "epoch": 1.2219068291868425,
+      "grad_norm": 0.41573848399930674,
+      "learning_rate": 7.3886823722361e-06,
+      "loss": 0.195,
+      "step": 3659
+    },
+    {
+      "epoch": 1.2222407747537152,
+      "grad_norm": 0.4196740807199658,
+      "learning_rate": 7.386974967327531e-06,
+      "loss": 0.1941,
+      "step": 3660
+    },
+    {
+      "epoch": 1.2225747203205877,
+      "grad_norm": 0.3842082607552933,
+      "learning_rate": 7.385267201838284e-06,
+      "loss": 0.1816,
+      "step": 3661
+    },
+    {
+      "epoch": 1.2229086658874604,
+      "grad_norm": 0.4210064079129366,
+      "learning_rate": 7.383559076026343e-06,
+      "loss": 0.2085,
+      "step": 3662
+    },
+    {
+      "epoch": 1.2232426114543329,
+      "grad_norm": 0.40011050080472454,
+      "learning_rate": 7.381850590149737e-06,
+      "loss": 0.1921,
+      "step": 3663
+    },
+    {
+      "epoch": 1.2235765570212056,
+      "grad_norm": 0.41455617156730395,
+      "learning_rate": 7.380141744466555e-06,
+      "loss": 0.1862,
+      "step": 3664
+    },
+    {
+      "epoch": 1.223910502588078,
+      "grad_norm": 0.42642811953196785,
+      "learning_rate": 7.378432539234936e-06,
+      "loss": 0.1976,
+      "step": 3665
+    },
+    {
+      "epoch": 1.2242444481549508,
+      "grad_norm": 0.4540552515895627,
+      "learning_rate": 7.376722974713078e-06,
+      "loss": 0.201,
+      "step": 3666
+    },
+    {
+      "epoch": 1.2245783937218233,
+      "grad_norm": 0.4442490667004041,
+      "learning_rate": 7.3750130511592275e-06,
+      "loss": 0.1957,
+      "step": 3667
+    },
+    {
+      "epoch": 1.224912339288696,
+      "grad_norm": 0.49385689080880146,
+      "learning_rate": 7.373302768831694e-06,
+      "loss": 0.2133,
+      "step": 3668
+    },
+    {
+      "epoch": 1.2252462848555685,
+      "grad_norm": 0.434217175678413,
+      "learning_rate": 7.371592127988831e-06,
+      "loss": 0.2016,
+      "step": 3669
+    },
+    {
+      "epoch": 1.225580230422441,
+      "grad_norm": 0.39530828895439507,
+      "learning_rate": 7.369881128889052e-06,
+      "loss": 0.1923,
+      "step": 3670
+    },
+    {
+      "epoch": 1.2259141759893137,
+      "grad_norm": 0.40983470561623925,
+      "learning_rate": 7.368169771790825e-06,
+      "loss": 0.2084,
+      "step": 3671
+    },
+    {
+      "epoch": 1.2262481215561865,
+      "grad_norm": 0.4571876738058572,
+      "learning_rate": 7.366458056952668e-06,
+      "loss": 0.1997,
+      "step": 3672
+    },
+    {
+      "epoch": 1.226582067123059,
+      "grad_norm": 0.426026877765367,
+      "learning_rate": 7.36474598463316e-06,
+      "loss": 0.2005,
+      "step": 3673
+    },
+    {
+      "epoch": 1.2269160126899314,
+      "grad_norm": 0.4118963504455984,
+      "learning_rate": 7.363033555090925e-06,
+      "loss": 0.1886,
+      "step": 3674
+    },
+    {
+      "epoch": 1.2272499582568042,
+      "grad_norm": 0.40449592628596953,
+      "learning_rate": 7.361320768584648e-06,
+      "loss": 0.1994,
+      "step": 3675
+    },
+    {
+      "epoch": 1.2275839038236767,
+      "grad_norm": 0.4197205384614,
+      "learning_rate": 7.359607625373065e-06,
+      "loss": 0.1891,
+      "step": 3676
+    },
+    {
+      "epoch": 1.2279178493905494,
+      "grad_norm": 0.4325982836766197,
+      "learning_rate": 7.357894125714967e-06,
+      "loss": 0.2011,
+      "step": 3677
+    },
+    {
+      "epoch": 1.2282517949574219,
+      "grad_norm": 0.44776808435438165,
+      "learning_rate": 7.3561802698691976e-06,
+      "loss": 0.1929,
+      "step": 3678
+    },
+    {
+      "epoch": 1.2285857405242946,
+      "grad_norm": 0.4529997887257035,
+      "learning_rate": 7.354466058094656e-06,
+      "loss": 0.2019,
+      "step": 3679
+    },
+    {
+      "epoch": 1.228919686091167,
+      "grad_norm": 0.4780531134449325,
+      "learning_rate": 7.352751490650294e-06,
+      "loss": 0.2334,
+      "step": 3680
+    },
+    {
+      "epoch": 1.2292536316580398,
+      "grad_norm": 0.4327464219707674,
+      "learning_rate": 7.3510365677951155e-06,
+      "loss": 0.1923,
+      "step": 3681
+    },
+    {
+      "epoch": 1.2295875772249123,
+      "grad_norm": 0.4494358023469274,
+      "learning_rate": 7.349321289788181e-06,
+      "loss": 0.2073,
+      "step": 3682
+    },
+    {
+      "epoch": 1.229921522791785,
+      "grad_norm": 0.4186072999519254,
+      "learning_rate": 7.3476056568886036e-06,
+      "loss": 0.1897,
+      "step": 3683
+    },
+    {
+      "epoch": 1.2302554683586575,
+      "grad_norm": 0.4505587151771832,
+      "learning_rate": 7.34588966935555e-06,
+      "loss": 0.2074,
+      "step": 3684
+    },
+    {
+      "epoch": 1.2305894139255302,
+      "grad_norm": 0.4712172452986157,
+      "learning_rate": 7.344173327448238e-06,
+      "loss": 0.1955,
+      "step": 3685
+    },
+    {
+      "epoch": 1.2309233594924027,
+      "grad_norm": 0.42151892488420034,
+      "learning_rate": 7.342456631425945e-06,
+      "loss": 0.1968,
+      "step": 3686
+    },
+    {
+      "epoch": 1.2312573050592754,
+      "grad_norm": 0.4858147778552328,
+      "learning_rate": 7.340739581547996e-06,
+      "loss": 0.2112,
+      "step": 3687
+    },
+    {
+      "epoch": 1.231591250626148,
+      "grad_norm": 0.39299160270288175,
+      "learning_rate": 7.339022178073772e-06,
+      "loss": 0.1783,
+      "step": 3688
+    },
+    {
+      "epoch": 1.2319251961930204,
+      "grad_norm": 0.4347870834866963,
+      "learning_rate": 7.337304421262706e-06,
+      "loss": 0.2005,
+      "step": 3689
+    },
+    {
+      "epoch": 1.2322591417598932,
+      "grad_norm": 0.3944954395837461,
+      "learning_rate": 7.335586311374287e-06,
+      "loss": 0.1804,
+      "step": 3690
+    },
+    {
+      "epoch": 1.2325930873267656,
+      "grad_norm": 0.4511832507539604,
+      "learning_rate": 7.3338678486680545e-06,
+      "loss": 0.226,
+      "step": 3691
+    },
+    {
+      "epoch": 1.2329270328936384,
+      "grad_norm": 0.4951489024207722,
+      "learning_rate": 7.3321490334036035e-06,
+      "loss": 0.233,
+      "step": 3692
+    },
+    {
+      "epoch": 1.2332609784605109,
+      "grad_norm": 0.42695922666442165,
+      "learning_rate": 7.3304298658405815e-06,
+      "loss": 0.1988,
+      "step": 3693
+    },
+    {
+      "epoch": 1.2335949240273836,
+      "grad_norm": 0.4559841391821671,
+      "learning_rate": 7.328710346238688e-06,
+      "loss": 0.1873,
+      "step": 3694
+    },
+    {
+      "epoch": 1.233928869594256,
+      "grad_norm": 0.4176584911033391,
+      "learning_rate": 7.326990474857676e-06,
+      "loss": 0.1755,
+      "step": 3695
+    },
+    {
+      "epoch": 1.2342628151611288,
+      "grad_norm": 0.388833205756847,
+      "learning_rate": 7.3252702519573545e-06,
+      "loss": 0.1818,
+      "step": 3696
+    },
+    {
+      "epoch": 1.2345967607280013,
+      "grad_norm": 0.4741753202108229,
+      "learning_rate": 7.323549677797582e-06,
+      "loss": 0.2035,
+      "step": 3697
+    },
+    {
+      "epoch": 1.234930706294874,
+      "grad_norm": 0.41197049335397234,
+      "learning_rate": 7.3218287526382716e-06,
+      "loss": 0.1884,
+      "step": 3698
+    },
+    {
+      "epoch": 1.2352646518617465,
+      "grad_norm": 0.4519660783569149,
+      "learning_rate": 7.320107476739389e-06,
+      "loss": 0.1987,
+      "step": 3699
+    },
+    {
+      "epoch": 1.2355985974286192,
+      "grad_norm": 0.4441870961138195,
+      "learning_rate": 7.318385850360954e-06,
+      "loss": 0.2002,
+      "step": 3700
+    },
+    {
+      "epoch": 1.2359325429954917,
+      "grad_norm": 0.4771636776284061,
+      "learning_rate": 7.316663873763039e-06,
+      "loss": 0.2005,
+      "step": 3701
+    },
+    {
+      "epoch": 1.2362664885623644,
+      "grad_norm": 0.421346363486521,
+      "learning_rate": 7.314941547205767e-06,
+      "loss": 0.2057,
+      "step": 3702
+    },
+    {
+      "epoch": 1.236600434129237,
+      "grad_norm": 0.3981895128659755,
+      "learning_rate": 7.313218870949317e-06,
+      "loss": 0.1943,
+      "step": 3703
+    },
+    {
+      "epoch": 1.2369343796961094,
+      "grad_norm": 0.43622076090027484,
+      "learning_rate": 7.31149584525392e-06,
+      "loss": 0.2047,
+      "step": 3704
+    },
+    {
+      "epoch": 1.2372683252629821,
+      "grad_norm": 0.39492812624758417,
+      "learning_rate": 7.309772470379856e-06,
+      "loss": 0.2045,
+      "step": 3705
+    },
+    {
+      "epoch": 1.2376022708298549,
+      "grad_norm": 0.39980899433349076,
+      "learning_rate": 7.308048746587466e-06,
+      "loss": 0.1904,
+      "step": 3706
+    },
+    {
+      "epoch": 1.2379362163967274,
+      "grad_norm": 0.4226276960308284,
+      "learning_rate": 7.3063246741371365e-06,
+      "loss": 0.1934,
+      "step": 3707
+    },
+    {
+      "epoch": 1.2382701619635998,
+      "grad_norm": 0.4416996943577303,
+      "learning_rate": 7.304600253289308e-06,
+      "loss": 0.1986,
+      "step": 3708
+    },
+    {
+      "epoch": 1.2386041075304726,
+      "grad_norm": 0.4084631963077389,
+      "learning_rate": 7.302875484304476e-06,
+      "loss": 0.2003,
+      "step": 3709
+    },
+    {
+      "epoch": 1.238938053097345,
+      "grad_norm": 0.43539905341193225,
+      "learning_rate": 7.301150367443186e-06,
+      "loss": 0.2037,
+      "step": 3710
+    },
+    {
+      "epoch": 1.2392719986642178,
+      "grad_norm": 0.4253528674527778,
+      "learning_rate": 7.299424902966039e-06,
+      "loss": 0.2084,
+      "step": 3711
+    },
+    {
+      "epoch": 1.2396059442310903,
+      "grad_norm": 0.4139818088928367,
+      "learning_rate": 7.297699091133685e-06,
+      "loss": 0.2113,
+      "step": 3712
+    },
+    {
+      "epoch": 1.239939889797963,
+      "grad_norm": 0.42337946777331986,
+      "learning_rate": 7.295972932206827e-06,
+      "loss": 0.1946,
+      "step": 3713
+    },
+    {
+      "epoch": 1.2402738353648355,
+      "grad_norm": 0.42234185270273344,
+      "learning_rate": 7.2942464264462255e-06,
+      "loss": 0.1877,
+      "step": 3714
+    },
+    {
+      "epoch": 1.2406077809317082,
+      "grad_norm": 0.42468102269690094,
+      "learning_rate": 7.292519574112688e-06,
+      "loss": 0.1952,
+      "step": 3715
+    },
+    {
+      "epoch": 1.2409417264985807,
+      "grad_norm": 0.45061158016324065,
+      "learning_rate": 7.290792375467074e-06,
+      "loss": 0.2077,
+      "step": 3716
+    },
+    {
+      "epoch": 1.2412756720654534,
+      "grad_norm": 0.4374334755090788,
+      "learning_rate": 7.2890648307702985e-06,
+      "loss": 0.208,
+      "step": 3717
+    },
+    {
+      "epoch": 1.241609617632326,
+      "grad_norm": 0.44401302086261285,
+      "learning_rate": 7.287336940283327e-06,
+      "loss": 0.2005,
+      "step": 3718
+    },
+    {
+      "epoch": 1.2419435631991984,
+      "grad_norm": 0.4366476902185395,
+      "learning_rate": 7.28560870426718e-06,
+      "loss": 0.2007,
+      "step": 3719
+    },
+    {
+      "epoch": 1.2422775087660711,
+      "grad_norm": 0.43419928883049946,
+      "learning_rate": 7.2838801229829245e-06,
+      "loss": 0.1975,
+      "step": 3720
+    },
+    {
+      "epoch": 1.2426114543329438,
+      "grad_norm": 0.40640252159976503,
+      "learning_rate": 7.2821511966916845e-06,
+      "loss": 0.1994,
+      "step": 3721
+    },
+    {
+      "epoch": 1.2429453998998163,
+      "grad_norm": 0.42257019966830034,
+      "learning_rate": 7.280421925654635e-06,
+      "loss": 0.2021,
+      "step": 3722
+    },
+    {
+      "epoch": 1.2432793454666888,
+      "grad_norm": 0.40500129240346916,
+      "learning_rate": 7.278692310133003e-06,
+      "loss": 0.1919,
+      "step": 3723
+    },
+    {
+      "epoch": 1.2436132910335616,
+      "grad_norm": 0.42012490022988247,
+      "learning_rate": 7.276962350388067e-06,
+      "loss": 0.1966,
+      "step": 3724
+    },
+    {
+      "epoch": 1.243947236600434,
+      "grad_norm": 0.4029324763396112,
+      "learning_rate": 7.275232046681157e-06,
+      "loss": 0.1896,
+      "step": 3725
+    },
+    {
+      "epoch": 1.2442811821673068,
+      "grad_norm": 0.4113849025835322,
+      "learning_rate": 7.273501399273656e-06,
+      "loss": 0.1969,
+      "step": 3726
+    },
+    {
+      "epoch": 1.2446151277341793,
+      "grad_norm": 0.4507044786866005,
+      "learning_rate": 7.271770408427e-06,
+      "loss": 0.2062,
+      "step": 3727
+    },
+    {
+      "epoch": 1.244949073301052,
+      "grad_norm": 0.45701381903540794,
+      "learning_rate": 7.2700390744026735e-06,
+      "loss": 0.2079,
+      "step": 3728
+    },
+    {
+      "epoch": 1.2452830188679245,
+      "grad_norm": 0.47264780094018927,
+      "learning_rate": 7.2683073974622165e-06,
+      "loss": 0.21,
+      "step": 3729
+    },
+    {
+      "epoch": 1.2456169644347972,
+      "grad_norm": 0.46485397157026925,
+      "learning_rate": 7.26657537786722e-06,
+      "loss": 0.2165,
+      "step": 3730
+    },
+    {
+      "epoch": 1.2459509100016697,
+      "grad_norm": 0.40938039577104884,
+      "learning_rate": 7.264843015879321e-06,
+      "loss": 0.1923,
+      "step": 3731
+    },
+    {
+      "epoch": 1.2462848555685424,
+      "grad_norm": 0.5185887899593448,
+      "learning_rate": 7.263110311760221e-06,
+      "loss": 0.2086,
+      "step": 3732
+    },
+    {
+      "epoch": 1.246618801135415,
+      "grad_norm": 0.4628231707028568,
+      "learning_rate": 7.2613772657716585e-06,
+      "loss": 0.1927,
+      "step": 3733
+    },
+    {
+      "epoch": 1.2469527467022876,
+      "grad_norm": 0.49261373578200623,
+      "learning_rate": 7.259643878175434e-06,
+      "loss": 0.2183,
+      "step": 3734
+    },
+    {
+      "epoch": 1.2472866922691601,
+      "grad_norm": 0.4642674170595037,
+      "learning_rate": 7.2579101492333956e-06,
+      "loss": 0.194,
+      "step": 3735
+    },
+    {
+      "epoch": 1.2476206378360328,
+      "grad_norm": 0.4697562787124826,
+      "learning_rate": 7.256176079207442e-06,
+      "loss": 0.2052,
+      "step": 3736
+    },
+    {
+      "epoch": 1.2479545834029053,
+      "grad_norm": 0.43296428720065006,
+      "learning_rate": 7.254441668359527e-06,
+      "loss": 0.2019,
+      "step": 3737
+    },
+    {
+      "epoch": 1.2482885289697778,
+      "grad_norm": 0.5097000021139282,
+      "learning_rate": 7.252706916951653e-06,
+      "loss": 0.2192,
+      "step": 3738
+    },
+    {
+      "epoch": 1.2486224745366505,
+      "grad_norm": 0.4864215004236803,
+      "learning_rate": 7.250971825245874e-06,
+      "loss": 0.2168,
+      "step": 3739
+    },
+    {
+      "epoch": 1.248956420103523,
+      "grad_norm": 0.41838422249957957,
+      "learning_rate": 7.249236393504296e-06,
+      "loss": 0.2018,
+      "step": 3740
+    },
+    {
+      "epoch": 1.2492903656703958,
+      "grad_norm": 0.39261054860167127,
+      "learning_rate": 7.247500621989078e-06,
+      "loss": 0.1887,
+      "step": 3741
+    },
+    {
+      "epoch": 1.2496243112372682,
+      "grad_norm": 0.420859082136698,
+      "learning_rate": 7.245764510962426e-06,
+      "loss": 0.2072,
+      "step": 3742
+    },
+    {
+      "epoch": 1.249958256804141,
+      "grad_norm": 0.44399344347580727,
+      "learning_rate": 7.244028060686603e-06,
+      "loss": 0.2007,
+      "step": 3743
+    },
+    {
+      "epoch": 1.2502922023710135,
+      "grad_norm": 0.403279573354582,
+      "learning_rate": 7.242291271423919e-06,
+      "loss": 0.1983,
+      "step": 3744
+    },
+    {
+      "epoch": 1.2506261479378862,
+      "grad_norm": 0.40558518807409527,
+      "learning_rate": 7.240554143436735e-06,
+      "loss": 0.1942,
+      "step": 3745
+    },
+    {
+      "epoch": 1.2509600935047587,
+      "grad_norm": 0.45525307076929966,
+      "learning_rate": 7.238816676987467e-06,
+      "loss": 0.196,
+      "step": 3746
+    },
+    {
+      "epoch": 1.2512940390716314,
+      "grad_norm": 0.47302635450133046,
+      "learning_rate": 7.237078872338579e-06,
+      "loss": 0.2172,
+      "step": 3747
+    },
+    {
+      "epoch": 1.2516279846385039,
+      "grad_norm": 0.4115995099581479,
+      "learning_rate": 7.235340729752584e-06,
+      "loss": 0.202,
+      "step": 3748
+    },
+    {
+      "epoch": 1.2519619302053766,
+      "grad_norm": 0.40330391629967777,
+      "learning_rate": 7.233602249492055e-06,
+      "loss": 0.1861,
+      "step": 3749
+    },
+    {
+      "epoch": 1.252295875772249,
+      "grad_norm": 0.4611901621435869,
+      "learning_rate": 7.2318634318196045e-06,
+      "loss": 0.1993,
+      "step": 3750
+    },
+    {
+      "epoch": 1.2526298213391218,
+      "grad_norm": 0.444872798872787,
+      "learning_rate": 7.230124276997903e-06,
+      "loss": 0.1939,
+      "step": 3751
+    },
+    {
+      "epoch": 1.2529637669059943,
+      "grad_norm": 0.39062409444486607,
+      "learning_rate": 7.228384785289671e-06,
+      "loss": 0.1878,
+      "step": 3752
+    },
+    {
+      "epoch": 1.2532977124728668,
+      "grad_norm": 0.38750288835226704,
+      "learning_rate": 7.2266449569576804e-06,
+      "loss": 0.2006,
+      "step": 3753
+    },
+    {
+      "epoch": 1.2536316580397395,
+      "grad_norm": 0.4739440204799354,
+      "learning_rate": 7.224904792264748e-06,
+      "loss": 0.1972,
+      "step": 3754
+    },
+    {
+      "epoch": 1.2539656036066122,
+      "grad_norm": 0.43430096931368434,
+      "learning_rate": 7.223164291473752e-06,
+      "loss": 0.1988,
+      "step": 3755
+    },
+    {
+      "epoch": 1.2542995491734847,
+      "grad_norm": 0.4797983430080601,
+      "learning_rate": 7.221423454847611e-06,
+      "loss": 0.2029,
+      "step": 3756
+    },
+    {
+      "epoch": 1.2546334947403572,
+      "grad_norm": 0.43552537544104014,
+      "learning_rate": 7.219682282649302e-06,
+      "loss": 0.1982,
+      "step": 3757
+    },
+    {
+      "epoch": 1.25496744030723,
+      "grad_norm": 0.44330876808250125,
+      "learning_rate": 7.2179407751418485e-06,
+      "loss": 0.1996,
+      "step": 3758
+    },
+    {
+      "epoch": 1.2553013858741024,
+      "grad_norm": 0.40349763911316494,
+      "learning_rate": 7.216198932588325e-06,
+      "loss": 0.1898,
+      "step": 3759
+    },
+    {
+      "epoch": 1.2556353314409752,
+      "grad_norm": 0.4307633819537129,
+      "learning_rate": 7.214456755251858e-06,
+      "loss": 0.2011,
+      "step": 3760
+    },
+    {
+      "epoch": 1.2559692770078477,
+      "grad_norm": 0.43747344877924577,
+      "learning_rate": 7.212714243395623e-06,
+      "loss": 0.1935,
+      "step": 3761
+    },
+    {
+      "epoch": 1.2563032225747204,
+      "grad_norm": 0.439147201532576,
+      "learning_rate": 7.210971397282848e-06,
+      "loss": 0.1968,
+      "step": 3762
+    },
+    {
+      "epoch": 1.2566371681415929,
+      "grad_norm": 0.4091513057684739,
+      "learning_rate": 7.20922821717681e-06,
+      "loss": 0.1975,
+      "step": 3763
+    },
+    {
+      "epoch": 1.2569711137084656,
+      "grad_norm": 0.4001065589997931,
+      "learning_rate": 7.207484703340838e-06,
+      "loss": 0.1789,
+      "step": 3764
+    },
+    {
+      "epoch": 1.257305059275338,
+      "grad_norm": 0.41731976095296924,
+      "learning_rate": 7.205740856038308e-06,
+      "loss": 0.2007,
+      "step": 3765
+    },
+    {
+      "epoch": 1.2576390048422108,
+      "grad_norm": 0.4303382469501401,
+      "learning_rate": 7.2039966755326515e-06,
+      "loss": 0.2048,
+      "step": 3766
+    },
+    {
+      "epoch": 1.2579729504090833,
+      "grad_norm": 0.43856878267294563,
+      "learning_rate": 7.2022521620873456e-06,
+      "loss": 0.1904,
+      "step": 3767
+    },
+    {
+      "epoch": 1.2583068959759558,
+      "grad_norm": 0.4183896525795676,
+      "learning_rate": 7.2005073159659186e-06,
+      "loss": 0.1931,
+      "step": 3768
+    },
+    {
+      "epoch": 1.2586408415428285,
+      "grad_norm": 0.4647365271907042,
+      "learning_rate": 7.198762137431952e-06,
+      "loss": 0.2066,
+      "step": 3769
+    },
+    {
+      "epoch": 1.2589747871097012,
+      "grad_norm": 0.40618009076146466,
+      "learning_rate": 7.197016626749076e-06,
+      "loss": 0.185,
+      "step": 3770
+    },
+    {
+      "epoch": 1.2593087326765737,
+      "grad_norm": 0.4063402599280868,
+      "learning_rate": 7.195270784180968e-06,
+      "loss": 0.1964,
+      "step": 3771
+    },
+    {
+      "epoch": 1.2596426782434462,
+      "grad_norm": 0.4980873393441178,
+      "learning_rate": 7.193524609991359e-06,
+      "loss": 0.2167,
+      "step": 3772
+    },
+    {
+      "epoch": 1.259976623810319,
+      "grad_norm": 0.4203495972272948,
+      "learning_rate": 7.191778104444031e-06,
+      "loss": 0.1996,
+      "step": 3773
+    },
+    {
+      "epoch": 1.2603105693771914,
+      "grad_norm": 0.4691853538909953,
+      "learning_rate": 7.190031267802814e-06,
+      "loss": 0.2068,
+      "step": 3774
+    },
+    {
+      "epoch": 1.2606445149440642,
+      "grad_norm": 0.429747934431845,
+      "learning_rate": 7.188284100331585e-06,
+      "loss": 0.197,
+      "step": 3775
+    },
+    {
+      "epoch": 1.2609784605109366,
+      "grad_norm": 0.3989541459642685,
+      "learning_rate": 7.186536602294278e-06,
+      "loss": 0.1866,
+      "step": 3776
+    },
+    {
+      "epoch": 1.2613124060778094,
+      "grad_norm": 0.4079076916479964,
+      "learning_rate": 7.184788773954871e-06,
+      "loss": 0.191,
+      "step": 3777
+    },
+    {
+      "epoch": 1.2616463516446819,
+      "grad_norm": 0.37280995002049727,
+      "learning_rate": 7.1830406155773946e-06,
+      "loss": 0.1842,
+      "step": 3778
+    },
+    {
+      "epoch": 1.2619802972115546,
+      "grad_norm": 0.47418669121238,
+      "learning_rate": 7.181292127425928e-06,
+      "loss": 0.2017,
+      "step": 3779
+    },
+    {
+      "epoch": 1.262314242778427,
+      "grad_norm": 0.4557782177670249,
+      "learning_rate": 7.179543309764604e-06,
+      "loss": 0.2095,
+      "step": 3780
+    },
+    {
+      "epoch": 1.2626481883452998,
+      "grad_norm": 0.4038382372590161,
+      "learning_rate": 7.177794162857598e-06,
+      "loss": 0.1895,
+      "step": 3781
+    },
+    {
+      "epoch": 1.2629821339121723,
+      "grad_norm": 0.4257072698432259,
+      "learning_rate": 7.176044686969141e-06,
+      "loss": 0.2084,
+      "step": 3782
+    },
+    {
+      "epoch": 1.2633160794790448,
+      "grad_norm": 0.40296656493114924,
+      "learning_rate": 7.174294882363513e-06,
+      "loss": 0.2078,
+      "step": 3783
+    },
+    {
+      "epoch": 1.2636500250459175,
+      "grad_norm": 0.3972652707207378,
+      "learning_rate": 7.172544749305039e-06,
+      "loss": 0.2005,
+      "step": 3784
+    },
+    {
+      "epoch": 1.2639839706127902,
+      "grad_norm": 0.40141205047962275,
+      "learning_rate": 7.170794288058103e-06,
+      "loss": 0.1899,
+      "step": 3785
+    },
+    {
+      "epoch": 1.2643179161796627,
+      "grad_norm": 0.3924619438453352,
+      "learning_rate": 7.169043498887126e-06,
+      "loss": 0.1928,
+      "step": 3786
+    },
+    {
+      "epoch": 1.2646518617465352,
+      "grad_norm": 0.4310553864330334,
+      "learning_rate": 7.1672923820565925e-06,
+      "loss": 0.1945,
+      "step": 3787
+    },
+    {
+      "epoch": 1.264985807313408,
+      "grad_norm": 0.42295352764743993,
+      "learning_rate": 7.165540937831024e-06,
+      "loss": 0.1984,
+      "step": 3788
+    },
+    {
+      "epoch": 1.2653197528802806,
+      "grad_norm": 0.38478960335529383,
+      "learning_rate": 7.163789166474998e-06,
+      "loss": 0.1907,
+      "step": 3789
+    },
+    {
+      "epoch": 1.2656536984471531,
+      "grad_norm": 0.3964245436953759,
+      "learning_rate": 7.162037068253141e-06,
+      "loss": 0.1824,
+      "step": 3790
+    },
+    {
+      "epoch": 1.2659876440140256,
+      "grad_norm": 0.4650615379634907,
+      "learning_rate": 7.160284643430129e-06,
+      "loss": 0.2234,
+      "step": 3791
+    },
+    {
+      "epoch": 1.2663215895808984,
+      "grad_norm": 0.4988659718160951,
+      "learning_rate": 7.158531892270682e-06,
+      "loss": 0.1991,
+      "step": 3792
+    },
+    {
+      "epoch": 1.2666555351477709,
+      "grad_norm": 0.40544641816954785,
+      "learning_rate": 7.156778815039579e-06,
+      "loss": 0.1871,
+      "step": 3793
+    },
+    {
+      "epoch": 1.2669894807146436,
+      "grad_norm": 0.4331677677975272,
+      "learning_rate": 7.15502541200164e-06,
+      "loss": 0.2045,
+      "step": 3794
+    },
+    {
+      "epoch": 1.267323426281516,
+      "grad_norm": 0.4427416226804204,
+      "learning_rate": 7.153271683421738e-06,
+      "loss": 0.2012,
+      "step": 3795
+    },
+    {
+      "epoch": 1.2676573718483888,
+      "grad_norm": 0.43060810240200514,
+      "learning_rate": 7.151517629564795e-06,
+      "loss": 0.2067,
+      "step": 3796
+    },
+    {
+      "epoch": 1.2679913174152613,
+      "grad_norm": 0.4162223937732087,
+      "learning_rate": 7.14976325069578e-06,
+      "loss": 0.2001,
+      "step": 3797
+    },
+    {
+      "epoch": 1.268325262982134,
+      "grad_norm": 0.4283496532813782,
+      "learning_rate": 7.148008547079713e-06,
+      "loss": 0.1942,
+      "step": 3798
+    },
+    {
+      "epoch": 1.2686592085490065,
+      "grad_norm": 0.3919278069619941,
+      "learning_rate": 7.1462535189816636e-06,
+      "loss": 0.1978,
+      "step": 3799
+    },
+    {
+      "epoch": 1.2689931541158792,
+      "grad_norm": 0.4030159316869139,
+      "learning_rate": 7.14449816666675e-06,
+      "loss": 0.2024,
+      "step": 3800
+    },
+    {
+      "epoch": 1.2693270996827517,
+      "grad_norm": 0.4122621176678283,
+      "learning_rate": 7.142742490400135e-06,
+      "loss": 0.1939,
+      "step": 3801
+    },
+    {
+      "epoch": 1.2696610452496242,
+      "grad_norm": 0.4242214438614236,
+      "learning_rate": 7.140986490447039e-06,
+      "loss": 0.2014,
+      "step": 3802
+    },
+    {
+      "epoch": 1.269994990816497,
+      "grad_norm": 0.41850239250671534,
+      "learning_rate": 7.139230167072724e-06,
+      "loss": 0.2037,
+      "step": 3803
+    },
+    {
+      "epoch": 1.2703289363833696,
+      "grad_norm": 0.4148927143434032,
+      "learning_rate": 7.137473520542503e-06,
+      "loss": 0.1982,
+      "step": 3804
+    },
+    {
+      "epoch": 1.2706628819502421,
+      "grad_norm": 0.42901971140755585,
+      "learning_rate": 7.135716551121739e-06,
+      "loss": 0.1966,
+      "step": 3805
+    },
+    {
+      "epoch": 1.2709968275171146,
+      "grad_norm": 0.45317109667412,
+      "learning_rate": 7.133959259075844e-06,
+      "loss": 0.2161,
+      "step": 3806
+    },
+    {
+      "epoch": 1.2713307730839873,
+      "grad_norm": 0.38629809789963193,
+      "learning_rate": 7.132201644670274e-06,
+      "loss": 0.1946,
+      "step": 3807
+    },
+    {
+      "epoch": 1.2716647186508598,
+      "grad_norm": 0.43946380323346784,
+      "learning_rate": 7.13044370817054e-06,
+      "loss": 0.2185,
+      "step": 3808
+    },
+    {
+      "epoch": 1.2719986642177326,
+      "grad_norm": 0.4153870460305795,
+      "learning_rate": 7.128685449842201e-06,
+      "loss": 0.1829,
+      "step": 3809
+    },
+    {
+      "epoch": 1.272332609784605,
+      "grad_norm": 0.4070835430938735,
+      "learning_rate": 7.1269268699508574e-06,
+      "loss": 0.1914,
+      "step": 3810
+    },
+    {
+      "epoch": 1.2726665553514778,
+      "grad_norm": 0.5294790454069535,
+      "learning_rate": 7.1251679687621685e-06,
+      "loss": 0.197,
+      "step": 3811
+    },
+    {
+      "epoch": 1.2730005009183503,
+      "grad_norm": 0.43195916234280074,
+      "learning_rate": 7.123408746541835e-06,
+      "loss": 0.2067,
+      "step": 3812
+    },
+    {
+      "epoch": 1.273334446485223,
+      "grad_norm": 0.4223072878291587,
+      "learning_rate": 7.1216492035556075e-06,
+      "loss": 0.2024,
+      "step": 3813
+    },
+    {
+      "epoch": 1.2736683920520955,
+      "grad_norm": 0.43487357461786935,
+      "learning_rate": 7.119889340069286e-06,
+      "loss": 0.1966,
+      "step": 3814
+    },
+    {
+      "epoch": 1.2740023376189682,
+      "grad_norm": 0.4051402525802372,
+      "learning_rate": 7.1181291563487175e-06,
+      "loss": 0.1887,
+      "step": 3815
+    },
+    {
+      "epoch": 1.2743362831858407,
+      "grad_norm": 0.4456616551512561,
+      "learning_rate": 7.116368652659802e-06,
+      "loss": 0.2077,
+      "step": 3816
+    },
+    {
+      "epoch": 1.2746702287527132,
+      "grad_norm": 0.407498695448717,
+      "learning_rate": 7.114607829268481e-06,
+      "loss": 0.1967,
+      "step": 3817
+    },
+    {
+      "epoch": 1.275004174319586,
+      "grad_norm": 0.40906980972589485,
+      "learning_rate": 7.1128466864407486e-06,
+      "loss": 0.1986,
+      "step": 3818
+    },
+    {
+      "epoch": 1.2753381198864586,
+      "grad_norm": 0.38020921049437223,
+      "learning_rate": 7.111085224442647e-06,
+      "loss": 0.1864,
+      "step": 3819
+    },
+    {
+      "epoch": 1.2756720654533311,
+      "grad_norm": 0.41376376861187264,
+      "learning_rate": 7.109323443540263e-06,
+      "loss": 0.2022,
+      "step": 3820
+    },
+    {
+      "epoch": 1.2760060110202036,
+      "grad_norm": 0.4085034643758407,
+      "learning_rate": 7.107561343999739e-06,
+      "loss": 0.2002,
+      "step": 3821
+    },
+    {
+      "epoch": 1.2763399565870763,
+      "grad_norm": 0.4147881532976798,
+      "learning_rate": 7.105798926087257e-06,
+      "loss": 0.1929,
+      "step": 3822
+    },
+    {
+      "epoch": 1.2766739021539488,
+      "grad_norm": 0.4164403555278612,
+      "learning_rate": 7.104036190069052e-06,
+      "loss": 0.1979,
+      "step": 3823
+    },
+    {
+      "epoch": 1.2770078477208215,
+      "grad_norm": 0.388973867320647,
+      "learning_rate": 7.102273136211407e-06,
+      "loss": 0.1947,
+      "step": 3824
+    },
+    {
+      "epoch": 1.277341793287694,
+      "grad_norm": 0.43615613297218037,
+      "learning_rate": 7.10050976478065e-06,
+      "loss": 0.1904,
+      "step": 3825
+    },
+    {
+      "epoch": 1.2776757388545668,
+      "grad_norm": 0.4217476905256785,
+      "learning_rate": 7.098746076043162e-06,
+      "loss": 0.1976,
+      "step": 3826
+    },
+    {
+      "epoch": 1.2780096844214393,
+      "grad_norm": 0.4410477775249162,
+      "learning_rate": 7.096982070265366e-06,
+      "loss": 0.2111,
+      "step": 3827
+    },
+    {
+      "epoch": 1.278343629988312,
+      "grad_norm": 0.407553596713063,
+      "learning_rate": 7.0952177477137374e-06,
+      "loss": 0.1877,
+      "step": 3828
+    },
+    {
+      "epoch": 1.2786775755551845,
+      "grad_norm": 0.4226143283409012,
+      "learning_rate": 7.093453108654798e-06,
+      "loss": 0.1942,
+      "step": 3829
+    },
+    {
+      "epoch": 1.2790115211220572,
+      "grad_norm": 0.44448160702216344,
+      "learning_rate": 7.091688153355116e-06,
+      "loss": 0.2117,
+      "step": 3830
+    },
+    {
+      "epoch": 1.2793454666889297,
+      "grad_norm": 0.44023312970202644,
+      "learning_rate": 7.08992288208131e-06,
+      "loss": 0.2061,
+      "step": 3831
+    },
+    {
+      "epoch": 1.2796794122558022,
+      "grad_norm": 0.4289817202503459,
+      "learning_rate": 7.088157295100046e-06,
+      "loss": 0.1982,
+      "step": 3832
+    },
+    {
+      "epoch": 1.280013357822675,
+      "grad_norm": 0.41168121667242447,
+      "learning_rate": 7.0863913926780335e-06,
+      "loss": 0.1965,
+      "step": 3833
+    },
+    {
+      "epoch": 1.2803473033895476,
+      "grad_norm": 0.389066292024194,
+      "learning_rate": 7.084625175082036e-06,
+      "loss": 0.1969,
+      "step": 3834
+    },
+    {
+      "epoch": 1.28068124895642,
+      "grad_norm": 0.42211956082879903,
+      "learning_rate": 7.082858642578861e-06,
+      "loss": 0.1927,
+      "step": 3835
+    },
+    {
+      "epoch": 1.2810151945232926,
+      "grad_norm": 0.41487931684211865,
+      "learning_rate": 7.081091795435361e-06,
+      "loss": 0.2018,
+      "step": 3836
+    },
+    {
+      "epoch": 1.2813491400901653,
+      "grad_norm": 0.4355951362057153,
+      "learning_rate": 7.079324633918443e-06,
+      "loss": 0.187,
+      "step": 3837
+    },
+    {
+      "epoch": 1.281683085657038,
+      "grad_norm": 0.40177334082527777,
+      "learning_rate": 7.077557158295053e-06,
+      "loss": 0.1943,
+      "step": 3838
+    },
+    {
+      "epoch": 1.2820170312239105,
+      "grad_norm": 0.4688295048980153,
+      "learning_rate": 7.075789368832194e-06,
+      "loss": 0.2056,
+      "step": 3839
+    },
+    {
+      "epoch": 1.282350976790783,
+      "grad_norm": 0.45332907753482743,
+      "learning_rate": 7.074021265796909e-06,
+      "loss": 0.2151,
+      "step": 3840
+    },
+    {
+      "epoch": 1.2826849223576557,
+      "grad_norm": 0.4011388285723007,
+      "learning_rate": 7.072252849456291e-06,
+      "loss": 0.1948,
+      "step": 3841
+    },
+    {
+      "epoch": 1.2830188679245282,
+      "grad_norm": 0.4397419364652935,
+      "learning_rate": 7.07048412007748e-06,
+      "loss": 0.1973,
+      "step": 3842
+    },
+    {
+      "epoch": 1.283352813491401,
+      "grad_norm": 0.4191526878534424,
+      "learning_rate": 7.068715077927664e-06,
+      "loss": 0.1922,
+      "step": 3843
+    },
+    {
+      "epoch": 1.2836867590582735,
+      "grad_norm": 0.4622949980235757,
+      "learning_rate": 7.066945723274077e-06,
+      "loss": 0.1966,
+      "step": 3844
+    },
+    {
+      "epoch": 1.2840207046251462,
+      "grad_norm": 0.39544208070452724,
+      "learning_rate": 7.065176056383999e-06,
+      "loss": 0.1978,
+      "step": 3845
+    },
+    {
+      "epoch": 1.2843546501920187,
+      "grad_norm": 0.4669748844624701,
+      "learning_rate": 7.063406077524764e-06,
+      "loss": 0.2106,
+      "step": 3846
+    },
+    {
+      "epoch": 1.2846885957588914,
+      "grad_norm": 0.4045890789425728,
+      "learning_rate": 7.061635786963743e-06,
+      "loss": 0.1927,
+      "step": 3847
+    },
+    {
+      "epoch": 1.2850225413257639,
+      "grad_norm": 0.4297051183096818,
+      "learning_rate": 7.059865184968362e-06,
+      "loss": 0.1991,
+      "step": 3848
+    },
+    {
+      "epoch": 1.2853564868926366,
+      "grad_norm": 0.44503532628630693,
+      "learning_rate": 7.058094271806091e-06,
+      "loss": 0.2127,
+      "step": 3849
+    },
+    {
+      "epoch": 1.285690432459509,
+      "grad_norm": 0.38682645112211256,
+      "learning_rate": 7.056323047744447e-06,
+      "loss": 0.1753,
+      "step": 3850
+    },
+    {
+      "epoch": 1.2860243780263816,
+      "grad_norm": 0.44630819787955817,
+      "learning_rate": 7.054551513050993e-06,
+      "loss": 0.2156,
+      "step": 3851
+    },
+    {
+      "epoch": 1.2863583235932543,
+      "grad_norm": 0.4378283545797013,
+      "learning_rate": 7.052779667993342e-06,
+      "loss": 0.2057,
+      "step": 3852
+    },
+    {
+      "epoch": 1.286692269160127,
+      "grad_norm": 0.4304232637199404,
+      "learning_rate": 7.051007512839153e-06,
+      "loss": 0.2205,
+      "step": 3853
+    },
+    {
+      "epoch": 1.2870262147269995,
+      "grad_norm": 0.5159648005944082,
+      "learning_rate": 7.0492350478561275e-06,
+      "loss": 0.1995,
+      "step": 3854
+    },
+    {
+      "epoch": 1.287360160293872,
+      "grad_norm": 0.457847641860177,
+      "learning_rate": 7.04746227331202e-06,
+      "loss": 0.2067,
+      "step": 3855
+    },
+    {
+      "epoch": 1.2876941058607447,
+      "grad_norm": 0.41872580138211385,
+      "learning_rate": 7.045689189474628e-06,
+      "loss": 0.1934,
+      "step": 3856
+    },
+    {
+      "epoch": 1.2880280514276172,
+      "grad_norm": 0.4110936768730782,
+      "learning_rate": 7.0439157966117955e-06,
+      "loss": 0.1858,
+      "step": 3857
+    },
+    {
+      "epoch": 1.28836199699449,
+      "grad_norm": 0.38185640035698354,
+      "learning_rate": 7.042142094991418e-06,
+      "loss": 0.1816,
+      "step": 3858
+    },
+    {
+      "epoch": 1.2886959425613624,
+      "grad_norm": 0.3873522432517064,
+      "learning_rate": 7.04036808488143e-06,
+      "loss": 0.1964,
+      "step": 3859
+    },
+    {
+      "epoch": 1.2890298881282352,
+      "grad_norm": 0.4232734771641415,
+      "learning_rate": 7.038593766549817e-06,
+      "loss": 0.2104,
+      "step": 3860
+    },
+    {
+      "epoch": 1.2893638336951077,
+      "grad_norm": 0.4746183445203052,
+      "learning_rate": 7.0368191402646145e-06,
+      "loss": 0.2097,
+      "step": 3861
+    },
+    {
+      "epoch": 1.2896977792619804,
+      "grad_norm": 0.42621702835077613,
+      "learning_rate": 7.035044206293898e-06,
+      "loss": 0.2194,
+      "step": 3862
+    },
+    {
+      "epoch": 1.2900317248288529,
+      "grad_norm": 0.42742800253037144,
+      "learning_rate": 7.0332689649057905e-06,
+      "loss": 0.1985,
+      "step": 3863
+    },
+    {
+      "epoch": 1.2903656703957256,
+      "grad_norm": 0.4320688142367941,
+      "learning_rate": 7.031493416368466e-06,
+      "loss": 0.1984,
+      "step": 3864
+    },
+    {
+      "epoch": 1.290699615962598,
+      "grad_norm": 0.4611896460687471,
+      "learning_rate": 7.029717560950141e-06,
+      "loss": 0.2105,
+      "step": 3865
+    },
+    {
+      "epoch": 1.2910335615294706,
+      "grad_norm": 0.42167209030771285,
+      "learning_rate": 7.027941398919078e-06,
+      "loss": 0.2007,
+      "step": 3866
+    },
+    {
+      "epoch": 1.2913675070963433,
+      "grad_norm": 0.4486720053192747,
+      "learning_rate": 7.0261649305435895e-06,
+      "loss": 0.2107,
+      "step": 3867
+    },
+    {
+      "epoch": 1.291701452663216,
+      "grad_norm": 0.43703159019530075,
+      "learning_rate": 7.02438815609203e-06,
+      "loss": 0.2002,
+      "step": 3868
+    },
+    {
+      "epoch": 1.2920353982300885,
+      "grad_norm": 0.4206235426460402,
+      "learning_rate": 7.022611075832804e-06,
+      "loss": 0.1824,
+      "step": 3869
+    },
+    {
+      "epoch": 1.292369343796961,
+      "grad_norm": 0.45899661384875784,
+      "learning_rate": 7.02083369003436e-06,
+      "loss": 0.2091,
+      "step": 3870
+    },
+    {
+      "epoch": 1.2927032893638337,
+      "grad_norm": 0.4163150817565343,
+      "learning_rate": 7.019055998965191e-06,
+      "loss": 0.2037,
+      "step": 3871
+    },
+    {
+      "epoch": 1.2930372349307062,
+      "grad_norm": 0.48635183034398205,
+      "learning_rate": 7.017278002893841e-06,
+      "loss": 0.2072,
+      "step": 3872
+    },
+    {
+      "epoch": 1.293371180497579,
+      "grad_norm": 0.4210590900618153,
+      "learning_rate": 7.015499702088896e-06,
+      "loss": 0.1871,
+      "step": 3873
+    },
+    {
+      "epoch": 1.2937051260644514,
+      "grad_norm": 0.42679469662211317,
+      "learning_rate": 7.013721096818988e-06,
+      "loss": 0.2057,
+      "step": 3874
+    },
+    {
+      "epoch": 1.2940390716313241,
+      "grad_norm": 0.43109509020232506,
+      "learning_rate": 7.011942187352798e-06,
+      "loss": 0.1961,
+      "step": 3875
+    },
+    {
+      "epoch": 1.2943730171981966,
+      "grad_norm": 0.4261404803605642,
+      "learning_rate": 7.010162973959052e-06,
+      "loss": 0.1987,
+      "step": 3876
+    },
+    {
+      "epoch": 1.2947069627650694,
+      "grad_norm": 0.3984455245433112,
+      "learning_rate": 7.008383456906518e-06,
+      "loss": 0.1915,
+      "step": 3877
+    },
+    {
+      "epoch": 1.2950409083319419,
+      "grad_norm": 0.44213008748103527,
+      "learning_rate": 7.0066036364640165e-06,
+      "loss": 0.2063,
+      "step": 3878
+    },
+    {
+      "epoch": 1.2953748538988146,
+      "grad_norm": 0.4340151247449952,
+      "learning_rate": 7.004823512900408e-06,
+      "loss": 0.2061,
+      "step": 3879
+    },
+    {
+      "epoch": 1.295708799465687,
+      "grad_norm": 0.4353715360420167,
+      "learning_rate": 7.003043086484602e-06,
+      "loss": 0.2192,
+      "step": 3880
+    },
+    {
+      "epoch": 1.2960427450325596,
+      "grad_norm": 0.40389370415658,
+      "learning_rate": 7.001262357485553e-06,
+      "loss": 0.1915,
+      "step": 3881
+    },
+    {
+      "epoch": 1.2963766905994323,
+      "grad_norm": 0.4385006533052746,
+      "learning_rate": 6.99948132617226e-06,
+      "loss": 0.211,
+      "step": 3882
+    },
+    {
+      "epoch": 1.296710636166305,
+      "grad_norm": 0.42962689985450564,
+      "learning_rate": 6.99769999281377e-06,
+      "loss": 0.2004,
+      "step": 3883
+    },
+    {
+      "epoch": 1.2970445817331775,
+      "grad_norm": 0.3988243836403783,
+      "learning_rate": 6.9959183576791745e-06,
+      "loss": 0.1935,
+      "step": 3884
+    },
+    {
+      "epoch": 1.29737852730005,
+      "grad_norm": 0.46588350704208614,
+      "learning_rate": 6.9941364210376095e-06,
+      "loss": 0.2027,
+      "step": 3885
+    },
+    {
+      "epoch": 1.2977124728669227,
+      "grad_norm": 0.4423870351069402,
+      "learning_rate": 6.992354183158258e-06,
+      "loss": 0.2006,
+      "step": 3886
+    },
+    {
+      "epoch": 1.2980464184337954,
+      "grad_norm": 0.44696770311545875,
+      "learning_rate": 6.9905716443103475e-06,
+      "loss": 0.1977,
+      "step": 3887
+    },
+    {
+      "epoch": 1.298380364000668,
+      "grad_norm": 0.43550098856674935,
+      "learning_rate": 6.9887888047631525e-06,
+      "loss": 0.1935,
+      "step": 3888
+    },
+    {
+      "epoch": 1.2987143095675404,
+      "grad_norm": 0.45209412899280554,
+      "learning_rate": 6.987005664785991e-06,
+      "loss": 0.1968,
+      "step": 3889
+    },
+    {
+      "epoch": 1.2990482551344131,
+      "grad_norm": 0.44580123030250995,
+      "learning_rate": 6.985222224648227e-06,
+      "loss": 0.2001,
+      "step": 3890
+    },
+    {
+      "epoch": 1.2993822007012856,
+      "grad_norm": 0.45139711177023095,
+      "learning_rate": 6.983438484619272e-06,
+      "loss": 0.2037,
+      "step": 3891
+    },
+    {
+      "epoch": 1.2997161462681583,
+      "grad_norm": 0.4219811830085104,
+      "learning_rate": 6.981654444968578e-06,
+      "loss": 0.1982,
+      "step": 3892
+    },
+    {
+      "epoch": 1.3000500918350308,
+      "grad_norm": 0.44396824829876985,
+      "learning_rate": 6.979870105965648e-06,
+      "loss": 0.2021,
+      "step": 3893
+    },
+    {
+      "epoch": 1.3003840374019036,
+      "grad_norm": 0.41445057653912953,
+      "learning_rate": 6.978085467880027e-06,
+      "loss": 0.1894,
+      "step": 3894
+    },
+    {
+      "epoch": 1.300717982968776,
+      "grad_norm": 0.4947034290772902,
+      "learning_rate": 6.9763005309813025e-06,
+      "loss": 0.206,
+      "step": 3895
+    },
+    {
+      "epoch": 1.3010519285356488,
+      "grad_norm": 0.43818926332258024,
+      "learning_rate": 6.974515295539115e-06,
+      "loss": 0.2049,
+      "step": 3896
+    },
+    {
+      "epoch": 1.3013858741025213,
+      "grad_norm": 0.4498029142452314,
+      "learning_rate": 6.9727297618231416e-06,
+      "loss": 0.2033,
+      "step": 3897
+    },
+    {
+      "epoch": 1.301719819669394,
+      "grad_norm": 0.39991165215175273,
+      "learning_rate": 6.970943930103109e-06,
+      "loss": 0.2049,
+      "step": 3898
+    },
+    {
+      "epoch": 1.3020537652362665,
+      "grad_norm": 0.39753769966116426,
+      "learning_rate": 6.96915780064879e-06,
+      "loss": 0.1969,
+      "step": 3899
+    },
+    {
+      "epoch": 1.302387710803139,
+      "grad_norm": 0.4219028223395952,
+      "learning_rate": 6.96737137373e-06,
+      "loss": 0.2007,
+      "step": 3900
+    },
+    {
+      "epoch": 1.3027216563700117,
+      "grad_norm": 0.4405048022853998,
+      "learning_rate": 6.965584649616597e-06,
+      "loss": 0.2084,
+      "step": 3901
+    },
+    {
+      "epoch": 1.3030556019368844,
+      "grad_norm": 0.47780153655107577,
+      "learning_rate": 6.963797628578489e-06,
+      "loss": 0.189,
+      "step": 3902
+    },
+    {
+      "epoch": 1.303389547503757,
+      "grad_norm": 0.44621414800365433,
+      "learning_rate": 6.962010310885627e-06,
+      "loss": 0.201,
+      "step": 3903
+    },
+    {
+      "epoch": 1.3037234930706294,
+      "grad_norm": 0.4567887700056304,
+      "learning_rate": 6.960222696808004e-06,
+      "loss": 0.2142,
+      "step": 3904
+    },
+    {
+      "epoch": 1.3040574386375021,
+      "grad_norm": 0.4348089611652648,
+      "learning_rate": 6.958434786615663e-06,
+      "loss": 0.1969,
+      "step": 3905
+    },
+    {
+      "epoch": 1.3043913842043746,
+      "grad_norm": 0.4141676962452518,
+      "learning_rate": 6.956646580578687e-06,
+      "loss": 0.2024,
+      "step": 3906
+    },
+    {
+      "epoch": 1.3047253297712473,
+      "grad_norm": 0.5021452032817756,
+      "learning_rate": 6.954858078967207e-06,
+      "loss": 0.2167,
+      "step": 3907
+    },
+    {
+      "epoch": 1.3050592753381198,
+      "grad_norm": 0.42162857639320483,
+      "learning_rate": 6.953069282051397e-06,
+      "loss": 0.198,
+      "step": 3908
+    },
+    {
+      "epoch": 1.3053932209049925,
+      "grad_norm": 0.4012274684722401,
+      "learning_rate": 6.951280190101475e-06,
+      "loss": 0.1932,
+      "step": 3909
+    },
+    {
+      "epoch": 1.305727166471865,
+      "grad_norm": 0.4800296772397605,
+      "learning_rate": 6.949490803387704e-06,
+      "loss": 0.2146,
+      "step": 3910
+    },
+    {
+      "epoch": 1.3060611120387378,
+      "grad_norm": 0.430988613912435,
+      "learning_rate": 6.9477011221803935e-06,
+      "loss": 0.1926,
+      "step": 3911
+    },
+    {
+      "epoch": 1.3063950576056103,
+      "grad_norm": 0.5275440853144523,
+      "learning_rate": 6.945911146749894e-06,
+      "loss": 0.1949,
+      "step": 3912
+    },
+    {
+      "epoch": 1.306729003172483,
+      "grad_norm": 0.4382678861181215,
+      "learning_rate": 6.944120877366605e-06,
+      "loss": 0.2023,
+      "step": 3913
+    },
+    {
+      "epoch": 1.3070629487393555,
+      "grad_norm": 0.44380835569655847,
+      "learning_rate": 6.9423303143009644e-06,
+      "loss": 0.2155,
+      "step": 3914
+    },
+    {
+      "epoch": 1.307396894306228,
+      "grad_norm": 0.4166446811542404,
+      "learning_rate": 6.940539457823459e-06,
+      "loss": 0.2038,
+      "step": 3915
+    },
+    {
+      "epoch": 1.3077308398731007,
+      "grad_norm": 0.4268607935295313,
+      "learning_rate": 6.938748308204622e-06,
+      "loss": 0.2065,
+      "step": 3916
+    },
+    {
+      "epoch": 1.3080647854399734,
+      "grad_norm": 0.44848540251287616,
+      "learning_rate": 6.936956865715024e-06,
+      "loss": 0.1959,
+      "step": 3917
+    },
+    {
+      "epoch": 1.308398731006846,
+      "grad_norm": 0.4524449751352037,
+      "learning_rate": 6.9351651306252836e-06,
+      "loss": 0.2078,
+      "step": 3918
+    },
+    {
+      "epoch": 1.3087326765737184,
+      "grad_norm": 0.4111904238885648,
+      "learning_rate": 6.933373103206064e-06,
+      "loss": 0.1935,
+      "step": 3919
+    },
+    {
+      "epoch": 1.309066622140591,
+      "grad_norm": 0.49171063968711426,
+      "learning_rate": 6.931580783728075e-06,
+      "loss": 0.2076,
+      "step": 3920
+    },
+    {
+      "epoch": 1.3094005677074636,
+      "grad_norm": 0.42276347946771226,
+      "learning_rate": 6.929788172462063e-06,
+      "loss": 0.2004,
+      "step": 3921
+    },
+    {
+      "epoch": 1.3097345132743363,
+      "grad_norm": 0.45493648816395343,
+      "learning_rate": 6.927995269678826e-06,
+      "loss": 0.2091,
+      "step": 3922
+    },
+    {
+      "epoch": 1.3100684588412088,
+      "grad_norm": 0.4905831176410732,
+      "learning_rate": 6.926202075649202e-06,
+      "loss": 0.2073,
+      "step": 3923
+    },
+    {
+      "epoch": 1.3104024044080815,
+      "grad_norm": 0.4280595066960188,
+      "learning_rate": 6.924408590644073e-06,
+      "loss": 0.2102,
+      "step": 3924
+    },
+    {
+      "epoch": 1.310736349974954,
+      "grad_norm": 0.4519193608106046,
+      "learning_rate": 6.922614814934367e-06,
+      "loss": 0.2056,
+      "step": 3925
+    },
+    {
+      "epoch": 1.3110702955418267,
+      "grad_norm": 0.4404172492976198,
+      "learning_rate": 6.920820748791057e-06,
+      "loss": 0.1964,
+      "step": 3926
+    },
+    {
+      "epoch": 1.3114042411086992,
+      "grad_norm": 0.42868822460623446,
+      "learning_rate": 6.919026392485154e-06,
+      "loss": 0.1973,
+      "step": 3927
+    },
+    {
+      "epoch": 1.311738186675572,
+      "grad_norm": 0.43305908963203626,
+      "learning_rate": 6.91723174628772e-06,
+      "loss": 0.1928,
+      "step": 3928
+    },
+    {
+      "epoch": 1.3120721322424445,
+      "grad_norm": 0.4477015665448263,
+      "learning_rate": 6.915436810469856e-06,
+      "loss": 0.2061,
+      "step": 3929
+    },
+    {
+      "epoch": 1.312406077809317,
+      "grad_norm": 0.4061799452628013,
+      "learning_rate": 6.913641585302708e-06,
+      "loss": 0.1882,
+      "step": 3930
+    },
+    {
+      "epoch": 1.3127400233761897,
+      "grad_norm": 0.3957076000775846,
+      "learning_rate": 6.9118460710574665e-06,
+      "loss": 0.197,
+      "step": 3931
+    },
+    {
+      "epoch": 1.3130739689430624,
+      "grad_norm": 0.4529677675383033,
+      "learning_rate": 6.910050268005364e-06,
+      "loss": 0.1999,
+      "step": 3932
+    },
+    {
+      "epoch": 1.3134079145099349,
+      "grad_norm": 0.45692150570119494,
+      "learning_rate": 6.908254176417679e-06,
+      "loss": 0.1991,
+      "step": 3933
+    },
+    {
+      "epoch": 1.3137418600768074,
+      "grad_norm": 0.3908259281335957,
+      "learning_rate": 6.906457796565732e-06,
+      "loss": 0.1917,
+      "step": 3934
+    },
+    {
+      "epoch": 1.31407580564368,
+      "grad_norm": 0.44642695936390425,
+      "learning_rate": 6.904661128720887e-06,
+      "loss": 0.2102,
+      "step": 3935
+    },
+    {
+      "epoch": 1.3144097512105528,
+      "grad_norm": 0.4140422793919292,
+      "learning_rate": 6.902864173154551e-06,
+      "loss": 0.1954,
+      "step": 3936
+    },
+    {
+      "epoch": 1.3147436967774253,
+      "grad_norm": 0.4175994886561666,
+      "learning_rate": 6.9010669301381765e-06,
+      "loss": 0.197,
+      "step": 3937
+    },
+    {
+      "epoch": 1.3150776423442978,
+      "grad_norm": 0.43189349728431153,
+      "learning_rate": 6.899269399943258e-06,
+      "loss": 0.2212,
+      "step": 3938
+    },
+    {
+      "epoch": 1.3154115879111705,
+      "grad_norm": 0.4035715995235755,
+      "learning_rate": 6.897471582841333e-06,
+      "loss": 0.1968,
+      "step": 3939
+    },
+    {
+      "epoch": 1.315745533478043,
+      "grad_norm": 0.4664423426609285,
+      "learning_rate": 6.895673479103983e-06,
+      "loss": 0.2087,
+      "step": 3940
+    },
+    {
+      "epoch": 1.3160794790449157,
+      "grad_norm": 0.4664300811353399,
+      "learning_rate": 6.893875089002835e-06,
+      "loss": 0.2107,
+      "step": 3941
+    },
+    {
+      "epoch": 1.3164134246117882,
+      "grad_norm": 0.3890368808127858,
+      "learning_rate": 6.892076412809553e-06,
+      "loss": 0.1894,
+      "step": 3942
+    },
+    {
+      "epoch": 1.316747370178661,
+      "grad_norm": 0.4327141924488333,
+      "learning_rate": 6.890277450795851e-06,
+      "loss": 0.1968,
+      "step": 3943
+    },
+    {
+      "epoch": 1.3170813157455334,
+      "grad_norm": 0.42540004436257756,
+      "learning_rate": 6.888478203233484e-06,
+      "loss": 0.1837,
+      "step": 3944
+    },
+    {
+      "epoch": 1.3174152613124062,
+      "grad_norm": 0.40663174769799976,
+      "learning_rate": 6.886678670394247e-06,
+      "loss": 0.1884,
+      "step": 3945
+    },
+    {
+      "epoch": 1.3177492068792787,
+      "grad_norm": 0.4240067688960512,
+      "learning_rate": 6.884878852549982e-06,
+      "loss": 0.2086,
+      "step": 3946
+    },
+    {
+      "epoch": 1.3180831524461514,
+      "grad_norm": 0.46538495365988414,
+      "learning_rate": 6.883078749972573e-06,
+      "loss": 0.1947,
+      "step": 3947
+    },
+    {
+      "epoch": 1.3184170980130239,
+      "grad_norm": 0.39694236009597017,
+      "learning_rate": 6.881278362933947e-06,
+      "loss": 0.1926,
+      "step": 3948
+    },
+    {
+      "epoch": 1.3187510435798964,
+      "grad_norm": 0.4084027414378485,
+      "learning_rate": 6.879477691706071e-06,
+      "loss": 0.1912,
+      "step": 3949
+    },
+    {
+      "epoch": 1.319084989146769,
+      "grad_norm": 0.4483001951410644,
+      "learning_rate": 6.877676736560961e-06,
+      "loss": 0.2115,
+      "step": 3950
+    },
+    {
+      "epoch": 1.3194189347136418,
+      "grad_norm": 0.45057352996667105,
+      "learning_rate": 6.87587549777067e-06,
+      "loss": 0.2042,
+      "step": 3951
+    },
+    {
+      "epoch": 1.3197528802805143,
+      "grad_norm": 0.42211307538365817,
+      "learning_rate": 6.874073975607298e-06,
+      "loss": 0.2053,
+      "step": 3952
+    },
+    {
+      "epoch": 1.3200868258473868,
+      "grad_norm": 0.41256970624902645,
+      "learning_rate": 6.872272170342985e-06,
+      "loss": 0.2016,
+      "step": 3953
+    },
+    {
+      "epoch": 1.3204207714142595,
+      "grad_norm": 0.5930226936995437,
+      "learning_rate": 6.870470082249917e-06,
+      "loss": 0.2153,
+      "step": 3954
+    },
+    {
+      "epoch": 1.320754716981132,
+      "grad_norm": 0.4080686250723333,
+      "learning_rate": 6.868667711600318e-06,
+      "loss": 0.1957,
+      "step": 3955
+    },
+    {
+      "epoch": 1.3210886625480047,
+      "grad_norm": 0.4330981620220748,
+      "learning_rate": 6.866865058666459e-06,
+      "loss": 0.2001,
+      "step": 3956
+    },
+    {
+      "epoch": 1.3214226081148772,
+      "grad_norm": 0.41330007823441606,
+      "learning_rate": 6.86506212372065e-06,
+      "loss": 0.2064,
+      "step": 3957
+    },
+    {
+      "epoch": 1.32175655368175,
+      "grad_norm": 0.4314166600814895,
+      "learning_rate": 6.863258907035246e-06,
+      "loss": 0.2024,
+      "step": 3958
+    },
+    {
+      "epoch": 1.3220904992486224,
+      "grad_norm": 0.45440336445114765,
+      "learning_rate": 6.861455408882647e-06,
+      "loss": 0.2149,
+      "step": 3959
+    },
+    {
+      "epoch": 1.3224244448154951,
+      "grad_norm": 0.4156203304442072,
+      "learning_rate": 6.85965162953529e-06,
+      "loss": 0.1992,
+      "step": 3960
+    },
+    {
+      "epoch": 1.3227583903823676,
+      "grad_norm": 0.40586715685586255,
+      "learning_rate": 6.857847569265657e-06,
+      "loss": 0.1912,
+      "step": 3961
+    },
+    {
+      "epoch": 1.3230923359492404,
+      "grad_norm": 0.4198856216324367,
+      "learning_rate": 6.8560432283462745e-06,
+      "loss": 0.2046,
+      "step": 3962
+    },
+    {
+      "epoch": 1.3234262815161129,
+      "grad_norm": 0.4506385780580963,
+      "learning_rate": 6.854238607049707e-06,
+      "loss": 0.2107,
+      "step": 3963
+    },
+    {
+      "epoch": 1.3237602270829854,
+      "grad_norm": 0.4144978365544001,
+      "learning_rate": 6.852433705648566e-06,
+      "loss": 0.2053,
+      "step": 3964
+    },
+    {
+      "epoch": 1.324094172649858,
+      "grad_norm": 0.4206897818109025,
+      "learning_rate": 6.8506285244155e-06,
+      "loss": 0.195,
+      "step": 3965
+    },
+    {
+      "epoch": 1.3244281182167308,
+      "grad_norm": 0.4367717494660734,
+      "learning_rate": 6.848823063623207e-06,
+      "loss": 0.1964,
+      "step": 3966
+    },
+    {
+      "epoch": 1.3247620637836033,
+      "grad_norm": 0.42784053652612,
+      "learning_rate": 6.84701732354442e-06,
+      "loss": 0.2128,
+      "step": 3967
+    },
+    {
+      "epoch": 1.3250960093504758,
+      "grad_norm": 0.4303998202870131,
+      "learning_rate": 6.845211304451919e-06,
+      "loss": 0.1982,
+      "step": 3968
+    },
+    {
+      "epoch": 1.3254299549173485,
+      "grad_norm": 0.41585888294104695,
+      "learning_rate": 6.843405006618523e-06,
+      "loss": 0.1887,
+      "step": 3969
+    },
+    {
+      "epoch": 1.325763900484221,
+      "grad_norm": 0.3962726406636628,
+      "learning_rate": 6.841598430317096e-06,
+      "loss": 0.1932,
+      "step": 3970
+    },
+    {
+      "epoch": 1.3260978460510937,
+      "grad_norm": 0.41713419116526673,
+      "learning_rate": 6.839791575820541e-06,
+      "loss": 0.1938,
+      "step": 3971
+    },
+    {
+      "epoch": 1.3264317916179662,
+      "grad_norm": 0.3966219464596358,
+      "learning_rate": 6.837984443401807e-06,
+      "loss": 0.2006,
+      "step": 3972
+    },
+    {
+      "epoch": 1.326765737184839,
+      "grad_norm": 0.4215570099402985,
+      "learning_rate": 6.836177033333882e-06,
+      "loss": 0.2002,
+      "step": 3973
+    },
+    {
+      "epoch": 1.3270996827517114,
+      "grad_norm": 0.39620589259657973,
+      "learning_rate": 6.834369345889793e-06,
+      "loss": 0.1938,
+      "step": 3974
+    },
+    {
+      "epoch": 1.3274336283185841,
+      "grad_norm": 0.43563140461015154,
+      "learning_rate": 6.832561381342617e-06,
+      "loss": 0.2061,
+      "step": 3975
+    },
+    {
+      "epoch": 1.3277675738854566,
+      "grad_norm": 0.4041140648181498,
+      "learning_rate": 6.830753139965467e-06,
+      "loss": 0.1975,
+      "step": 3976
+    },
+    {
+      "epoch": 1.3281015194523293,
+      "grad_norm": 0.41885286235310515,
+      "learning_rate": 6.828944622031497e-06,
+      "loss": 0.2032,
+      "step": 3977
+    },
+    {
+      "epoch": 1.3284354650192018,
+      "grad_norm": 0.4670424648659507,
+      "learning_rate": 6.827135827813909e-06,
+      "loss": 0.2011,
+      "step": 3978
+    },
+    {
+      "epoch": 1.3287694105860743,
+      "grad_norm": 0.45994411686683434,
+      "learning_rate": 6.825326757585939e-06,
+      "loss": 0.2069,
+      "step": 3979
+    },
+    {
+      "epoch": 1.329103356152947,
+      "grad_norm": 0.4039507808169467,
+      "learning_rate": 6.823517411620871e-06,
+      "loss": 0.1889,
+      "step": 3980
+    },
+    {
+      "epoch": 1.3294373017198198,
+      "grad_norm": 0.40372411165182154,
+      "learning_rate": 6.821707790192025e-06,
+      "loss": 0.1921,
+      "step": 3981
+    },
+    {
+      "epoch": 1.3297712472866923,
+      "grad_norm": 0.44376795314218465,
+      "learning_rate": 6.819897893572769e-06,
+      "loss": 0.1962,
+      "step": 3982
+    },
+    {
+      "epoch": 1.3301051928535648,
+      "grad_norm": 0.39179303819655353,
+      "learning_rate": 6.818087722036507e-06,
+      "loss": 0.1891,
+      "step": 3983
+    },
+    {
+      "epoch": 1.3304391384204375,
+      "grad_norm": 0.44009288949714437,
+      "learning_rate": 6.8162772758566875e-06,
+      "loss": 0.2119,
+      "step": 3984
+    },
+    {
+      "epoch": 1.3307730839873102,
+      "grad_norm": 0.4444703006216181,
+      "learning_rate": 6.8144665553067975e-06,
+      "loss": 0.2195,
+      "step": 3985
+    },
+    {
+      "epoch": 1.3311070295541827,
+      "grad_norm": 0.4277024217685211,
+      "learning_rate": 6.812655560660373e-06,
+      "loss": 0.1933,
+      "step": 3986
+    },
+    {
+      "epoch": 1.3314409751210552,
+      "grad_norm": 0.43496112243374163,
+      "learning_rate": 6.810844292190982e-06,
+      "loss": 0.2074,
+      "step": 3987
+    },
+    {
+      "epoch": 1.331774920687928,
+      "grad_norm": 0.45009831755053864,
+      "learning_rate": 6.809032750172236e-06,
+      "loss": 0.2111,
+      "step": 3988
+    },
+    {
+      "epoch": 1.3321088662548004,
+      "grad_norm": 0.43335127708055415,
+      "learning_rate": 6.807220934877794e-06,
+      "loss": 0.2008,
+      "step": 3989
+    },
+    {
+      "epoch": 1.3324428118216731,
+      "grad_norm": 0.4212267467609862,
+      "learning_rate": 6.80540884658135e-06,
+      "loss": 0.1898,
+      "step": 3990
+    },
+    {
+      "epoch": 1.3327767573885456,
+      "grad_norm": 0.4700750950129188,
+      "learning_rate": 6.803596485556643e-06,
+      "loss": 0.2029,
+      "step": 3991
+    },
+    {
+      "epoch": 1.3331107029554183,
+      "grad_norm": 0.415428277497626,
+      "learning_rate": 6.8017838520774494e-06,
+      "loss": 0.1884,
+      "step": 3992
+    },
+    {
+      "epoch": 1.3334446485222908,
+      "grad_norm": 0.43456686746455836,
+      "learning_rate": 6.79997094641759e-06,
+      "loss": 0.2062,
+      "step": 3993
+    },
+    {
+      "epoch": 1.3337785940891635,
+      "grad_norm": 0.477054485573292,
+      "learning_rate": 6.798157768850924e-06,
+      "loss": 0.1951,
+      "step": 3994
+    },
+    {
+      "epoch": 1.334112539656036,
+      "grad_norm": 0.4216294974695778,
+      "learning_rate": 6.796344319651356e-06,
+      "loss": 0.2044,
+      "step": 3995
+    },
+    {
+      "epoch": 1.3344464852229088,
+      "grad_norm": 0.6541115831812444,
+      "learning_rate": 6.794530599092826e-06,
+      "loss": 0.2053,
+      "step": 3996
+    },
+    {
+      "epoch": 1.3347804307897813,
+      "grad_norm": 0.4641533174704176,
+      "learning_rate": 6.792716607449319e-06,
+      "loss": 0.2049,
+      "step": 3997
+    },
+    {
+      "epoch": 1.3351143763566538,
+      "grad_norm": 0.469313295525281,
+      "learning_rate": 6.790902344994861e-06,
+      "loss": 0.2025,
+      "step": 3998
+    },
+    {
+      "epoch": 1.3354483219235265,
+      "grad_norm": 0.4146298442288413,
+      "learning_rate": 6.789087812003516e-06,
+      "loss": 0.2056,
+      "step": 3999
+    },
+    {
+      "epoch": 1.3357822674903992,
+      "grad_norm": 0.4017430523249004,
+      "learning_rate": 6.787273008749391e-06,
+      "loss": 0.1972,
+      "step": 4000
+    },
+    {
+      "epoch": 1.3361162130572717,
+      "grad_norm": 0.4038683460405391,
+      "learning_rate": 6.785457935506634e-06,
+      "loss": 0.1851,
+      "step": 4001
+    },
+    {
+      "epoch": 1.3364501586241442,
+      "grad_norm": 0.4195513584233563,
+      "learning_rate": 6.783642592549433e-06,
+      "loss": 0.1981,
+      "step": 4002
+    },
+    {
+      "epoch": 1.336784104191017,
+      "grad_norm": 0.47255883489969897,
+      "learning_rate": 6.781826980152015e-06,
+      "loss": 0.1976,
+      "step": 4003
+    },
+    {
+      "epoch": 1.3371180497578894,
+      "grad_norm": 0.39785977899861913,
+      "learning_rate": 6.780011098588654e-06,
+      "loss": 0.1893,
+      "step": 4004
+    },
+    {
+      "epoch": 1.337451995324762,
+      "grad_norm": 0.4059364052664352,
+      "learning_rate": 6.778194948133656e-06,
+      "loss": 0.1934,
+      "step": 4005
+    },
+    {
+      "epoch": 1.3377859408916346,
+      "grad_norm": 0.38153108505480937,
+      "learning_rate": 6.776378529061374e-06,
+      "loss": 0.1838,
+      "step": 4006
+    },
+    {
+      "epoch": 1.3381198864585073,
+      "grad_norm": 0.4191943863581517,
+      "learning_rate": 6.774561841646199e-06,
+      "loss": 0.1969,
+      "step": 4007
+    },
+    {
+      "epoch": 1.3384538320253798,
+      "grad_norm": 0.43300160469021504,
+      "learning_rate": 6.772744886162563e-06,
+      "loss": 0.2016,
+      "step": 4008
+    },
+    {
+      "epoch": 1.3387877775922525,
+      "grad_norm": 0.4461504516470582,
+      "learning_rate": 6.770927662884937e-06,
+      "loss": 0.2038,
+      "step": 4009
+    },
+    {
+      "epoch": 1.339121723159125,
+      "grad_norm": 0.4468710082082139,
+      "learning_rate": 6.769110172087838e-06,
+      "loss": 0.2118,
+      "step": 4010
+    },
+    {
+      "epoch": 1.3394556687259978,
+      "grad_norm": 0.42922253720928116,
+      "learning_rate": 6.767292414045816e-06,
+      "loss": 0.1836,
+      "step": 4011
+    },
+    {
+      "epoch": 1.3397896142928702,
+      "grad_norm": 0.7166284001551869,
+      "learning_rate": 6.765474389033464e-06,
+      "loss": 0.1962,
+      "step": 4012
+    },
+    {
+      "epoch": 1.3401235598597427,
+      "grad_norm": 0.5407365657709906,
+      "learning_rate": 6.7636560973254195e-06,
+      "loss": 0.1867,
+      "step": 4013
+    },
+    {
+      "epoch": 1.3404575054266155,
+      "grad_norm": 0.4164038811937711,
+      "learning_rate": 6.761837539196355e-06,
+      "loss": 0.1919,
+      "step": 4014
+    },
+    {
+      "epoch": 1.3407914509934882,
+      "grad_norm": 0.42826189965553885,
+      "learning_rate": 6.760018714920985e-06,
+      "loss": 0.1958,
+      "step": 4015
+    },
+    {
+      "epoch": 1.3411253965603607,
+      "grad_norm": 0.434608355144472,
+      "learning_rate": 6.758199624774065e-06,
+      "loss": 0.1896,
+      "step": 4016
+    },
+    {
+      "epoch": 1.3414593421272332,
+      "grad_norm": 0.4068014354296992,
+      "learning_rate": 6.7563802690303895e-06,
+      "loss": 0.2008,
+      "step": 4017
+    },
+    {
+      "epoch": 1.3417932876941059,
+      "grad_norm": 0.4095906278662473,
+      "learning_rate": 6.7545606479647915e-06,
+      "loss": 0.1943,
+      "step": 4018
+    },
+    {
+      "epoch": 1.3421272332609784,
+      "grad_norm": 0.4169936089160469,
+      "learning_rate": 6.752740761852151e-06,
+      "loss": 0.1997,
+      "step": 4019
+    },
+    {
+      "epoch": 1.342461178827851,
+      "grad_norm": 0.4142264028021034,
+      "learning_rate": 6.7509206109673794e-06,
+      "loss": 0.1978,
+      "step": 4020
+    },
+    {
+      "epoch": 1.3427951243947236,
+      "grad_norm": 0.4059367415693497,
+      "learning_rate": 6.749100195585433e-06,
+      "loss": 0.1994,
+      "step": 4021
+    },
+    {
+      "epoch": 1.3431290699615963,
+      "grad_norm": 0.44463105396489705,
+      "learning_rate": 6.747279515981307e-06,
+      "loss": 0.2101,
+      "step": 4022
+    },
+    {
+      "epoch": 1.3434630155284688,
+      "grad_norm": 0.44157463328609936,
+      "learning_rate": 6.745458572430038e-06,
+      "loss": 0.1992,
+      "step": 4023
+    },
+    {
+      "epoch": 1.3437969610953415,
+      "grad_norm": 0.38829291132548344,
+      "learning_rate": 6.743637365206698e-06,
+      "loss": 0.199,
+      "step": 4024
+    },
+    {
+      "epoch": 1.344130906662214,
+      "grad_norm": 0.4474724309977797,
+      "learning_rate": 6.741815894586404e-06,
+      "loss": 0.2131,
+      "step": 4025
+    },
+    {
+      "epoch": 1.3444648522290867,
+      "grad_norm": 0.5159836340150592,
+      "learning_rate": 6.7399941608443096e-06,
+      "loss": 0.2019,
+      "step": 4026
+    },
+    {
+      "epoch": 1.3447987977959592,
+      "grad_norm": 0.4128853305002455,
+      "learning_rate": 6.7381721642556095e-06,
+      "loss": 0.184,
+      "step": 4027
+    },
+    {
+      "epoch": 1.3451327433628317,
+      "grad_norm": 0.47630481187131113,
+      "learning_rate": 6.736349905095538e-06,
+      "loss": 0.1956,
+      "step": 4028
+    },
+    {
+      "epoch": 1.3454666889297044,
+      "grad_norm": 0.43170867694495907,
+      "learning_rate": 6.734527383639369e-06,
+      "loss": 0.2004,
+      "step": 4029
+    },
+    {
+      "epoch": 1.3458006344965772,
+      "grad_norm": 0.44362403957101815,
+      "learning_rate": 6.732704600162414e-06,
+      "loss": 0.22,
+      "step": 4030
+    },
+    {
+      "epoch": 1.3461345800634497,
+      "grad_norm": 0.4137671915130916,
+      "learning_rate": 6.730881554940029e-06,
+      "loss": 0.202,
+      "step": 4031
+    },
+    {
+      "epoch": 1.3464685256303222,
+      "grad_norm": 0.43747702708097363,
+      "learning_rate": 6.729058248247602e-06,
+      "loss": 0.2066,
+      "step": 4032
+    },
+    {
+      "epoch": 1.3468024711971949,
+      "grad_norm": 0.4331403596331942,
+      "learning_rate": 6.727234680360569e-06,
+      "loss": 0.2067,
+      "step": 4033
+    },
+    {
+      "epoch": 1.3471364167640676,
+      "grad_norm": 0.3868967897202734,
+      "learning_rate": 6.725410851554401e-06,
+      "loss": 0.188,
+      "step": 4034
+    },
+    {
+      "epoch": 1.34747036233094,
+      "grad_norm": 0.5018421015205833,
+      "learning_rate": 6.7235867621046055e-06,
+      "loss": 0.1986,
+      "step": 4035
+    },
+    {
+      "epoch": 1.3478043078978126,
+      "grad_norm": 0.3939057183783324,
+      "learning_rate": 6.721762412286738e-06,
+      "loss": 0.1932,
+      "step": 4036
+    },
+    {
+      "epoch": 1.3481382534646853,
+      "grad_norm": 0.42393788403562216,
+      "learning_rate": 6.719937802376383e-06,
+      "loss": 0.1912,
+      "step": 4037
+    },
+    {
+      "epoch": 1.3484721990315578,
+      "grad_norm": 0.41288963328543327,
+      "learning_rate": 6.718112932649171e-06,
+      "loss": 0.1964,
+      "step": 4038
+    },
+    {
+      "epoch": 1.3488061445984305,
+      "grad_norm": 0.4732934923796236,
+      "learning_rate": 6.716287803380771e-06,
+      "loss": 0.2072,
+      "step": 4039
+    },
+    {
+      "epoch": 1.349140090165303,
+      "grad_norm": 0.4372044972174929,
+      "learning_rate": 6.714462414846891e-06,
+      "loss": 0.2041,
+      "step": 4040
+    },
+    {
+      "epoch": 1.3494740357321757,
+      "grad_norm": 0.4355557384762028,
+      "learning_rate": 6.712636767323273e-06,
+      "loss": 0.2052,
+      "step": 4041
+    },
+    {
+      "epoch": 1.3498079812990482,
+      "grad_norm": 0.4392920698842665,
+      "learning_rate": 6.710810861085708e-06,
+      "loss": 0.2097,
+      "step": 4042
+    },
+    {
+      "epoch": 1.3501419268659207,
+      "grad_norm": 0.406740095621287,
+      "learning_rate": 6.708984696410018e-06,
+      "loss": 0.1893,
+      "step": 4043
+    },
+    {
+      "epoch": 1.3504758724327934,
+      "grad_norm": 0.373673142911883,
+      "learning_rate": 6.707158273572066e-06,
+      "loss": 0.1867,
+      "step": 4044
+    },
+    {
+      "epoch": 1.3508098179996662,
+      "grad_norm": 0.40677087019915287,
+      "learning_rate": 6.7053315928477566e-06,
+      "loss": 0.1944,
+      "step": 4045
+    },
+    {
+      "epoch": 1.3511437635665386,
+      "grad_norm": 0.4062089141523285,
+      "learning_rate": 6.703504654513031e-06,
+      "loss": 0.1898,
+      "step": 4046
+    },
+    {
+      "epoch": 1.3514777091334111,
+      "grad_norm": 0.41024823502203445,
+      "learning_rate": 6.701677458843868e-06,
+      "loss": 0.1926,
+      "step": 4047
+    },
+    {
+      "epoch": 1.3518116547002839,
+      "grad_norm": 0.426250767229244,
+      "learning_rate": 6.6998500061162884e-06,
+      "loss": 0.2071,
+      "step": 4048
+    },
+    {
+      "epoch": 1.3521456002671566,
+      "grad_norm": 0.4039386500414676,
+      "learning_rate": 6.6980222966063516e-06,
+      "loss": 0.1885,
+      "step": 4049
+    },
+    {
+      "epoch": 1.352479545834029,
+      "grad_norm": 0.42655519261390634,
+      "learning_rate": 6.6961943305901515e-06,
+      "loss": 0.1971,
+      "step": 4050
+    },
+    {
+      "epoch": 1.3528134914009016,
+      "grad_norm": 0.5067100050466419,
+      "learning_rate": 6.694366108343827e-06,
+      "loss": 0.2232,
+      "step": 4051
+    },
+    {
+      "epoch": 1.3531474369677743,
+      "grad_norm": 0.48278513038763404,
+      "learning_rate": 6.692537630143551e-06,
+      "loss": 0.1997,
+      "step": 4052
+    },
+    {
+      "epoch": 1.3534813825346468,
+      "grad_norm": 0.4507838778276989,
+      "learning_rate": 6.6907088962655375e-06,
+      "loss": 0.1962,
+      "step": 4053
+    },
+    {
+      "epoch": 1.3538153281015195,
+      "grad_norm": 0.4551173960396972,
+      "learning_rate": 6.688879906986036e-06,
+      "loss": 0.2083,
+      "step": 4054
+    },
+    {
+      "epoch": 1.354149273668392,
+      "grad_norm": 0.4569334112005147,
+      "learning_rate": 6.687050662581341e-06,
+      "loss": 0.2214,
+      "step": 4055
+    },
+    {
+      "epoch": 1.3544832192352647,
+      "grad_norm": 0.4806277133503981,
+      "learning_rate": 6.685221163327778e-06,
+      "loss": 0.1979,
+      "step": 4056
+    },
+    {
+      "epoch": 1.3548171648021372,
+      "grad_norm": 0.4563806677226338,
+      "learning_rate": 6.683391409501715e-06,
+      "loss": 0.2132,
+      "step": 4057
+    },
+    {
+      "epoch": 1.35515111036901,
+      "grad_norm": 0.5042972618637712,
+      "learning_rate": 6.6815614013795595e-06,
+      "loss": 0.2192,
+      "step": 4058
+    },
+    {
+      "epoch": 1.3554850559358824,
+      "grad_norm": 0.44352802619886555,
+      "learning_rate": 6.679731139237753e-06,
+      "loss": 0.2,
+      "step": 4059
+    },
+    {
+      "epoch": 1.3558190015027551,
+      "grad_norm": 0.4350886660962178,
+      "learning_rate": 6.67790062335278e-06,
+      "loss": 0.1952,
+      "step": 4060
+    },
+    {
+      "epoch": 1.3561529470696276,
+      "grad_norm": 0.40422153509801434,
+      "learning_rate": 6.676069854001162e-06,
+      "loss": 0.195,
+      "step": 4061
+    },
+    {
+      "epoch": 1.3564868926365001,
+      "grad_norm": 0.41453093337842917,
+      "learning_rate": 6.674238831459456e-06,
+      "loss": 0.2004,
+      "step": 4062
+    },
+    {
+      "epoch": 1.3568208382033728,
+      "grad_norm": 0.43625037543673406,
+      "learning_rate": 6.672407556004262e-06,
+      "loss": 0.2025,
+      "step": 4063
+    },
+    {
+      "epoch": 1.3571547837702456,
+      "grad_norm": 0.4481372504164815,
+      "learning_rate": 6.670576027912215e-06,
+      "loss": 0.1952,
+      "step": 4064
+    },
+    {
+      "epoch": 1.357488729337118,
+      "grad_norm": 0.4335407448052665,
+      "learning_rate": 6.668744247459988e-06,
+      "loss": 0.1961,
+      "step": 4065
+    },
+    {
+      "epoch": 1.3578226749039906,
+      "grad_norm": 0.44753482938235967,
+      "learning_rate": 6.666912214924295e-06,
+      "loss": 0.2144,
+      "step": 4066
+    },
+    {
+      "epoch": 1.3581566204708633,
+      "grad_norm": 0.3991995293082414,
+      "learning_rate": 6.665079930581883e-06,
+      "loss": 0.1853,
+      "step": 4067
+    },
+    {
+      "epoch": 1.3584905660377358,
+      "grad_norm": 0.42486029468962033,
+      "learning_rate": 6.663247394709542e-06,
+      "loss": 0.1962,
+      "step": 4068
+    },
+    {
+      "epoch": 1.3588245116046085,
+      "grad_norm": 0.4266054882931973,
+      "learning_rate": 6.661414607584099e-06,
+      "loss": 0.1975,
+      "step": 4069
+    },
+    {
+      "epoch": 1.359158457171481,
+      "grad_norm": 0.432581270411452,
+      "learning_rate": 6.659581569482415e-06,
+      "loss": 0.2044,
+      "step": 4070
+    },
+    {
+      "epoch": 1.3594924027383537,
+      "grad_norm": 0.40150080646226216,
+      "learning_rate": 6.657748280681395e-06,
+      "loss": 0.2039,
+      "step": 4071
+    },
+    {
+      "epoch": 1.3598263483052262,
+      "grad_norm": 0.4225694559806857,
+      "learning_rate": 6.65591474145798e-06,
+      "loss": 0.204,
+      "step": 4072
+    },
+    {
+      "epoch": 1.360160293872099,
+      "grad_norm": 0.45371512908538403,
+      "learning_rate": 6.6540809520891425e-06,
+      "loss": 0.1996,
+      "step": 4073
+    },
+    {
+      "epoch": 1.3604942394389714,
+      "grad_norm": 0.4263029427537738,
+      "learning_rate": 6.652246912851903e-06,
+      "loss": 0.1985,
+      "step": 4074
+    },
+    {
+      "epoch": 1.3608281850058441,
+      "grad_norm": 0.4403693400801846,
+      "learning_rate": 6.650412624023311e-06,
+      "loss": 0.2079,
+      "step": 4075
+    },
+    {
+      "epoch": 1.3611621305727166,
+      "grad_norm": 0.4672726128852996,
+      "learning_rate": 6.648578085880461e-06,
+      "loss": 0.2114,
+      "step": 4076
+    },
+    {
+      "epoch": 1.3614960761395891,
+      "grad_norm": 0.42076978881075816,
+      "learning_rate": 6.64674329870048e-06,
+      "loss": 0.1915,
+      "step": 4077
+    },
+    {
+      "epoch": 1.3618300217064618,
+      "grad_norm": 0.4074124100582979,
+      "learning_rate": 6.644908262760531e-06,
+      "loss": 0.196,
+      "step": 4078
+    },
+    {
+      "epoch": 1.3621639672733346,
+      "grad_norm": 0.43709488826547427,
+      "learning_rate": 6.643072978337823e-06,
+      "loss": 0.2024,
+      "step": 4079
+    },
+    {
+      "epoch": 1.362497912840207,
+      "grad_norm": 0.4410887006441817,
+      "learning_rate": 6.641237445709595e-06,
+      "loss": 0.2083,
+      "step": 4080
+    },
+    {
+      "epoch": 1.3628318584070795,
+      "grad_norm": 0.41075833370114767,
+      "learning_rate": 6.639401665153126e-06,
+      "loss": 0.1874,
+      "step": 4081
+    },
+    {
+      "epoch": 1.3631658039739523,
+      "grad_norm": 0.4314059293595324,
+      "learning_rate": 6.637565636945731e-06,
+      "loss": 0.1969,
+      "step": 4082
+    },
+    {
+      "epoch": 1.363499749540825,
+      "grad_norm": 0.3962845145124328,
+      "learning_rate": 6.635729361364765e-06,
+      "loss": 0.182,
+      "step": 4083
+    },
+    {
+      "epoch": 1.3638336951076975,
+      "grad_norm": 0.45085289607224943,
+      "learning_rate": 6.633892838687621e-06,
+      "loss": 0.2153,
+      "step": 4084
+    },
+    {
+      "epoch": 1.36416764067457,
+      "grad_norm": 0.442502443264247,
+      "learning_rate": 6.632056069191723e-06,
+      "loss": 0.203,
+      "step": 4085
+    },
+    {
+      "epoch": 1.3645015862414427,
+      "grad_norm": 0.4331715292207473,
+      "learning_rate": 6.6302190531545395e-06,
+      "loss": 0.1989,
+      "step": 4086
+    },
+    {
+      "epoch": 1.3648355318083152,
+      "grad_norm": 0.4526521135216819,
+      "learning_rate": 6.628381790853573e-06,
+      "loss": 0.2094,
+      "step": 4087
+    },
+    {
+      "epoch": 1.365169477375188,
+      "grad_norm": 0.40449183201031125,
+      "learning_rate": 6.626544282566363e-06,
+      "loss": 0.1949,
+      "step": 4088
+    },
+    {
+      "epoch": 1.3655034229420604,
+      "grad_norm": 0.40387227018258237,
+      "learning_rate": 6.624706528570487e-06,
+      "loss": 0.1988,
+      "step": 4089
+    },
+    {
+      "epoch": 1.3658373685089331,
+      "grad_norm": 0.5711678934538252,
+      "learning_rate": 6.6228685291435605e-06,
+      "loss": 0.2188,
+      "step": 4090
+    },
+    {
+      "epoch": 1.3661713140758056,
+      "grad_norm": 0.4421358421448025,
+      "learning_rate": 6.621030284563232e-06,
+      "loss": 0.1824,
+      "step": 4091
+    },
+    {
+      "epoch": 1.366505259642678,
+      "grad_norm": 0.4559202259210313,
+      "learning_rate": 6.619191795107195e-06,
+      "loss": 0.217,
+      "step": 4092
+    },
+    {
+      "epoch": 1.3668392052095508,
+      "grad_norm": 0.4244083311195419,
+      "learning_rate": 6.617353061053171e-06,
+      "loss": 0.2009,
+      "step": 4093
+    },
+    {
+      "epoch": 1.3671731507764235,
+      "grad_norm": 0.4173771853428039,
+      "learning_rate": 6.615514082678922e-06,
+      "loss": 0.1872,
+      "step": 4094
+    },
+    {
+      "epoch": 1.367507096343296,
+      "grad_norm": 0.4377390377861039,
+      "learning_rate": 6.613674860262249e-06,
+      "loss": 0.2141,
+      "step": 4095
+    },
+    {
+      "epoch": 1.3678410419101685,
+      "grad_norm": 0.406886823763542,
+      "learning_rate": 6.61183539408099e-06,
+      "loss": 0.1906,
+      "step": 4096
+    },
+    {
+      "epoch": 1.3681749874770412,
+      "grad_norm": 0.42851766250941675,
+      "learning_rate": 6.609995684413013e-06,
+      "loss": 0.1965,
+      "step": 4097
+    },
+    {
+      "epoch": 1.368508933043914,
+      "grad_norm": 0.622902793386304,
+      "learning_rate": 6.608155731536233e-06,
+      "loss": 0.1968,
+      "step": 4098
+    },
+    {
+      "epoch": 1.3688428786107865,
+      "grad_norm": 0.3853625281666951,
+      "learning_rate": 6.606315535728594e-06,
+      "loss": 0.1816,
+      "step": 4099
+    },
+    {
+      "epoch": 1.369176824177659,
+      "grad_norm": 0.4666928758678988,
+      "learning_rate": 6.604475097268079e-06,
+      "loss": 0.202,
+      "step": 4100
+    },
+    {
+      "epoch": 1.3695107697445317,
+      "grad_norm": 0.3771661413941458,
+      "learning_rate": 6.602634416432708e-06,
+      "loss": 0.1874,
+      "step": 4101
+    },
+    {
+      "epoch": 1.3698447153114042,
+      "grad_norm": 0.4610010980819958,
+      "learning_rate": 6.600793493500539e-06,
+      "loss": 0.2075,
+      "step": 4102
+    },
+    {
+      "epoch": 1.3701786608782769,
+      "grad_norm": 0.41246814662170656,
+      "learning_rate": 6.5989523287496645e-06,
+      "loss": 0.1907,
+      "step": 4103
+    },
+    {
+      "epoch": 1.3705126064451494,
+      "grad_norm": 0.44901774277549816,
+      "learning_rate": 6.597110922458214e-06,
+      "loss": 0.1955,
+      "step": 4104
+    },
+    {
+      "epoch": 1.370846552012022,
+      "grad_norm": 0.3989788461433798,
+      "learning_rate": 6.595269274904351e-06,
+      "loss": 0.1971,
+      "step": 4105
+    },
+    {
+      "epoch": 1.3711804975788946,
+      "grad_norm": 0.4601759741943717,
+      "learning_rate": 6.593427386366282e-06,
+      "loss": 0.2044,
+      "step": 4106
+    },
+    {
+      "epoch": 1.3715144431457673,
+      "grad_norm": 0.4075278252757786,
+      "learning_rate": 6.591585257122244e-06,
+      "loss": 0.1929,
+      "step": 4107
+    },
+    {
+      "epoch": 1.3718483887126398,
+      "grad_norm": 0.45268827886168284,
+      "learning_rate": 6.589742887450512e-06,
+      "loss": 0.2012,
+      "step": 4108
+    },
+    {
+      "epoch": 1.3721823342795125,
+      "grad_norm": 0.39829538551301213,
+      "learning_rate": 6.5879002776294e-06,
+      "loss": 0.1934,
+      "step": 4109
+    },
+    {
+      "epoch": 1.372516279846385,
+      "grad_norm": 0.4075145132486534,
+      "learning_rate": 6.586057427937252e-06,
+      "loss": 0.2013,
+      "step": 4110
+    },
+    {
+      "epoch": 1.3728502254132575,
+      "grad_norm": 0.413224264908335,
+      "learning_rate": 6.584214338652455e-06,
+      "loss": 0.1885,
+      "step": 4111
+    },
+    {
+      "epoch": 1.3731841709801302,
+      "grad_norm": 0.42233907535924026,
+      "learning_rate": 6.582371010053429e-06,
+      "loss": 0.2007,
+      "step": 4112
+    },
+    {
+      "epoch": 1.373518116547003,
+      "grad_norm": 0.4689784474170459,
+      "learning_rate": 6.58052744241863e-06,
+      "loss": 0.2026,
+      "step": 4113
+    },
+    {
+      "epoch": 1.3738520621138754,
+      "grad_norm": 0.42477561232876815,
+      "learning_rate": 6.578683636026551e-06,
+      "loss": 0.203,
+      "step": 4114
+    },
+    {
+      "epoch": 1.374186007680748,
+      "grad_norm": 0.4309943871533634,
+      "learning_rate": 6.576839591155719e-06,
+      "loss": 0.1938,
+      "step": 4115
+    },
+    {
+      "epoch": 1.3745199532476207,
+      "grad_norm": 0.5615884354542333,
+      "learning_rate": 6.574995308084702e-06,
+      "loss": 0.1806,
+      "step": 4116
+    },
+    {
+      "epoch": 1.3748538988144932,
+      "grad_norm": 0.3813788802372199,
+      "learning_rate": 6.573150787092097e-06,
+      "loss": 0.1733,
+      "step": 4117
+    },
+    {
+      "epoch": 1.3751878443813659,
+      "grad_norm": 0.42148123715265845,
+      "learning_rate": 6.5713060284565435e-06,
+      "loss": 0.1896,
+      "step": 4118
+    },
+    {
+      "epoch": 1.3755217899482384,
+      "grad_norm": 0.431883451577582,
+      "learning_rate": 6.569461032456713e-06,
+      "loss": 0.2186,
+      "step": 4119
+    },
+    {
+      "epoch": 1.375855735515111,
+      "grad_norm": 0.4066532875757682,
+      "learning_rate": 6.567615799371313e-06,
+      "loss": 0.1902,
+      "step": 4120
+    },
+    {
+      "epoch": 1.3761896810819836,
+      "grad_norm": 0.4370833319403715,
+      "learning_rate": 6.565770329479089e-06,
+      "loss": 0.2035,
+      "step": 4121
+    },
+    {
+      "epoch": 1.3765236266488563,
+      "grad_norm": 0.39098258449611045,
+      "learning_rate": 6.5639246230588205e-06,
+      "loss": 0.1918,
+      "step": 4122
+    },
+    {
+      "epoch": 1.3768575722157288,
+      "grad_norm": 0.39463295296495227,
+      "learning_rate": 6.562078680389323e-06,
+      "loss": 0.1874,
+      "step": 4123
+    },
+    {
+      "epoch": 1.3771915177826015,
+      "grad_norm": 0.7838749092433236,
+      "learning_rate": 6.560232501749446e-06,
+      "loss": 0.2093,
+      "step": 4124
+    },
+    {
+      "epoch": 1.377525463349474,
+      "grad_norm": 0.45124048458861776,
+      "learning_rate": 6.558386087418082e-06,
+      "loss": 0.2058,
+      "step": 4125
+    },
+    {
+      "epoch": 1.3778594089163465,
+      "grad_norm": 0.44508050673501187,
+      "learning_rate": 6.556539437674147e-06,
+      "loss": 0.2119,
+      "step": 4126
+    },
+    {
+      "epoch": 1.3781933544832192,
+      "grad_norm": 0.4532245305883286,
+      "learning_rate": 6.554692552796604e-06,
+      "loss": 0.1988,
+      "step": 4127
+    },
+    {
+      "epoch": 1.378527300050092,
+      "grad_norm": 0.4366039970271216,
+      "learning_rate": 6.552845433064445e-06,
+      "loss": 0.2016,
+      "step": 4128
+    },
+    {
+      "epoch": 1.3788612456169644,
+      "grad_norm": 0.43991279957977425,
+      "learning_rate": 6.550998078756698e-06,
+      "loss": 0.2125,
+      "step": 4129
+    },
+    {
+      "epoch": 1.379195191183837,
+      "grad_norm": 0.45075051528764204,
+      "learning_rate": 6.549150490152429e-06,
+      "loss": 0.2124,
+      "step": 4130
+    },
+    {
+      "epoch": 1.3795291367507097,
+      "grad_norm": 0.4697620488513748,
+      "learning_rate": 6.5473026675307394e-06,
+      "loss": 0.214,
+      "step": 4131
+    },
+    {
+      "epoch": 1.3798630823175824,
+      "grad_norm": 0.4650812525046603,
+      "learning_rate": 6.545454611170762e-06,
+      "loss": 0.2046,
+      "step": 4132
+    },
+    {
+      "epoch": 1.3801970278844549,
+      "grad_norm": 0.4473205624182132,
+      "learning_rate": 6.543606321351668e-06,
+      "loss": 0.1995,
+      "step": 4133
+    },
+    {
+      "epoch": 1.3805309734513274,
+      "grad_norm": 0.43793988430119773,
+      "learning_rate": 6.541757798352664e-06,
+      "loss": 0.1964,
+      "step": 4134
+    },
+    {
+      "epoch": 1.3808649190182,
+      "grad_norm": 0.46122603750033925,
+      "learning_rate": 6.539909042452991e-06,
+      "loss": 0.2021,
+      "step": 4135
+    },
+    {
+      "epoch": 1.3811988645850726,
+      "grad_norm": 0.4431099270741144,
+      "learning_rate": 6.538060053931925e-06,
+      "loss": 0.2151,
+      "step": 4136
+    },
+    {
+      "epoch": 1.3815328101519453,
+      "grad_norm": 0.44155698593464704,
+      "learning_rate": 6.536210833068779e-06,
+      "loss": 0.1947,
+      "step": 4137
+    },
+    {
+      "epoch": 1.3818667557188178,
+      "grad_norm": 0.4109387527355864,
+      "learning_rate": 6.534361380142896e-06,
+      "loss": 0.1998,
+      "step": 4138
+    },
+    {
+      "epoch": 1.3822007012856905,
+      "grad_norm": 0.4603948218667672,
+      "learning_rate": 6.532511695433662e-06,
+      "loss": 0.1966,
+      "step": 4139
+    },
+    {
+      "epoch": 1.382534646852563,
+      "grad_norm": 0.44282145799352707,
+      "learning_rate": 6.5306617792204915e-06,
+      "loss": 0.1896,
+      "step": 4140
+    },
+    {
+      "epoch": 1.3828685924194355,
+      "grad_norm": 0.44104795387567414,
+      "learning_rate": 6.528811631782835e-06,
+      "loss": 0.2068,
+      "step": 4141
+    },
+    {
+      "epoch": 1.3832025379863082,
+      "grad_norm": 0.48858635612357,
+      "learning_rate": 6.526961253400181e-06,
+      "loss": 0.2144,
+      "step": 4142
+    },
+    {
+      "epoch": 1.383536483553181,
+      "grad_norm": 0.44312917050300416,
+      "learning_rate": 6.525110644352052e-06,
+      "loss": 0.2069,
+      "step": 4143
+    },
+    {
+      "epoch": 1.3838704291200534,
+      "grad_norm": 0.3981097773309502,
+      "learning_rate": 6.523259804918001e-06,
+      "loss": 0.1897,
+      "step": 4144
+    },
+    {
+      "epoch": 1.384204374686926,
+      "grad_norm": 0.4204274275796497,
+      "learning_rate": 6.52140873537762e-06,
+      "loss": 0.1838,
+      "step": 4145
+    },
+    {
+      "epoch": 1.3845383202537986,
+      "grad_norm": 0.46326487504021546,
+      "learning_rate": 6.519557436010535e-06,
+      "loss": 0.2145,
+      "step": 4146
+    },
+    {
+      "epoch": 1.3848722658206714,
+      "grad_norm": 0.4722683700335305,
+      "learning_rate": 6.51770590709641e-06,
+      "loss": 0.2081,
+      "step": 4147
+    },
+    {
+      "epoch": 1.3852062113875439,
+      "grad_norm": 0.4311480901777335,
+      "learning_rate": 6.515854148914935e-06,
+      "loss": 0.2024,
+      "step": 4148
+    },
+    {
+      "epoch": 1.3855401569544163,
+      "grad_norm": 0.46985472062528916,
+      "learning_rate": 6.514002161745844e-06,
+      "loss": 0.2183,
+      "step": 4149
+    },
+    {
+      "epoch": 1.385874102521289,
+      "grad_norm": 0.4803730015949404,
+      "learning_rate": 6.512149945868898e-06,
+      "loss": 0.2193,
+      "step": 4150
+    },
+    {
+      "epoch": 1.3862080480881616,
+      "grad_norm": 0.4597435238205787,
+      "learning_rate": 6.510297501563899e-06,
+      "loss": 0.2072,
+      "step": 4151
+    },
+    {
+      "epoch": 1.3865419936550343,
+      "grad_norm": 0.43218955473225323,
+      "learning_rate": 6.5084448291106785e-06,
+      "loss": 0.2076,
+      "step": 4152
+    },
+    {
+      "epoch": 1.3868759392219068,
+      "grad_norm": 0.4150408431997899,
+      "learning_rate": 6.506591928789105e-06,
+      "loss": 0.1936,
+      "step": 4153
+    },
+    {
+      "epoch": 1.3872098847887795,
+      "grad_norm": 0.4312232452910037,
+      "learning_rate": 6.504738800879081e-06,
+      "loss": 0.1933,
+      "step": 4154
+    },
+    {
+      "epoch": 1.387543830355652,
+      "grad_norm": 0.4064866596866989,
+      "learning_rate": 6.502885445660544e-06,
+      "loss": 0.184,
+      "step": 4155
+    },
+    {
+      "epoch": 1.3878777759225247,
+      "grad_norm": 0.3829518324598277,
+      "learning_rate": 6.501031863413464e-06,
+      "loss": 0.1746,
+      "step": 4156
+    },
+    {
+      "epoch": 1.3882117214893972,
+      "grad_norm": 0.45492095092003937,
+      "learning_rate": 6.499178054417847e-06,
+      "loss": 0.2152,
+      "step": 4157
+    },
+    {
+      "epoch": 1.38854566705627,
+      "grad_norm": 0.4457187903715593,
+      "learning_rate": 6.497324018953732e-06,
+      "loss": 0.2061,
+      "step": 4158
+    },
+    {
+      "epoch": 1.3888796126231424,
+      "grad_norm": 0.4466194927024366,
+      "learning_rate": 6.495469757301196e-06,
+      "loss": 0.1999,
+      "step": 4159
+    },
+    {
+      "epoch": 1.389213558190015,
+      "grad_norm": 0.4355326082302892,
+      "learning_rate": 6.493615269740343e-06,
+      "loss": 0.2141,
+      "step": 4160
+    },
+    {
+      "epoch": 1.3895475037568876,
+      "grad_norm": 0.3940978415054905,
+      "learning_rate": 6.491760556551315e-06,
+      "loss": 0.188,
+      "step": 4161
+    },
+    {
+      "epoch": 1.3898814493237603,
+      "grad_norm": 0.4292342258128804,
+      "learning_rate": 6.489905618014293e-06,
+      "loss": 0.2138,
+      "step": 4162
+    },
+    {
+      "epoch": 1.3902153948906328,
+      "grad_norm": 0.43189024487894684,
+      "learning_rate": 6.488050454409483e-06,
+      "loss": 0.1972,
+      "step": 4163
+    },
+    {
+      "epoch": 1.3905493404575053,
+      "grad_norm": 0.4053167522949468,
+      "learning_rate": 6.486195066017129e-06,
+      "loss": 0.1958,
+      "step": 4164
+    },
+    {
+      "epoch": 1.390883286024378,
+      "grad_norm": 0.45743707193697686,
+      "learning_rate": 6.484339453117514e-06,
+      "loss": 0.2064,
+      "step": 4165
+    },
+    {
+      "epoch": 1.3912172315912505,
+      "grad_norm": 0.3727514433378026,
+      "learning_rate": 6.482483615990945e-06,
+      "loss": 0.1789,
+      "step": 4166
+    },
+    {
+      "epoch": 1.3915511771581233,
+      "grad_norm": 0.38894137550668284,
+      "learning_rate": 6.480627554917771e-06,
+      "loss": 0.1897,
+      "step": 4167
+    },
+    {
+      "epoch": 1.3918851227249958,
+      "grad_norm": 0.3949005314394,
+      "learning_rate": 6.47877127017837e-06,
+      "loss": 0.184,
+      "step": 4168
+    },
+    {
+      "epoch": 1.3922190682918685,
+      "grad_norm": 0.5483421957144745,
+      "learning_rate": 6.476914762053158e-06,
+      "loss": 0.2248,
+      "step": 4169
+    },
+    {
+      "epoch": 1.392553013858741,
+      "grad_norm": 0.4349485020907497,
+      "learning_rate": 6.47505803082258e-06,
+      "loss": 0.2087,
+      "step": 4170
+    },
+    {
+      "epoch": 1.3928869594256137,
+      "grad_norm": 0.494753044097071,
+      "learning_rate": 6.473201076767119e-06,
+      "loss": 0.2029,
+      "step": 4171
+    },
+    {
+      "epoch": 1.3932209049924862,
+      "grad_norm": 0.402889777781755,
+      "learning_rate": 6.471343900167289e-06,
+      "loss": 0.1958,
+      "step": 4172
+    },
+    {
+      "epoch": 1.393554850559359,
+      "grad_norm": 0.42334287046577884,
+      "learning_rate": 6.469486501303639e-06,
+      "loss": 0.2183,
+      "step": 4173
+    },
+    {
+      "epoch": 1.3938887961262314,
+      "grad_norm": 0.3947978066614757,
+      "learning_rate": 6.467628880456749e-06,
+      "loss": 0.1865,
+      "step": 4174
+    },
+    {
+      "epoch": 1.394222741693104,
+      "grad_norm": 0.47629578193143574,
+      "learning_rate": 6.465771037907236e-06,
+      "loss": 0.2095,
+      "step": 4175
+    },
+    {
+      "epoch": 1.3945566872599766,
+      "grad_norm": 0.3880989484453497,
+      "learning_rate": 6.463912973935749e-06,
+      "loss": 0.1855,
+      "step": 4176
+    },
+    {
+      "epoch": 1.3948906328268493,
+      "grad_norm": 0.47037211387441913,
+      "learning_rate": 6.462054688822971e-06,
+      "loss": 0.2078,
+      "step": 4177
+    },
+    {
+      "epoch": 1.3952245783937218,
+      "grad_norm": 0.382314312114383,
+      "learning_rate": 6.460196182849616e-06,
+      "loss": 0.1982,
+      "step": 4178
+    },
+    {
+      "epoch": 1.3955585239605943,
+      "grad_norm": 0.41109789004350694,
+      "learning_rate": 6.458337456296434e-06,
+      "loss": 0.2024,
+      "step": 4179
+    },
+    {
+      "epoch": 1.395892469527467,
+      "grad_norm": 0.4899142261094805,
+      "learning_rate": 6.456478509444209e-06,
+      "loss": 0.1993,
+      "step": 4180
+    },
+    {
+      "epoch": 1.3962264150943398,
+      "grad_norm": 0.42665421483230975,
+      "learning_rate": 6.454619342573756e-06,
+      "loss": 0.2041,
+      "step": 4181
+    },
+    {
+      "epoch": 1.3965603606612123,
+      "grad_norm": 0.41216556645235136,
+      "learning_rate": 6.452759955965922e-06,
+      "loss": 0.1884,
+      "step": 4182
+    },
+    {
+      "epoch": 1.3968943062280847,
+      "grad_norm": 0.4322046656673873,
+      "learning_rate": 6.450900349901592e-06,
+      "loss": 0.2001,
+      "step": 4183
+    },
+    {
+      "epoch": 1.3972282517949575,
+      "grad_norm": 0.4285030575773755,
+      "learning_rate": 6.449040524661681e-06,
+      "loss": 0.1906,
+      "step": 4184
+    },
+    {
+      "epoch": 1.39756219736183,
+      "grad_norm": 0.43465596091020026,
+      "learning_rate": 6.447180480527135e-06,
+      "loss": 0.2021,
+      "step": 4185
+    },
+    {
+      "epoch": 1.3978961429287027,
+      "grad_norm": 0.41971935181118564,
+      "learning_rate": 6.445320217778939e-06,
+      "loss": 0.1957,
+      "step": 4186
+    },
+    {
+      "epoch": 1.3982300884955752,
+      "grad_norm": 0.4583440890267203,
+      "learning_rate": 6.443459736698106e-06,
+      "loss": 0.2051,
+      "step": 4187
+    },
+    {
+      "epoch": 1.398564034062448,
+      "grad_norm": 0.4406778185384973,
+      "learning_rate": 6.4415990375656826e-06,
+      "loss": 0.2003,
+      "step": 4188
+    },
+    {
+      "epoch": 1.3988979796293204,
+      "grad_norm": 0.5268157377750877,
+      "learning_rate": 6.4397381206627505e-06,
+      "loss": 0.2256,
+      "step": 4189
+    },
+    {
+      "epoch": 1.3992319251961929,
+      "grad_norm": 0.40175298789569275,
+      "learning_rate": 6.437876986270424e-06,
+      "loss": 0.1926,
+      "step": 4190
+    },
+    {
+      "epoch": 1.3995658707630656,
+      "grad_norm": 0.43113117207329726,
+      "learning_rate": 6.436015634669848e-06,
+      "loss": 0.2101,
+      "step": 4191
+    },
+    {
+      "epoch": 1.3998998163299383,
+      "grad_norm": 0.40792889401372284,
+      "learning_rate": 6.434154066142201e-06,
+      "loss": 0.1982,
+      "step": 4192
+    },
+    {
+      "epoch": 1.4002337618968108,
+      "grad_norm": 0.3980086772220684,
+      "learning_rate": 6.432292280968695e-06,
+      "loss": 0.2006,
+      "step": 4193
+    },
+    {
+      "epoch": 1.4005677074636833,
+      "grad_norm": 0.45321556751445036,
+      "learning_rate": 6.430430279430577e-06,
+      "loss": 0.2076,
+      "step": 4194
+    },
+    {
+      "epoch": 1.400901653030556,
+      "grad_norm": 0.3962084842360654,
+      "learning_rate": 6.428568061809122e-06,
+      "loss": 0.2022,
+      "step": 4195
+    },
+    {
+      "epoch": 1.4012355985974287,
+      "grad_norm": 0.38721195579949064,
+      "learning_rate": 6.426705628385641e-06,
+      "loss": 0.1928,
+      "step": 4196
+    },
+    {
+      "epoch": 1.4015695441643012,
+      "grad_norm": 0.43865314302481156,
+      "learning_rate": 6.4248429794414745e-06,
+      "loss": 0.1984,
+      "step": 4197
+    },
+    {
+      "epoch": 1.4019034897311737,
+      "grad_norm": 0.4245277931246651,
+      "learning_rate": 6.422980115258e-06,
+      "loss": 0.2047,
+      "step": 4198
+    },
+    {
+      "epoch": 1.4022374352980465,
+      "grad_norm": 0.40245781599580605,
+      "learning_rate": 6.421117036116624e-06,
+      "loss": 0.1939,
+      "step": 4199
+    },
+    {
+      "epoch": 1.402571380864919,
+      "grad_norm": 0.42437101178064845,
+      "learning_rate": 6.4192537422987864e-06,
+      "loss": 0.1898,
+      "step": 4200
+    },
+    {
+      "epoch": 1.4029053264317917,
+      "grad_norm": 0.41477729386611645,
+      "learning_rate": 6.417390234085961e-06,
+      "loss": 0.2016,
+      "step": 4201
+    },
+    {
+      "epoch": 1.4032392719986642,
+      "grad_norm": 0.39992406322586277,
+      "learning_rate": 6.415526511759649e-06,
+      "loss": 0.1922,
+      "step": 4202
+    },
+    {
+      "epoch": 1.4035732175655369,
+      "grad_norm": 0.45466666242275616,
+      "learning_rate": 6.413662575601391e-06,
+      "loss": 0.2148,
+      "step": 4203
+    },
+    {
+      "epoch": 1.4039071631324094,
+      "grad_norm": 0.3778303865454364,
+      "learning_rate": 6.4117984258927565e-06,
+      "loss": 0.1831,
+      "step": 4204
+    },
+    {
+      "epoch": 1.404241108699282,
+      "grad_norm": 0.4005745223294027,
+      "learning_rate": 6.409934062915345e-06,
+      "loss": 0.1847,
+      "step": 4205
+    },
+    {
+      "epoch": 1.4045750542661546,
+      "grad_norm": 0.4584571575972361,
+      "learning_rate": 6.408069486950793e-06,
+      "loss": 0.2045,
+      "step": 4206
+    },
+    {
+      "epoch": 1.4049089998330273,
+      "grad_norm": 0.41082051621126475,
+      "learning_rate": 6.406204698280766e-06,
+      "loss": 0.1912,
+      "step": 4207
+    },
+    {
+      "epoch": 1.4052429453998998,
+      "grad_norm": 0.3901252288030012,
+      "learning_rate": 6.40433969718696e-06,
+      "loss": 0.1807,
+      "step": 4208
+    },
+    {
+      "epoch": 1.4055768909667723,
+      "grad_norm": 0.4501067185058659,
+      "learning_rate": 6.402474483951109e-06,
+      "loss": 0.2018,
+      "step": 4209
+    },
+    {
+      "epoch": 1.405910836533645,
+      "grad_norm": 0.38450772473450306,
+      "learning_rate": 6.400609058854973e-06,
+      "loss": 0.1823,
+      "step": 4210
+    },
+    {
+      "epoch": 1.4062447821005177,
+      "grad_norm": 0.40515453066428003,
+      "learning_rate": 6.398743422180346e-06,
+      "loss": 0.2034,
+      "step": 4211
+    },
+    {
+      "epoch": 1.4065787276673902,
+      "grad_norm": 0.4239306652790811,
+      "learning_rate": 6.396877574209057e-06,
+      "loss": 0.2087,
+      "step": 4212
+    },
+    {
+      "epoch": 1.4069126732342627,
+      "grad_norm": 0.4566772627357021,
+      "learning_rate": 6.395011515222962e-06,
+      "loss": 0.1948,
+      "step": 4213
+    },
+    {
+      "epoch": 1.4072466188011354,
+      "grad_norm": 0.4000662579941561,
+      "learning_rate": 6.393145245503951e-06,
+      "loss": 0.1955,
+      "step": 4214
+    },
+    {
+      "epoch": 1.407580564368008,
+      "grad_norm": 0.4201143343811621,
+      "learning_rate": 6.391278765333948e-06,
+      "loss": 0.2011,
+      "step": 4215
+    },
+    {
+      "epoch": 1.4079145099348807,
+      "grad_norm": 0.39637080003077907,
+      "learning_rate": 6.389412074994906e-06,
+      "loss": 0.1937,
+      "step": 4216
+    },
+    {
+      "epoch": 1.4082484555017531,
+      "grad_norm": 0.4258908660767943,
+      "learning_rate": 6.387545174768809e-06,
+      "loss": 0.201,
+      "step": 4217
+    },
+    {
+      "epoch": 1.4085824010686259,
+      "grad_norm": 0.46190972235436634,
+      "learning_rate": 6.385678064937677e-06,
+      "loss": 0.219,
+      "step": 4218
+    },
+    {
+      "epoch": 1.4089163466354984,
+      "grad_norm": 0.40798704121606794,
+      "learning_rate": 6.383810745783556e-06,
+      "loss": 0.2036,
+      "step": 4219
+    },
+    {
+      "epoch": 1.409250292202371,
+      "grad_norm": 0.3900248306224213,
+      "learning_rate": 6.38194321758853e-06,
+      "loss": 0.182,
+      "step": 4220
+    },
+    {
+      "epoch": 1.4095842377692436,
+      "grad_norm": 0.41143107768977444,
+      "learning_rate": 6.3800754806347065e-06,
+      "loss": 0.2046,
+      "step": 4221
+    },
+    {
+      "epoch": 1.4099181833361163,
+      "grad_norm": 0.4227528210432267,
+      "learning_rate": 6.378207535204234e-06,
+      "loss": 0.2068,
+      "step": 4222
+    },
+    {
+      "epoch": 1.4102521289029888,
+      "grad_norm": 0.4496275000832063,
+      "learning_rate": 6.376339381579285e-06,
+      "loss": 0.188,
+      "step": 4223
+    },
+    {
+      "epoch": 1.4105860744698613,
+      "grad_norm": 0.3920155752449724,
+      "learning_rate": 6.374471020042067e-06,
+      "loss": 0.1915,
+      "step": 4224
+    },
+    {
+      "epoch": 1.410920020036734,
+      "grad_norm": 0.4893281271897962,
+      "learning_rate": 6.372602450874816e-06,
+      "loss": 0.2051,
+      "step": 4225
+    },
+    {
+      "epoch": 1.4112539656036067,
+      "grad_norm": 0.3895814494862588,
+      "learning_rate": 6.370733674359803e-06,
+      "loss": 0.2008,
+      "step": 4226
+    },
+    {
+      "epoch": 1.4115879111704792,
+      "grad_norm": 0.4464740149177025,
+      "learning_rate": 6.36886469077933e-06,
+      "loss": 0.1996,
+      "step": 4227
+    },
+    {
+      "epoch": 1.4119218567373517,
+      "grad_norm": 0.41982386936019905,
+      "learning_rate": 6.366995500415727e-06,
+      "loss": 0.1789,
+      "step": 4228
+    },
+    {
+      "epoch": 1.4122558023042244,
+      "grad_norm": 0.48692131344551154,
+      "learning_rate": 6.365126103551358e-06,
+      "loss": 0.1914,
+      "step": 4229
+    },
+    {
+      "epoch": 1.4125897478710971,
+      "grad_norm": 0.4188619004865136,
+      "learning_rate": 6.363256500468617e-06,
+      "loss": 0.1926,
+      "step": 4230
+    },
+    {
+      "epoch": 1.4129236934379696,
+      "grad_norm": 0.38428575637852025,
+      "learning_rate": 6.3613866914499285e-06,
+      "loss": 0.1913,
+      "step": 4231
+    },
+    {
+      "epoch": 1.4132576390048421,
+      "grad_norm": 0.4192523891713591,
+      "learning_rate": 6.359516676777751e-06,
+      "loss": 0.2058,
+      "step": 4232
+    },
+    {
+      "epoch": 1.4135915845717149,
+      "grad_norm": 0.4161796743739333,
+      "learning_rate": 6.357646456734574e-06,
+      "loss": 0.1986,
+      "step": 4233
+    },
+    {
+      "epoch": 1.4139255301385873,
+      "grad_norm": 0.412845390363555,
+      "learning_rate": 6.3557760316029115e-06,
+      "loss": 0.192,
+      "step": 4234
+    },
+    {
+      "epoch": 1.41425947570546,
+      "grad_norm": 0.3764465690310885,
+      "learning_rate": 6.353905401665317e-06,
+      "loss": 0.1832,
+      "step": 4235
+    },
+    {
+      "epoch": 1.4145934212723326,
+      "grad_norm": 0.3984816244207758,
+      "learning_rate": 6.35203456720437e-06,
+      "loss": 0.1959,
+      "step": 4236
+    },
+    {
+      "epoch": 1.4149273668392053,
+      "grad_norm": 0.4660411346747864,
+      "learning_rate": 6.35016352850268e-06,
+      "loss": 0.208,
+      "step": 4237
+    },
+    {
+      "epoch": 1.4152613124060778,
+      "grad_norm": 0.4208257324589308,
+      "learning_rate": 6.3482922858428915e-06,
+      "loss": 0.1906,
+      "step": 4238
+    },
+    {
+      "epoch": 1.4155952579729503,
+      "grad_norm": 0.4093879464060409,
+      "learning_rate": 6.34642083950768e-06,
+      "loss": 0.1919,
+      "step": 4239
+    },
+    {
+      "epoch": 1.415929203539823,
+      "grad_norm": 0.4169795445243217,
+      "learning_rate": 6.344549189779745e-06,
+      "loss": 0.1934,
+      "step": 4240
+    },
+    {
+      "epoch": 1.4162631491066957,
+      "grad_norm": 0.3999232718854513,
+      "learning_rate": 6.342677336941825e-06,
+      "loss": 0.201,
+      "step": 4241
+    },
+    {
+      "epoch": 1.4165970946735682,
+      "grad_norm": 0.43188532288830084,
+      "learning_rate": 6.340805281276683e-06,
+      "loss": 0.1915,
+      "step": 4242
+    },
+    {
+      "epoch": 1.4169310402404407,
+      "grad_norm": 0.4212897378007842,
+      "learning_rate": 6.338933023067114e-06,
+      "loss": 0.2031,
+      "step": 4243
+    },
+    {
+      "epoch": 1.4172649858073134,
+      "grad_norm": 0.39104495072727397,
+      "learning_rate": 6.337060562595949e-06,
+      "loss": 0.1838,
+      "step": 4244
+    },
+    {
+      "epoch": 1.4175989313741861,
+      "grad_norm": 0.4279064144901453,
+      "learning_rate": 6.3351879001460425e-06,
+      "loss": 0.203,
+      "step": 4245
+    },
+    {
+      "epoch": 1.4179328769410586,
+      "grad_norm": 0.44299570054290127,
+      "learning_rate": 6.333315036000281e-06,
+      "loss": 0.1909,
+      "step": 4246
+    },
+    {
+      "epoch": 1.4182668225079311,
+      "grad_norm": 0.4518229529900438,
+      "learning_rate": 6.331441970441585e-06,
+      "loss": 0.2016,
+      "step": 4247
+    },
+    {
+      "epoch": 1.4186007680748038,
+      "grad_norm": 0.437754191582659,
+      "learning_rate": 6.329568703752902e-06,
+      "loss": 0.2023,
+      "step": 4248
+    },
+    {
+      "epoch": 1.4189347136416763,
+      "grad_norm": 0.39214812360430923,
+      "learning_rate": 6.32769523621721e-06,
+      "loss": 0.1946,
+      "step": 4249
+    },
+    {
+      "epoch": 1.419268659208549,
+      "grad_norm": 0.40032628961599576,
+      "learning_rate": 6.3258215681175215e-06,
+      "loss": 0.1893,
+      "step": 4250
+    },
+    {
+      "epoch": 1.4196026047754216,
+      "grad_norm": 0.4135834268436745,
+      "learning_rate": 6.323947699736873e-06,
+      "loss": 0.1914,
+      "step": 4251
+    },
+    {
+      "epoch": 1.4199365503422943,
+      "grad_norm": 0.4408452075370049,
+      "learning_rate": 6.3220736313583345e-06,
+      "loss": 0.2073,
+      "step": 4252
+    },
+    {
+      "epoch": 1.4202704959091668,
+      "grad_norm": 0.46529501920804817,
+      "learning_rate": 6.320199363265008e-06,
+      "loss": 0.1994,
+      "step": 4253
+    },
+    {
+      "epoch": 1.4206044414760395,
+      "grad_norm": 0.4180415628920196,
+      "learning_rate": 6.318324895740023e-06,
+      "loss": 0.1979,
+      "step": 4254
+    },
+    {
+      "epoch": 1.420938387042912,
+      "grad_norm": 0.40285815975229106,
+      "learning_rate": 6.31645022906654e-06,
+      "loss": 0.1972,
+      "step": 4255
+    },
+    {
+      "epoch": 1.4212723326097847,
+      "grad_norm": 0.4245223489154361,
+      "learning_rate": 6.314575363527748e-06,
+      "loss": 0.1955,
+      "step": 4256
+    },
+    {
+      "epoch": 1.4216062781766572,
+      "grad_norm": 0.4107284657056354,
+      "learning_rate": 6.312700299406871e-06,
+      "loss": 0.1956,
+      "step": 4257
+    },
+    {
+      "epoch": 1.4219402237435297,
+      "grad_norm": 0.4123475914189533,
+      "learning_rate": 6.310825036987154e-06,
+      "loss": 0.1987,
+      "step": 4258
+    },
+    {
+      "epoch": 1.4222741693104024,
+      "grad_norm": 0.4113685742620576,
+      "learning_rate": 6.308949576551884e-06,
+      "loss": 0.189,
+      "step": 4259
+    },
+    {
+      "epoch": 1.4226081148772751,
+      "grad_norm": 0.4113493190674343,
+      "learning_rate": 6.3070739183843655e-06,
+      "loss": 0.1958,
+      "step": 4260
+    },
+    {
+      "epoch": 1.4229420604441476,
+      "grad_norm": 0.41318754392232565,
+      "learning_rate": 6.305198062767942e-06,
+      "loss": 0.1921,
+      "step": 4261
+    },
+    {
+      "epoch": 1.4232760060110201,
+      "grad_norm": 0.42444570405075827,
+      "learning_rate": 6.303322009985984e-06,
+      "loss": 0.2002,
+      "step": 4262
+    },
+    {
+      "epoch": 1.4236099515778928,
+      "grad_norm": 0.4036884323591814,
+      "learning_rate": 6.301445760321889e-06,
+      "loss": 0.1921,
+      "step": 4263
+    },
+    {
+      "epoch": 1.4239438971447653,
+      "grad_norm": 0.3868929082230903,
+      "learning_rate": 6.299569314059088e-06,
+      "loss": 0.1952,
+      "step": 4264
+    },
+    {
+      "epoch": 1.424277842711638,
+      "grad_norm": 0.3934660880793526,
+      "learning_rate": 6.297692671481042e-06,
+      "loss": 0.1961,
+      "step": 4265
+    },
+    {
+      "epoch": 1.4246117882785105,
+      "grad_norm": 0.5476517438338885,
+      "learning_rate": 6.295815832871235e-06,
+      "loss": 0.178,
+      "step": 4266
+    },
+    {
+      "epoch": 1.4249457338453833,
+      "grad_norm": 0.42403647877836415,
+      "learning_rate": 6.2939387985131905e-06,
+      "loss": 0.1926,
+      "step": 4267
+    },
+    {
+      "epoch": 1.4252796794122558,
+      "grad_norm": 0.4603384594375843,
+      "learning_rate": 6.292061568690455e-06,
+      "loss": 0.1962,
+      "step": 4268
+    },
+    {
+      "epoch": 1.4256136249791285,
+      "grad_norm": 0.42212567893770986,
+      "learning_rate": 6.290184143686606e-06,
+      "loss": 0.1825,
+      "step": 4269
+    },
+    {
+      "epoch": 1.425947570546001,
+      "grad_norm": 0.42720372519788125,
+      "learning_rate": 6.288306523785252e-06,
+      "loss": 0.2023,
+      "step": 4270
+    },
+    {
+      "epoch": 1.4262815161128737,
+      "grad_norm": 0.3850216851382417,
+      "learning_rate": 6.286428709270026e-06,
+      "loss": 0.19,
+      "step": 4271
+    },
+    {
+      "epoch": 1.4266154616797462,
+      "grad_norm": 0.42293406684487284,
+      "learning_rate": 6.284550700424597e-06,
+      "loss": 0.1934,
+      "step": 4272
+    },
+    {
+      "epoch": 1.4269494072466187,
+      "grad_norm": 0.4185602665867109,
+      "learning_rate": 6.282672497532659e-06,
+      "loss": 0.1862,
+      "step": 4273
+    },
+    {
+      "epoch": 1.4272833528134914,
+      "grad_norm": 0.42353635443489346,
+      "learning_rate": 6.280794100877938e-06,
+      "loss": 0.1976,
+      "step": 4274
+    },
+    {
+      "epoch": 1.427617298380364,
+      "grad_norm": 0.4193474465872973,
+      "learning_rate": 6.278915510744187e-06,
+      "loss": 0.1798,
+      "step": 4275
+    },
+    {
+      "epoch": 1.4279512439472366,
+      "grad_norm": 0.47731951370814735,
+      "learning_rate": 6.277036727415189e-06,
+      "loss": 0.2027,
+      "step": 4276
+    },
+    {
+      "epoch": 1.428285189514109,
+      "grad_norm": 0.41253445818583245,
+      "learning_rate": 6.2751577511747575e-06,
+      "loss": 0.1864,
+      "step": 4277
+    },
+    {
+      "epoch": 1.4286191350809818,
+      "grad_norm": 0.41072543870509975,
+      "learning_rate": 6.273278582306732e-06,
+      "loss": 0.2005,
+      "step": 4278
+    },
+    {
+      "epoch": 1.4289530806478545,
+      "grad_norm": 0.4464246228500059,
+      "learning_rate": 6.271399221094986e-06,
+      "loss": 0.1992,
+      "step": 4279
+    },
+    {
+      "epoch": 1.429287026214727,
+      "grad_norm": 0.4094540550192012,
+      "learning_rate": 6.269519667823416e-06,
+      "loss": 0.1973,
+      "step": 4280
+    },
+    {
+      "epoch": 1.4296209717815995,
+      "grad_norm": 0.41233321137674056,
+      "learning_rate": 6.267639922775952e-06,
+      "loss": 0.1979,
+      "step": 4281
+    },
+    {
+      "epoch": 1.4299549173484722,
+      "grad_norm": 0.424743809144049,
+      "learning_rate": 6.265759986236552e-06,
+      "loss": 0.2035,
+      "step": 4282
+    },
+    {
+      "epoch": 1.4302888629153447,
+      "grad_norm": 0.4005345858496865,
+      "learning_rate": 6.263879858489204e-06,
+      "loss": 0.1913,
+      "step": 4283
+    },
+    {
+      "epoch": 1.4306228084822175,
+      "grad_norm": 0.4053273447527923,
+      "learning_rate": 6.261999539817919e-06,
+      "loss": 0.1943,
+      "step": 4284
+    },
+    {
+      "epoch": 1.43095675404909,
+      "grad_norm": 0.4185492417076911,
+      "learning_rate": 6.260119030506746e-06,
+      "loss": 0.1975,
+      "step": 4285
+    },
+    {
+      "epoch": 1.4312906996159627,
+      "grad_norm": 0.3913801060486488,
+      "learning_rate": 6.258238330839754e-06,
+      "loss": 0.1798,
+      "step": 4286
+    },
+    {
+      "epoch": 1.4316246451828352,
+      "grad_norm": 0.426769833384345,
+      "learning_rate": 6.2563574411010485e-06,
+      "loss": 0.1964,
+      "step": 4287
+    },
+    {
+      "epoch": 1.4319585907497077,
+      "grad_norm": 0.41614648545349253,
+      "learning_rate": 6.254476361574757e-06,
+      "loss": 0.189,
+      "step": 4288
+    },
+    {
+      "epoch": 1.4322925363165804,
+      "grad_norm": 0.4241398703694006,
+      "learning_rate": 6.252595092545042e-06,
+      "loss": 0.1875,
+      "step": 4289
+    },
+    {
+      "epoch": 1.432626481883453,
+      "grad_norm": 0.4299458304622268,
+      "learning_rate": 6.250713634296087e-06,
+      "loss": 0.2018,
+      "step": 4290
+    },
+    {
+      "epoch": 1.4329604274503256,
+      "grad_norm": 0.4229363795139313,
+      "learning_rate": 6.248831987112113e-06,
+      "loss": 0.206,
+      "step": 4291
+    },
+    {
+      "epoch": 1.433294373017198,
+      "grad_norm": 0.401583294371539,
+      "learning_rate": 6.246950151277362e-06,
+      "loss": 0.1883,
+      "step": 4292
+    },
+    {
+      "epoch": 1.4336283185840708,
+      "grad_norm": 0.4519792653966738,
+      "learning_rate": 6.245068127076109e-06,
+      "loss": 0.2018,
+      "step": 4293
+    },
+    {
+      "epoch": 1.4339622641509435,
+      "grad_norm": 0.40227774137045846,
+      "learning_rate": 6.243185914792655e-06,
+      "loss": 0.1819,
+      "step": 4294
+    },
+    {
+      "epoch": 1.434296209717816,
+      "grad_norm": 0.559657939335022,
+      "learning_rate": 6.2413035147113295e-06,
+      "loss": 0.1913,
+      "step": 4295
+    },
+    {
+      "epoch": 1.4346301552846885,
+      "grad_norm": 0.3972251507876515,
+      "learning_rate": 6.239420927116493e-06,
+      "loss": 0.1886,
+      "step": 4296
+    },
+    {
+      "epoch": 1.4349641008515612,
+      "grad_norm": 0.41776370884070535,
+      "learning_rate": 6.2375381522925325e-06,
+      "loss": 0.1972,
+      "step": 4297
+    },
+    {
+      "epoch": 1.4352980464184337,
+      "grad_norm": 0.41015662981383944,
+      "learning_rate": 6.235655190523862e-06,
+      "loss": 0.1978,
+      "step": 4298
+    },
+    {
+      "epoch": 1.4356319919853064,
+      "grad_norm": 0.4229324755634245,
+      "learning_rate": 6.233772042094924e-06,
+      "loss": 0.1968,
+      "step": 4299
+    },
+    {
+      "epoch": 1.435965937552179,
+      "grad_norm": 0.4710852821204468,
+      "learning_rate": 6.231888707290194e-06,
+      "loss": 0.2016,
+      "step": 4300
+    },
+    {
+      "epoch": 1.4362998831190517,
+      "grad_norm": 0.5534094086271172,
+      "learning_rate": 6.230005186394169e-06,
+      "loss": 0.2034,
+      "step": 4301
+    },
+    {
+      "epoch": 1.4366338286859242,
+      "grad_norm": 0.40362990371947977,
+      "learning_rate": 6.228121479691377e-06,
+      "loss": 0.1976,
+      "step": 4302
+    },
+    {
+      "epoch": 1.4369677742527969,
+      "grad_norm": 0.46882248799019716,
+      "learning_rate": 6.226237587466375e-06,
+      "loss": 0.216,
+      "step": 4303
+    },
+    {
+      "epoch": 1.4373017198196694,
+      "grad_norm": 0.44983410588953954,
+      "learning_rate": 6.224353510003747e-06,
+      "loss": 0.1955,
+      "step": 4304
+    },
+    {
+      "epoch": 1.437635665386542,
+      "grad_norm": 0.4562795509116905,
+      "learning_rate": 6.222469247588105e-06,
+      "loss": 0.1959,
+      "step": 4305
+    },
+    {
+      "epoch": 1.4379696109534146,
+      "grad_norm": 0.40086144043627886,
+      "learning_rate": 6.220584800504091e-06,
+      "loss": 0.1897,
+      "step": 4306
+    },
+    {
+      "epoch": 1.438303556520287,
+      "grad_norm": 0.4220200763976442,
+      "learning_rate": 6.218700169036368e-06,
+      "loss": 0.2092,
+      "step": 4307
+    },
+    {
+      "epoch": 1.4386375020871598,
+      "grad_norm": 0.40218684127289284,
+      "learning_rate": 6.216815353469636e-06,
+      "loss": 0.1891,
+      "step": 4308
+    },
+    {
+      "epoch": 1.4389714476540325,
+      "grad_norm": 0.3699554890094108,
+      "learning_rate": 6.214930354088618e-06,
+      "loss": 0.1804,
+      "step": 4309
+    },
+    {
+      "epoch": 1.439305393220905,
+      "grad_norm": 0.46604002782555415,
+      "learning_rate": 6.213045171178063e-06,
+      "loss": 0.2052,
+      "step": 4310
+    },
+    {
+      "epoch": 1.4396393387877775,
+      "grad_norm": 0.4096947218921958,
+      "learning_rate": 6.2111598050227535e-06,
+      "loss": 0.1879,
+      "step": 4311
+    },
+    {
+      "epoch": 1.4399732843546502,
+      "grad_norm": 0.43814289672198536,
+      "learning_rate": 6.209274255907494e-06,
+      "loss": 0.2108,
+      "step": 4312
+    },
+    {
+      "epoch": 1.4403072299215227,
+      "grad_norm": 0.4349153906107477,
+      "learning_rate": 6.207388524117119e-06,
+      "loss": 0.1942,
+      "step": 4313
+    },
+    {
+      "epoch": 1.4406411754883954,
+      "grad_norm": 0.389711559079184,
+      "learning_rate": 6.205502609936491e-06,
+      "loss": 0.189,
+      "step": 4314
+    },
+    {
+      "epoch": 1.440975121055268,
+      "grad_norm": 0.39940637696014164,
+      "learning_rate": 6.2036165136505e-06,
+      "loss": 0.1912,
+      "step": 4315
+    },
+    {
+      "epoch": 1.4413090666221406,
+      "grad_norm": 0.4348476443490899,
+      "learning_rate": 6.201730235544062e-06,
+      "loss": 0.2026,
+      "step": 4316
+    },
+    {
+      "epoch": 1.4416430121890131,
+      "grad_norm": 0.4010344004022616,
+      "learning_rate": 6.1998437759021235e-06,
+      "loss": 0.19,
+      "step": 4317
+    },
+    {
+      "epoch": 1.4419769577558859,
+      "grad_norm": 0.4077069086603506,
+      "learning_rate": 6.197957135009653e-06,
+      "loss": 0.1992,
+      "step": 4318
+    },
+    {
+      "epoch": 1.4423109033227584,
+      "grad_norm": 0.454450027580997,
+      "learning_rate": 6.196070313151652e-06,
+      "loss": 0.2078,
+      "step": 4319
+    },
+    {
+      "epoch": 1.442644848889631,
+      "grad_norm": 0.42090639548690023,
+      "learning_rate": 6.194183310613147e-06,
+      "loss": 0.2042,
+      "step": 4320
+    },
+    {
+      "epoch": 1.4429787944565036,
+      "grad_norm": 0.3924425533047438,
+      "learning_rate": 6.1922961276791925e-06,
+      "loss": 0.1861,
+      "step": 4321
+    },
+    {
+      "epoch": 1.443312740023376,
+      "grad_norm": 0.43825434862695306,
+      "learning_rate": 6.190408764634869e-06,
+      "loss": 0.2036,
+      "step": 4322
+    },
+    {
+      "epoch": 1.4436466855902488,
+      "grad_norm": 0.4554298024237673,
+      "learning_rate": 6.188521221765285e-06,
+      "loss": 0.2075,
+      "step": 4323
+    },
+    {
+      "epoch": 1.4439806311571215,
+      "grad_norm": 0.40310377491021876,
+      "learning_rate": 6.186633499355576e-06,
+      "loss": 0.1849,
+      "step": 4324
+    },
+    {
+      "epoch": 1.444314576723994,
+      "grad_norm": 0.3958077216570012,
+      "learning_rate": 6.184745597690903e-06,
+      "loss": 0.2019,
+      "step": 4325
+    },
+    {
+      "epoch": 1.4446485222908665,
+      "grad_norm": 0.41846403968531265,
+      "learning_rate": 6.1828575170564595e-06,
+      "loss": 0.1921,
+      "step": 4326
+    },
+    {
+      "epoch": 1.4449824678577392,
+      "grad_norm": 0.4349377636186543,
+      "learning_rate": 6.18096925773746e-06,
+      "loss": 0.1972,
+      "step": 4327
+    },
+    {
+      "epoch": 1.445316413424612,
+      "grad_norm": 0.39603752468722014,
+      "learning_rate": 6.179080820019147e-06,
+      "loss": 0.1766,
+      "step": 4328
+    },
+    {
+      "epoch": 1.4456503589914844,
+      "grad_norm": 0.40669232085821777,
+      "learning_rate": 6.177192204186796e-06,
+      "loss": 0.1931,
+      "step": 4329
+    },
+    {
+      "epoch": 1.445984304558357,
+      "grad_norm": 0.3946887338662643,
+      "learning_rate": 6.1753034105257e-06,
+      "loss": 0.1957,
+      "step": 4330
+    },
+    {
+      "epoch": 1.4463182501252296,
+      "grad_norm": 0.41443060881225996,
+      "learning_rate": 6.173414439321185e-06,
+      "loss": 0.1964,
+      "step": 4331
+    },
+    {
+      "epoch": 1.4466521956921021,
+      "grad_norm": 0.4405029054870975,
+      "learning_rate": 6.171525290858602e-06,
+      "loss": 0.1927,
+      "step": 4332
+    },
+    {
+      "epoch": 1.4469861412589748,
+      "grad_norm": 0.4628203122774109,
+      "learning_rate": 6.169635965423331e-06,
+      "loss": 0.211,
+      "step": 4333
+    },
+    {
+      "epoch": 1.4473200868258473,
+      "grad_norm": 0.48872692289281494,
+      "learning_rate": 6.167746463300774e-06,
+      "loss": 0.2088,
+      "step": 4334
+    },
+    {
+      "epoch": 1.44765403239272,
+      "grad_norm": 0.4343259836847498,
+      "learning_rate": 6.1658567847763655e-06,
+      "loss": 0.2007,
+      "step": 4335
+    },
+    {
+      "epoch": 1.4479879779595926,
+      "grad_norm": 0.4398765922609618,
+      "learning_rate": 6.163966930135561e-06,
+      "loss": 0.2006,
+      "step": 4336
+    },
+    {
+      "epoch": 1.448321923526465,
+      "grad_norm": 0.45579456571854265,
+      "learning_rate": 6.162076899663846e-06,
+      "loss": 0.2107,
+      "step": 4337
+    },
+    {
+      "epoch": 1.4486558690933378,
+      "grad_norm": 0.40883179410081466,
+      "learning_rate": 6.160186693646732e-06,
+      "loss": 0.1849,
+      "step": 4338
+    },
+    {
+      "epoch": 1.4489898146602105,
+      "grad_norm": 0.49071716904008567,
+      "learning_rate": 6.158296312369759e-06,
+      "loss": 0.2147,
+      "step": 4339
+    },
+    {
+      "epoch": 1.449323760227083,
+      "grad_norm": 0.4314291467098978,
+      "learning_rate": 6.156405756118489e-06,
+      "loss": 0.2086,
+      "step": 4340
+    },
+    {
+      "epoch": 1.4496577057939555,
+      "grad_norm": 0.4593921097879424,
+      "learning_rate": 6.154515025178511e-06,
+      "loss": 0.2002,
+      "step": 4341
+    },
+    {
+      "epoch": 1.4499916513608282,
+      "grad_norm": 0.45040771494821547,
+      "learning_rate": 6.152624119835447e-06,
+      "loss": 0.2086,
+      "step": 4342
+    },
+    {
+      "epoch": 1.450325596927701,
+      "grad_norm": 0.4299923678581887,
+      "learning_rate": 6.150733040374937e-06,
+      "loss": 0.2014,
+      "step": 4343
+    },
+    {
+      "epoch": 1.4506595424945734,
+      "grad_norm": 0.400344990906151,
+      "learning_rate": 6.148841787082653e-06,
+      "loss": 0.1927,
+      "step": 4344
+    },
+    {
+      "epoch": 1.450993488061446,
+      "grad_norm": 0.5435972870488641,
+      "learning_rate": 6.146950360244288e-06,
+      "loss": 0.1967,
+      "step": 4345
+    },
+    {
+      "epoch": 1.4513274336283186,
+      "grad_norm": 0.5742306297652585,
+      "learning_rate": 6.145058760145568e-06,
+      "loss": 0.2234,
+      "step": 4346
+    },
+    {
+      "epoch": 1.4516613791951911,
+      "grad_norm": 0.39381531571315703,
+      "learning_rate": 6.14316698707224e-06,
+      "loss": 0.188,
+      "step": 4347
+    },
+    {
+      "epoch": 1.4519953247620638,
+      "grad_norm": 0.39767928183404444,
+      "learning_rate": 6.1412750413100754e-06,
+      "loss": 0.1967,
+      "step": 4348
+    },
+    {
+      "epoch": 1.4523292703289363,
+      "grad_norm": 0.42473106735652594,
+      "learning_rate": 6.13938292314488e-06,
+      "loss": 0.2017,
+      "step": 4349
+    },
+    {
+      "epoch": 1.452663215895809,
+      "grad_norm": 0.4608850855222034,
+      "learning_rate": 6.137490632862479e-06,
+      "loss": 0.2073,
+      "step": 4350
+    },
+    {
+      "epoch": 1.4529971614626815,
+      "grad_norm": 0.44194784147975535,
+      "learning_rate": 6.135598170748721e-06,
+      "loss": 0.1999,
+      "step": 4351
+    },
+    {
+      "epoch": 1.4533311070295543,
+      "grad_norm": 0.4336212713439643,
+      "learning_rate": 6.13370553708949e-06,
+      "loss": 0.21,
+      "step": 4352
+    },
+    {
+      "epoch": 1.4536650525964268,
+      "grad_norm": 0.4474198500961851,
+      "learning_rate": 6.13181273217069e-06,
+      "loss": 0.2107,
+      "step": 4353
+    },
+    {
+      "epoch": 1.4539989981632995,
+      "grad_norm": 0.3988946635025851,
+      "learning_rate": 6.129919756278248e-06,
+      "loss": 0.1998,
+      "step": 4354
+    },
+    {
+      "epoch": 1.454332943730172,
+      "grad_norm": 0.4289910244729747,
+      "learning_rate": 6.128026609698124e-06,
+      "loss": 0.1926,
+      "step": 4355
+    },
+    {
+      "epoch": 1.4546668892970445,
+      "grad_norm": 0.38926894257927563,
+      "learning_rate": 6.126133292716297e-06,
+      "loss": 0.1893,
+      "step": 4356
+    },
+    {
+      "epoch": 1.4550008348639172,
+      "grad_norm": 0.42221784850848676,
+      "learning_rate": 6.124239805618778e-06,
+      "loss": 0.1908,
+      "step": 4357
+    },
+    {
+      "epoch": 1.45533478043079,
+      "grad_norm": 0.39646952858495677,
+      "learning_rate": 6.122346148691598e-06,
+      "loss": 0.1877,
+      "step": 4358
+    },
+    {
+      "epoch": 1.4556687259976624,
+      "grad_norm": 0.40546148481183963,
+      "learning_rate": 6.120452322220818e-06,
+      "loss": 0.2004,
+      "step": 4359
+    },
+    {
+      "epoch": 1.456002671564535,
+      "grad_norm": 0.4547099784865291,
+      "learning_rate": 6.11855832649252e-06,
+      "loss": 0.2115,
+      "step": 4360
+    },
+    {
+      "epoch": 1.4563366171314076,
+      "grad_norm": 0.38112993327587635,
+      "learning_rate": 6.116664161792817e-06,
+      "loss": 0.1858,
+      "step": 4361
+    },
+    {
+      "epoch": 1.45667056269828,
+      "grad_norm": 0.4019263289774499,
+      "learning_rate": 6.114769828407845e-06,
+      "loss": 0.1955,
+      "step": 4362
+    },
+    {
+      "epoch": 1.4570045082651528,
+      "grad_norm": 0.41009801163114296,
+      "learning_rate": 6.112875326623763e-06,
+      "loss": 0.1882,
+      "step": 4363
+    },
+    {
+      "epoch": 1.4573384538320253,
+      "grad_norm": 0.4079447248544796,
+      "learning_rate": 6.110980656726759e-06,
+      "loss": 0.1889,
+      "step": 4364
+    },
+    {
+      "epoch": 1.457672399398898,
+      "grad_norm": 0.4516607016893233,
+      "learning_rate": 6.109085819003048e-06,
+      "loss": 0.2073,
+      "step": 4365
+    },
+    {
+      "epoch": 1.4580063449657705,
+      "grad_norm": 0.42566248368763016,
+      "learning_rate": 6.107190813738864e-06,
+      "loss": 0.2089,
+      "step": 4366
+    },
+    {
+      "epoch": 1.4583402905326432,
+      "grad_norm": 0.39147915382309184,
+      "learning_rate": 6.10529564122047e-06,
+      "loss": 0.1974,
+      "step": 4367
+    },
+    {
+      "epoch": 1.4586742360995157,
+      "grad_norm": 0.3982765548370061,
+      "learning_rate": 6.103400301734155e-06,
+      "loss": 0.1993,
+      "step": 4368
+    },
+    {
+      "epoch": 1.4590081816663885,
+      "grad_norm": 0.40681227759099087,
+      "learning_rate": 6.101504795566232e-06,
+      "loss": 0.2013,
+      "step": 4369
+    },
+    {
+      "epoch": 1.459342127233261,
+      "grad_norm": 0.44284559949396546,
+      "learning_rate": 6.099609123003041e-06,
+      "loss": 0.2055,
+      "step": 4370
+    },
+    {
+      "epoch": 1.4596760728001335,
+      "grad_norm": 0.3860492351094961,
+      "learning_rate": 6.097713284330944e-06,
+      "loss": 0.1948,
+      "step": 4371
+    },
+    {
+      "epoch": 1.4600100183670062,
+      "grad_norm": 0.41425735018239185,
+      "learning_rate": 6.095817279836329e-06,
+      "loss": 0.1912,
+      "step": 4372
+    },
+    {
+      "epoch": 1.4603439639338789,
+      "grad_norm": 0.38871439359382953,
+      "learning_rate": 6.093921109805612e-06,
+      "loss": 0.1928,
+      "step": 4373
+    },
+    {
+      "epoch": 1.4606779095007514,
+      "grad_norm": 0.4143118345237507,
+      "learning_rate": 6.092024774525231e-06,
+      "loss": 0.215,
+      "step": 4374
+    },
+    {
+      "epoch": 1.4610118550676239,
+      "grad_norm": 0.44030852410849036,
+      "learning_rate": 6.090128274281649e-06,
+      "loss": 0.2048,
+      "step": 4375
+    },
+    {
+      "epoch": 1.4613458006344966,
+      "grad_norm": 0.4075810714964566,
+      "learning_rate": 6.0882316093613555e-06,
+      "loss": 0.191,
+      "step": 4376
+    },
+    {
+      "epoch": 1.4616797462013693,
+      "grad_norm": 0.47178283797989656,
+      "learning_rate": 6.086334780050865e-06,
+      "loss": 0.2076,
+      "step": 4377
+    },
+    {
+      "epoch": 1.4620136917682418,
+      "grad_norm": 0.4385058940753501,
+      "learning_rate": 6.084437786636713e-06,
+      "loss": 0.1973,
+      "step": 4378
+    },
+    {
+      "epoch": 1.4623476373351143,
+      "grad_norm": 0.39454270917277967,
+      "learning_rate": 6.082540629405467e-06,
+      "loss": 0.1956,
+      "step": 4379
+    },
+    {
+      "epoch": 1.462681582901987,
+      "grad_norm": 0.4246463460236804,
+      "learning_rate": 6.08064330864371e-06,
+      "loss": 0.1897,
+      "step": 4380
+    },
+    {
+      "epoch": 1.4630155284688595,
+      "grad_norm": 0.4279872130659886,
+      "learning_rate": 6.078745824638058e-06,
+      "loss": 0.1909,
+      "step": 4381
+    },
+    {
+      "epoch": 1.4633494740357322,
+      "grad_norm": 0.4037107620637563,
+      "learning_rate": 6.076848177675148e-06,
+      "loss": 0.185,
+      "step": 4382
+    },
+    {
+      "epoch": 1.4636834196026047,
+      "grad_norm": 0.452895373409483,
+      "learning_rate": 6.07495036804164e-06,
+      "loss": 0.1987,
+      "step": 4383
+    },
+    {
+      "epoch": 1.4640173651694774,
+      "grad_norm": 0.46645780502615036,
+      "learning_rate": 6.073052396024222e-06,
+      "loss": 0.2091,
+      "step": 4384
+    },
+    {
+      "epoch": 1.46435131073635,
+      "grad_norm": 0.4041982376337457,
+      "learning_rate": 6.071154261909605e-06,
+      "loss": 0.1929,
+      "step": 4385
+    },
+    {
+      "epoch": 1.4646852563032224,
+      "grad_norm": 0.4356675688935076,
+      "learning_rate": 6.069255965984524e-06,
+      "loss": 0.2031,
+      "step": 4386
+    },
+    {
+      "epoch": 1.4650192018700952,
+      "grad_norm": 0.40081538939786715,
+      "learning_rate": 6.067357508535741e-06,
+      "loss": 0.1979,
+      "step": 4387
+    },
+    {
+      "epoch": 1.4653531474369679,
+      "grad_norm": 0.424597670187965,
+      "learning_rate": 6.065458889850037e-06,
+      "loss": 0.1921,
+      "step": 4388
+    },
+    {
+      "epoch": 1.4656870930038404,
+      "grad_norm": 0.4502702104798383,
+      "learning_rate": 6.063560110214224e-06,
+      "loss": 0.2006,
+      "step": 4389
+    },
+    {
+      "epoch": 1.4660210385707129,
+      "grad_norm": 0.4195663771268208,
+      "learning_rate": 6.061661169915132e-06,
+      "loss": 0.1984,
+      "step": 4390
+    },
+    {
+      "epoch": 1.4663549841375856,
+      "grad_norm": 0.4308479391309738,
+      "learning_rate": 6.05976206923962e-06,
+      "loss": 0.2084,
+      "step": 4391
+    },
+    {
+      "epoch": 1.4666889297044583,
+      "grad_norm": 0.4166693946797753,
+      "learning_rate": 6.057862808474569e-06,
+      "loss": 0.1985,
+      "step": 4392
+    },
+    {
+      "epoch": 1.4670228752713308,
+      "grad_norm": 0.4603888013323258,
+      "learning_rate": 6.055963387906884e-06,
+      "loss": 0.2015,
+      "step": 4393
+    },
+    {
+      "epoch": 1.4673568208382033,
+      "grad_norm": 0.4081769473587065,
+      "learning_rate": 6.054063807823497e-06,
+      "loss": 0.2027,
+      "step": 4394
+    },
+    {
+      "epoch": 1.467690766405076,
+      "grad_norm": 0.4134409770785201,
+      "learning_rate": 6.052164068511359e-06,
+      "loss": 0.1994,
+      "step": 4395
+    },
+    {
+      "epoch": 1.4680247119719485,
+      "grad_norm": 0.4692504389911519,
+      "learning_rate": 6.05026417025745e-06,
+      "loss": 0.2123,
+      "step": 4396
+    },
+    {
+      "epoch": 1.4683586575388212,
+      "grad_norm": 0.4493204264781914,
+      "learning_rate": 6.0483641133487736e-06,
+      "loss": 0.2065,
+      "step": 4397
+    },
+    {
+      "epoch": 1.4686926031056937,
+      "grad_norm": 0.42766406223034875,
+      "learning_rate": 6.046463898072351e-06,
+      "loss": 0.2048,
+      "step": 4398
+    },
+    {
+      "epoch": 1.4690265486725664,
+      "grad_norm": 0.5308801017267736,
+      "learning_rate": 6.044563524715237e-06,
+      "loss": 0.1971,
+      "step": 4399
+    },
+    {
+      "epoch": 1.469360494239439,
+      "grad_norm": 0.4262125710912439,
+      "learning_rate": 6.042662993564503e-06,
+      "loss": 0.1949,
+      "step": 4400
+    },
+    {
+      "epoch": 1.4696944398063116,
+      "grad_norm": 0.4025672852690035,
+      "learning_rate": 6.040762304907246e-06,
+      "loss": 0.1984,
+      "step": 4401
+    },
+    {
+      "epoch": 1.4700283853731841,
+      "grad_norm": 0.4026158150281825,
+      "learning_rate": 6.038861459030588e-06,
+      "loss": 0.1891,
+      "step": 4402
+    },
+    {
+      "epoch": 1.4703623309400569,
+      "grad_norm": 0.3956628314819311,
+      "learning_rate": 6.036960456221677e-06,
+      "loss": 0.1958,
+      "step": 4403
+    },
+    {
+      "epoch": 1.4706962765069294,
+      "grad_norm": 0.4044262098140246,
+      "learning_rate": 6.035059296767676e-06,
+      "loss": 0.1916,
+      "step": 4404
+    },
+    {
+      "epoch": 1.4710302220738019,
+      "grad_norm": 0.480407738770424,
+      "learning_rate": 6.033157980955782e-06,
+      "loss": 0.2004,
+      "step": 4405
+    },
+    {
+      "epoch": 1.4713641676406746,
+      "grad_norm": 0.46733025871686174,
+      "learning_rate": 6.0312565090732115e-06,
+      "loss": 0.2135,
+      "step": 4406
+    },
+    {
+      "epoch": 1.4716981132075473,
+      "grad_norm": 0.38309985347010694,
+      "learning_rate": 6.0293548814072004e-06,
+      "loss": 0.1973,
+      "step": 4407
+    },
+    {
+      "epoch": 1.4720320587744198,
+      "grad_norm": 0.43882038617597807,
+      "learning_rate": 6.0274530982450155e-06,
+      "loss": 0.2097,
+      "step": 4408
+    },
+    {
+      "epoch": 1.4723660043412923,
+      "grad_norm": 0.4763463732615725,
+      "learning_rate": 6.025551159873941e-06,
+      "loss": 0.2025,
+      "step": 4409
+    },
+    {
+      "epoch": 1.472699949908165,
+      "grad_norm": 0.4773269336687056,
+      "learning_rate": 6.023649066581288e-06,
+      "loss": 0.2237,
+      "step": 4410
+    },
+    {
+      "epoch": 1.4730338954750375,
+      "grad_norm": 0.4148665052375862,
+      "learning_rate": 6.021746818654393e-06,
+      "loss": 0.2008,
+      "step": 4411
+    },
+    {
+      "epoch": 1.4733678410419102,
+      "grad_norm": 0.43151291731531444,
+      "learning_rate": 6.019844416380609e-06,
+      "loss": 0.1983,
+      "step": 4412
+    },
+    {
+      "epoch": 1.4737017866087827,
+      "grad_norm": 0.4286672462162659,
+      "learning_rate": 6.017941860047318e-06,
+      "loss": 0.1993,
+      "step": 4413
+    },
+    {
+      "epoch": 1.4740357321756554,
+      "grad_norm": 0.42593830099460744,
+      "learning_rate": 6.016039149941924e-06,
+      "loss": 0.2,
+      "step": 4414
+    },
+    {
+      "epoch": 1.474369677742528,
+      "grad_norm": 0.44556539133547163,
+      "learning_rate": 6.01413628635185e-06,
+      "loss": 0.2224,
+      "step": 4415
+    },
+    {
+      "epoch": 1.4747036233094006,
+      "grad_norm": 0.40967464330179093,
+      "learning_rate": 6.012233269564551e-06,
+      "loss": 0.1948,
+      "step": 4416
+    },
+    {
+      "epoch": 1.4750375688762731,
+      "grad_norm": 0.43684591455417715,
+      "learning_rate": 6.010330099867497e-06,
+      "loss": 0.2075,
+      "step": 4417
+    },
+    {
+      "epoch": 1.4753715144431458,
+      "grad_norm": 0.4341408303231394,
+      "learning_rate": 6.008426777548186e-06,
+      "loss": 0.1982,
+      "step": 4418
+    },
+    {
+      "epoch": 1.4757054600100183,
+      "grad_norm": 0.4358707442878612,
+      "learning_rate": 6.0065233028941365e-06,
+      "loss": 0.2045,
+      "step": 4419
+    },
+    {
+      "epoch": 1.4760394055768908,
+      "grad_norm": 0.37328822493004127,
+      "learning_rate": 6.00461967619289e-06,
+      "loss": 0.1821,
+      "step": 4420
+    },
+    {
+      "epoch": 1.4763733511437636,
+      "grad_norm": 0.400014805788866,
+      "learning_rate": 6.002715897732013e-06,
+      "loss": 0.1847,
+      "step": 4421
+    },
+    {
+      "epoch": 1.4767072967106363,
+      "grad_norm": 0.4046246703398329,
+      "learning_rate": 6.000811967799092e-06,
+      "loss": 0.1899,
+      "step": 4422
+    },
+    {
+      "epoch": 1.4770412422775088,
+      "grad_norm": 0.40715797582521757,
+      "learning_rate": 5.99890788668174e-06,
+      "loss": 0.1941,
+      "step": 4423
+    },
+    {
+      "epoch": 1.4773751878443813,
+      "grad_norm": 0.4206959634130982,
+      "learning_rate": 5.997003654667589e-06,
+      "loss": 0.1991,
+      "step": 4424
+    },
+    {
+      "epoch": 1.477709133411254,
+      "grad_norm": 0.40769749828652163,
+      "learning_rate": 5.995099272044298e-06,
+      "loss": 0.1829,
+      "step": 4425
+    },
+    {
+      "epoch": 1.4780430789781267,
+      "grad_norm": 0.44404122589778716,
+      "learning_rate": 5.9931947390995435e-06,
+      "loss": 0.2092,
+      "step": 4426
+    },
+    {
+      "epoch": 1.4783770245449992,
+      "grad_norm": 0.42520813532123286,
+      "learning_rate": 5.99129005612103e-06,
+      "loss": 0.2081,
+      "step": 4427
+    },
+    {
+      "epoch": 1.4787109701118717,
+      "grad_norm": 0.4122469308763514,
+      "learning_rate": 5.989385223396482e-06,
+      "loss": 0.2011,
+      "step": 4428
+    },
+    {
+      "epoch": 1.4790449156787444,
+      "grad_norm": 0.3929382468133838,
+      "learning_rate": 5.987480241213646e-06,
+      "loss": 0.192,
+      "step": 4429
+    },
+    {
+      "epoch": 1.479378861245617,
+      "grad_norm": 0.5672122545325259,
+      "learning_rate": 5.985575109860292e-06,
+      "loss": 0.2116,
+      "step": 4430
+    },
+    {
+      "epoch": 1.4797128068124896,
+      "grad_norm": 0.4983677423174381,
+      "learning_rate": 5.983669829624214e-06,
+      "loss": 0.2018,
+      "step": 4431
+    },
+    {
+      "epoch": 1.4800467523793621,
+      "grad_norm": 0.42548655408845926,
+      "learning_rate": 5.981764400793224e-06,
+      "loss": 0.1971,
+      "step": 4432
+    },
+    {
+      "epoch": 1.4803806979462348,
+      "grad_norm": 0.39361510423237334,
+      "learning_rate": 5.9798588236551626e-06,
+      "loss": 0.1919,
+      "step": 4433
+    },
+    {
+      "epoch": 1.4807146435131073,
+      "grad_norm": 0.42721227153711655,
+      "learning_rate": 5.977953098497889e-06,
+      "loss": 0.2013,
+      "step": 4434
+    },
+    {
+      "epoch": 1.4810485890799798,
+      "grad_norm": 0.4093784539006786,
+      "learning_rate": 5.976047225609284e-06,
+      "loss": 0.1945,
+      "step": 4435
+    },
+    {
+      "epoch": 1.4813825346468525,
+      "grad_norm": 0.39877583576233977,
+      "learning_rate": 5.974141205277253e-06,
+      "loss": 0.1932,
+      "step": 4436
+    },
+    {
+      "epoch": 1.4817164802137253,
+      "grad_norm": 0.4241654493666607,
+      "learning_rate": 5.972235037789723e-06,
+      "loss": 0.1888,
+      "step": 4437
+    },
+    {
+      "epoch": 1.4820504257805978,
+      "grad_norm": 0.4127889957282086,
+      "learning_rate": 5.970328723434642e-06,
+      "loss": 0.1855,
+      "step": 4438
+    },
+    {
+      "epoch": 1.4823843713474703,
+      "grad_norm": 0.43974577195283593,
+      "learning_rate": 5.968422262499983e-06,
+      "loss": 0.2001,
+      "step": 4439
+    },
+    {
+      "epoch": 1.482718316914343,
+      "grad_norm": 0.43856634741945644,
+      "learning_rate": 5.966515655273739e-06,
+      "loss": 0.1909,
+      "step": 4440
+    },
+    {
+      "epoch": 1.4830522624812157,
+      "grad_norm": 0.43919243938383296,
+      "learning_rate": 5.9646089020439245e-06,
+      "loss": 0.1957,
+      "step": 4441
+    },
+    {
+      "epoch": 1.4833862080480882,
+      "grad_norm": 0.39521255006239037,
+      "learning_rate": 5.962702003098576e-06,
+      "loss": 0.1884,
+      "step": 4442
+    },
+    {
+      "epoch": 1.4837201536149607,
+      "grad_norm": 0.39010763406866694,
+      "learning_rate": 5.960794958725756e-06,
+      "loss": 0.1967,
+      "step": 4443
+    },
+    {
+      "epoch": 1.4840540991818334,
+      "grad_norm": 0.38308666558185406,
+      "learning_rate": 5.958887769213544e-06,
+      "loss": 0.1898,
+      "step": 4444
+    },
+    {
+      "epoch": 1.484388044748706,
+      "grad_norm": 0.3808395689544056,
+      "learning_rate": 5.956980434850044e-06,
+      "loss": 0.1926,
+      "step": 4445
+    },
+    {
+      "epoch": 1.4847219903155786,
+      "grad_norm": 0.4019120461912746,
+      "learning_rate": 5.955072955923381e-06,
+      "loss": 0.1901,
+      "step": 4446
+    },
+    {
+      "epoch": 1.485055935882451,
+      "grad_norm": 0.3914079755133911,
+      "learning_rate": 5.9531653327217035e-06,
+      "loss": 0.1858,
+      "step": 4447
+    },
+    {
+      "epoch": 1.4853898814493238,
+      "grad_norm": 0.4001292440001611,
+      "learning_rate": 5.951257565533177e-06,
+      "loss": 0.1942,
+      "step": 4448
+    },
+    {
+      "epoch": 1.4857238270161963,
+      "grad_norm": 0.44889699749599704,
+      "learning_rate": 5.949349654645997e-06,
+      "loss": 0.2144,
+      "step": 4449
+    },
+    {
+      "epoch": 1.486057772583069,
+      "grad_norm": 0.4376071738007426,
+      "learning_rate": 5.947441600348373e-06,
+      "loss": 0.2112,
+      "step": 4450
+    },
+    {
+      "epoch": 1.4863917181499415,
+      "grad_norm": 0.5002467028319895,
+      "learning_rate": 5.945533402928537e-06,
+      "loss": 0.2143,
+      "step": 4451
+    },
+    {
+      "epoch": 1.4867256637168142,
+      "grad_norm": 0.43261606233366906,
+      "learning_rate": 5.9436250626747505e-06,
+      "loss": 0.2231,
+      "step": 4452
+    },
+    {
+      "epoch": 1.4870596092836867,
+      "grad_norm": 0.42928106883312683,
+      "learning_rate": 5.941716579875286e-06,
+      "loss": 0.1944,
+      "step": 4453
+    },
+    {
+      "epoch": 1.4873935548505592,
+      "grad_norm": 0.44047447971144676,
+      "learning_rate": 5.939807954818443e-06,
+      "loss": 0.2087,
+      "step": 4454
+    },
+    {
+      "epoch": 1.487727500417432,
+      "grad_norm": 0.647500480455085,
+      "learning_rate": 5.937899187792544e-06,
+      "loss": 0.1897,
+      "step": 4455
+    },
+    {
+      "epoch": 1.4880614459843047,
+      "grad_norm": 0.39203680888737436,
+      "learning_rate": 5.935990279085928e-06,
+      "loss": 0.1926,
+      "step": 4456
+    },
+    {
+      "epoch": 1.4883953915511772,
+      "grad_norm": 0.7590048406114819,
+      "learning_rate": 5.93408122898696e-06,
+      "loss": 0.2066,
+      "step": 4457
+    },
+    {
+      "epoch": 1.4887293371180497,
+      "grad_norm": 0.3828687255820843,
+      "learning_rate": 5.9321720377840245e-06,
+      "loss": 0.1898,
+      "step": 4458
+    },
+    {
+      "epoch": 1.4890632826849224,
+      "grad_norm": 0.3989626077341179,
+      "learning_rate": 5.930262705765526e-06,
+      "loss": 0.1919,
+      "step": 4459
+    },
+    {
+      "epoch": 1.4893972282517949,
+      "grad_norm": 0.4435140197348797,
+      "learning_rate": 5.928353233219893e-06,
+      "loss": 0.2138,
+      "step": 4460
+    },
+    {
+      "epoch": 1.4897311738186676,
+      "grad_norm": 0.41263294889386565,
+      "learning_rate": 5.926443620435572e-06,
+      "loss": 0.199,
+      "step": 4461
+    },
+    {
+      "epoch": 1.49006511938554,
+      "grad_norm": 0.4331460485795739,
+      "learning_rate": 5.924533867701034e-06,
+      "loss": 0.188,
+      "step": 4462
+    },
+    {
+      "epoch": 1.4903990649524128,
+      "grad_norm": 0.41622877671317154,
+      "learning_rate": 5.922623975304771e-06,
+      "loss": 0.1985,
+      "step": 4463
+    },
+    {
+      "epoch": 1.4907330105192853,
+      "grad_norm": 0.4044199019408801,
+      "learning_rate": 5.920713943535291e-06,
+      "loss": 0.1978,
+      "step": 4464
+    },
+    {
+      "epoch": 1.491066956086158,
+      "grad_norm": 0.44762562378654125,
+      "learning_rate": 5.9188037726811285e-06,
+      "loss": 0.2139,
+      "step": 4465
+    },
+    {
+      "epoch": 1.4914009016530305,
+      "grad_norm": 0.4544481373494818,
+      "learning_rate": 5.9168934630308385e-06,
+      "loss": 0.2069,
+      "step": 4466
+    },
+    {
+      "epoch": 1.4917348472199032,
+      "grad_norm": 0.4158141847232926,
+      "learning_rate": 5.914983014872995e-06,
+      "loss": 0.2005,
+      "step": 4467
+    },
+    {
+      "epoch": 1.4920687927867757,
+      "grad_norm": 0.438988331436695,
+      "learning_rate": 5.9130724284961924e-06,
+      "loss": 0.1996,
+      "step": 4468
+    },
+    {
+      "epoch": 1.4924027383536482,
+      "grad_norm": 0.4092313246855316,
+      "learning_rate": 5.91116170418905e-06,
+      "loss": 0.1997,
+      "step": 4469
+    },
+    {
+      "epoch": 1.492736683920521,
+      "grad_norm": 0.5391760700288386,
+      "learning_rate": 5.909250842240203e-06,
+      "loss": 0.2231,
+      "step": 4470
+    },
+    {
+      "epoch": 1.4930706294873937,
+      "grad_norm": 0.42467270074369073,
+      "learning_rate": 5.907339842938309e-06,
+      "loss": 0.1921,
+      "step": 4471
+    },
+    {
+      "epoch": 1.4934045750542662,
+      "grad_norm": 0.4149774782856609,
+      "learning_rate": 5.90542870657205e-06,
+      "loss": 0.1953,
+      "step": 4472
+    },
+    {
+      "epoch": 1.4937385206211387,
+      "grad_norm": 0.4071411942291915,
+      "learning_rate": 5.903517433430123e-06,
+      "loss": 0.1945,
+      "step": 4473
+    },
+    {
+      "epoch": 1.4940724661880114,
+      "grad_norm": 0.4161620755423087,
+      "learning_rate": 5.901606023801248e-06,
+      "loss": 0.196,
+      "step": 4474
+    },
+    {
+      "epoch": 1.494406411754884,
+      "grad_norm": 0.44415983179036833,
+      "learning_rate": 5.899694477974168e-06,
+      "loss": 0.2036,
+      "step": 4475
+    },
+    {
+      "epoch": 1.4947403573217566,
+      "grad_norm": 0.48213305322394434,
+      "learning_rate": 5.897782796237645e-06,
+      "loss": 0.2017,
+      "step": 4476
+    },
+    {
+      "epoch": 1.495074302888629,
+      "grad_norm": 0.4565653405989367,
+      "learning_rate": 5.895870978880457e-06,
+      "loss": 0.2131,
+      "step": 4477
+    },
+    {
+      "epoch": 1.4954082484555018,
+      "grad_norm": 0.38413826906308257,
+      "learning_rate": 5.89395902619141e-06,
+      "loss": 0.1824,
+      "step": 4478
+    },
+    {
+      "epoch": 1.4957421940223743,
+      "grad_norm": 0.3865209015501584,
+      "learning_rate": 5.892046938459327e-06,
+      "loss": 0.1932,
+      "step": 4479
+    },
+    {
+      "epoch": 1.496076139589247,
+      "grad_norm": 0.4485955314433592,
+      "learning_rate": 5.890134715973049e-06,
+      "loss": 0.2067,
+      "step": 4480
+    },
+    {
+      "epoch": 1.4964100851561195,
+      "grad_norm": 0.4159945254787882,
+      "learning_rate": 5.888222359021443e-06,
+      "loss": 0.208,
+      "step": 4481
+    },
+    {
+      "epoch": 1.4967440307229922,
+      "grad_norm": 0.4705989200101088,
+      "learning_rate": 5.8863098678933896e-06,
+      "loss": 0.2105,
+      "step": 4482
+    },
+    {
+      "epoch": 1.4970779762898647,
+      "grad_norm": 0.4875225833597677,
+      "learning_rate": 5.884397242877795e-06,
+      "loss": 0.2036,
+      "step": 4483
+    },
+    {
+      "epoch": 1.4974119218567372,
+      "grad_norm": 0.46559568159684633,
+      "learning_rate": 5.882484484263584e-06,
+      "loss": 0.2078,
+      "step": 4484
+    },
+    {
+      "epoch": 1.49774586742361,
+      "grad_norm": 0.4104886972900785,
+      "learning_rate": 5.8805715923397e-06,
+      "loss": 0.2093,
+      "step": 4485
+    },
+    {
+      "epoch": 1.4980798129904827,
+      "grad_norm": 0.4270246748278954,
+      "learning_rate": 5.87865856739511e-06,
+      "loss": 0.1874,
+      "step": 4486
+    },
+    {
+      "epoch": 1.4984137585573551,
+      "grad_norm": 0.40156162163724024,
+      "learning_rate": 5.876745409718796e-06,
+      "loss": 0.1865,
+      "step": 4487
+    },
+    {
+      "epoch": 1.4987477041242276,
+      "grad_norm": 0.391157655694104,
+      "learning_rate": 5.874832119599766e-06,
+      "loss": 0.1849,
+      "step": 4488
+    },
+    {
+      "epoch": 1.4990816496911004,
+      "grad_norm": 0.4212459484944311,
+      "learning_rate": 5.872918697327042e-06,
+      "loss": 0.1948,
+      "step": 4489
+    },
+    {
+      "epoch": 1.499415595257973,
+      "grad_norm": 0.43743610694660373,
+      "learning_rate": 5.871005143189671e-06,
+      "loss": 0.2022,
+      "step": 4490
+    },
+    {
+      "epoch": 1.4997495408248456,
+      "grad_norm": 0.47954887154563275,
+      "learning_rate": 5.869091457476718e-06,
+      "loss": 0.2165,
+      "step": 4491
+    },
+    {
+      "epoch": 1.500083486391718,
+      "grad_norm": 0.40104355324589935,
+      "learning_rate": 5.8671776404772655e-06,
+      "loss": 0.2011,
+      "step": 4492
+    },
+    {
+      "epoch": 1.5004174319585908,
+      "grad_norm": 0.4065530502322676,
+      "learning_rate": 5.8652636924804206e-06,
+      "loss": 0.1917,
+      "step": 4493
+    },
+    {
+      "epoch": 1.5007513775254635,
+      "grad_norm": 0.39361953435018554,
+      "learning_rate": 5.863349613775308e-06,
+      "loss": 0.1871,
+      "step": 4494
+    },
+    {
+      "epoch": 1.501085323092336,
+      "grad_norm": 0.4220860375266858,
+      "learning_rate": 5.861435404651068e-06,
+      "loss": 0.197,
+      "step": 4495
+    },
+    {
+      "epoch": 1.5014192686592085,
+      "grad_norm": 0.4133784223860879,
+      "learning_rate": 5.859521065396869e-06,
+      "loss": 0.2039,
+      "step": 4496
+    },
+    {
+      "epoch": 1.5017532142260812,
+      "grad_norm": 0.46819851944987745,
+      "learning_rate": 5.857606596301892e-06,
+      "loss": 0.1887,
+      "step": 4497
+    },
+    {
+      "epoch": 1.5020871597929537,
+      "grad_norm": 0.4212627192628078,
+      "learning_rate": 5.85569199765534e-06,
+      "loss": 0.1992,
+      "step": 4498
+    },
+    {
+      "epoch": 1.5024211053598262,
+      "grad_norm": 0.5441159142194298,
+      "learning_rate": 5.853777269746438e-06,
+      "loss": 0.1945,
+      "step": 4499
+    },
+    {
+      "epoch": 1.502755050926699,
+      "grad_norm": 0.38319452570638174,
+      "learning_rate": 5.851862412864426e-06,
+      "loss": 0.1878,
+      "step": 4500
+    },
+    {
+      "epoch": 1.5030889964935716,
+      "grad_norm": 0.39792537321581023,
+      "learning_rate": 5.8499474272985654e-06,
+      "loss": 0.1871,
+      "step": 4501
+    },
+    {
+      "epoch": 1.5034229420604441,
+      "grad_norm": 0.42532885484991806,
+      "learning_rate": 5.848032313338139e-06,
+      "loss": 0.188,
+      "step": 4502
+    },
+    {
+      "epoch": 1.5037568876273166,
+      "grad_norm": 0.40589177813814775,
+      "learning_rate": 5.846117071272444e-06,
+      "loss": 0.1919,
+      "step": 4503
+    },
+    {
+      "epoch": 1.5040908331941893,
+      "grad_norm": 0.4038764967885529,
+      "learning_rate": 5.844201701390806e-06,
+      "loss": 0.1947,
+      "step": 4504
+    },
+    {
+      "epoch": 1.504424778761062,
+      "grad_norm": 0.42058966558438154,
+      "learning_rate": 5.842286203982559e-06,
+      "loss": 0.1875,
+      "step": 4505
+    },
+    {
+      "epoch": 1.5047587243279346,
+      "grad_norm": 0.44301115684433295,
+      "learning_rate": 5.840370579337063e-06,
+      "loss": 0.2088,
+      "step": 4506
+    },
+    {
+      "epoch": 1.505092669894807,
+      "grad_norm": 0.4506602942327276,
+      "learning_rate": 5.838454827743697e-06,
+      "loss": 0.214,
+      "step": 4507
+    },
+    {
+      "epoch": 1.5054266154616798,
+      "grad_norm": 0.41888527162391026,
+      "learning_rate": 5.8365389494918565e-06,
+      "loss": 0.1989,
+      "step": 4508
+    },
+    {
+      "epoch": 1.5057605610285525,
+      "grad_norm": 0.46234137248207496,
+      "learning_rate": 5.834622944870959e-06,
+      "loss": 0.2071,
+      "step": 4509
+    },
+    {
+      "epoch": 1.506094506595425,
+      "grad_norm": 0.3991349847409579,
+      "learning_rate": 5.832706814170437e-06,
+      "loss": 0.1921,
+      "step": 4510
+    },
+    {
+      "epoch": 1.5064284521622975,
+      "grad_norm": 0.45328443937762986,
+      "learning_rate": 5.830790557679746e-06,
+      "loss": 0.2094,
+      "step": 4511
+    },
+    {
+      "epoch": 1.5067623977291702,
+      "grad_norm": 0.4545739233729224,
+      "learning_rate": 5.8288741756883585e-06,
+      "loss": 0.2233,
+      "step": 4512
+    },
+    {
+      "epoch": 1.5070963432960427,
+      "grad_norm": 0.4173274151899909,
+      "learning_rate": 5.826957668485768e-06,
+      "loss": 0.1891,
+      "step": 4513
+    },
+    {
+      "epoch": 1.5074302888629152,
+      "grad_norm": 0.4514509261491327,
+      "learning_rate": 5.825041036361484e-06,
+      "loss": 0.2034,
+      "step": 4514
+    },
+    {
+      "epoch": 1.507764234429788,
+      "grad_norm": 0.4245100612780044,
+      "learning_rate": 5.823124279605037e-06,
+      "loss": 0.2083,
+      "step": 4515
+    },
+    {
+      "epoch": 1.5080981799966606,
+      "grad_norm": 0.41279249791467926,
+      "learning_rate": 5.821207398505976e-06,
+      "loss": 0.2131,
+      "step": 4516
+    },
+    {
+      "epoch": 1.5084321255635331,
+      "grad_norm": 0.4430510447851048,
+      "learning_rate": 5.819290393353867e-06,
+      "loss": 0.2233,
+      "step": 4517
+    },
+    {
+      "epoch": 1.5087660711304056,
+      "grad_norm": 0.4270138809636306,
+      "learning_rate": 5.817373264438297e-06,
+      "loss": 0.198,
+      "step": 4518
+    },
+    {
+      "epoch": 1.5091000166972783,
+      "grad_norm": 0.3855525070150281,
+      "learning_rate": 5.815456012048873e-06,
+      "loss": 0.1943,
+      "step": 4519
+    },
+    {
+      "epoch": 1.509433962264151,
+      "grad_norm": 0.3951585582354066,
+      "learning_rate": 5.8135386364752154e-06,
+      "loss": 0.1973,
+      "step": 4520
+    },
+    {
+      "epoch": 1.5097679078310235,
+      "grad_norm": 0.44097574717875243,
+      "learning_rate": 5.8116211380069675e-06,
+      "loss": 0.1974,
+      "step": 4521
+    },
+    {
+      "epoch": 1.510101853397896,
+      "grad_norm": 0.418568197709888,
+      "learning_rate": 5.809703516933791e-06,
+      "loss": 0.1981,
+      "step": 4522
+    },
+    {
+      "epoch": 1.5104357989647688,
+      "grad_norm": 0.39429989332524545,
+      "learning_rate": 5.807785773545364e-06,
+      "loss": 0.1885,
+      "step": 4523
+    },
+    {
+      "epoch": 1.5107697445316415,
+      "grad_norm": 0.5133283842688638,
+      "learning_rate": 5.805867908131384e-06,
+      "loss": 0.2054,
+      "step": 4524
+    },
+    {
+      "epoch": 1.511103690098514,
+      "grad_norm": 0.4138694184824036,
+      "learning_rate": 5.803949920981568e-06,
+      "loss": 0.2066,
+      "step": 4525
+    },
+    {
+      "epoch": 1.5114376356653865,
+      "grad_norm": 0.451090788654152,
+      "learning_rate": 5.802031812385651e-06,
+      "loss": 0.207,
+      "step": 4526
+    },
+    {
+      "epoch": 1.5117715812322592,
+      "grad_norm": 0.4391511075816685,
+      "learning_rate": 5.800113582633384e-06,
+      "loss": 0.1972,
+      "step": 4527
+    },
+    {
+      "epoch": 1.512105526799132,
+      "grad_norm": 0.4062998929666656,
+      "learning_rate": 5.7981952320145405e-06,
+      "loss": 0.2005,
+      "step": 4528
+    },
+    {
+      "epoch": 1.5124394723660042,
+      "grad_norm": 0.4147263705858617,
+      "learning_rate": 5.796276760818908e-06,
+      "loss": 0.2,
+      "step": 4529
+    },
+    {
+      "epoch": 1.512773417932877,
+      "grad_norm": 0.4184233692356123,
+      "learning_rate": 5.794358169336295e-06,
+      "loss": 0.2072,
+      "step": 4530
+    },
+    {
+      "epoch": 1.5131073634997496,
+      "grad_norm": 0.42606865720950565,
+      "learning_rate": 5.792439457856528e-06,
+      "loss": 0.2058,
+      "step": 4531
+    },
+    {
+      "epoch": 1.513441309066622,
+      "grad_norm": 0.4031899587249856,
+      "learning_rate": 5.790520626669449e-06,
+      "loss": 0.1865,
+      "step": 4532
+    },
+    {
+      "epoch": 1.5137752546334946,
+      "grad_norm": 0.39822141992740223,
+      "learning_rate": 5.788601676064922e-06,
+      "loss": 0.1849,
+      "step": 4533
+    },
+    {
+      "epoch": 1.5141092002003673,
+      "grad_norm": 0.41169793772637675,
+      "learning_rate": 5.786682606332827e-06,
+      "loss": 0.198,
+      "step": 4534
+    },
+    {
+      "epoch": 1.51444314576724,
+      "grad_norm": 0.41321607553783124,
+      "learning_rate": 5.78476341776306e-06,
+      "loss": 0.1988,
+      "step": 4535
+    },
+    {
+      "epoch": 1.5147770913341125,
+      "grad_norm": 0.4333225224257672,
+      "learning_rate": 5.782844110645539e-06,
+      "loss": 0.1957,
+      "step": 4536
+    },
+    {
+      "epoch": 1.515111036900985,
+      "grad_norm": 0.42052885937644136,
+      "learning_rate": 5.780924685270198e-06,
+      "loss": 0.198,
+      "step": 4537
+    },
+    {
+      "epoch": 1.5154449824678577,
+      "grad_norm": 0.4037698414311071,
+      "learning_rate": 5.779005141926988e-06,
+      "loss": 0.2,
+      "step": 4538
+    },
+    {
+      "epoch": 1.5157789280347305,
+      "grad_norm": 0.3818155530534464,
+      "learning_rate": 5.777085480905877e-06,
+      "loss": 0.1915,
+      "step": 4539
+    },
+    {
+      "epoch": 1.516112873601603,
+      "grad_norm": 0.3963973853574785,
+      "learning_rate": 5.7751657024968565e-06,
+      "loss": 0.1988,
+      "step": 4540
+    },
+    {
+      "epoch": 1.5164468191684755,
+      "grad_norm": 0.45356922202417344,
+      "learning_rate": 5.773245806989929e-06,
+      "loss": 0.2118,
+      "step": 4541
+    },
+    {
+      "epoch": 1.5167807647353482,
+      "grad_norm": 0.41828720268999303,
+      "learning_rate": 5.771325794675117e-06,
+      "loss": 0.2065,
+      "step": 4542
+    },
+    {
+      "epoch": 1.517114710302221,
+      "grad_norm": 0.38469626361124054,
+      "learning_rate": 5.769405665842461e-06,
+      "loss": 0.1849,
+      "step": 4543
+    },
+    {
+      "epoch": 1.5174486558690934,
+      "grad_norm": 0.3785273566123282,
+      "learning_rate": 5.767485420782021e-06,
+      "loss": 0.1876,
+      "step": 4544
+    },
+    {
+      "epoch": 1.5177826014359659,
+      "grad_norm": 0.41324123757089976,
+      "learning_rate": 5.7655650597838704e-06,
+      "loss": 0.2107,
+      "step": 4545
+    },
+    {
+      "epoch": 1.5181165470028386,
+      "grad_norm": 0.4023559102615698,
+      "learning_rate": 5.7636445831381034e-06,
+      "loss": 0.1958,
+      "step": 4546
+    },
+    {
+      "epoch": 1.518450492569711,
+      "grad_norm": 0.4268003997100768,
+      "learning_rate": 5.761723991134831e-06,
+      "loss": 0.2005,
+      "step": 4547
+    },
+    {
+      "epoch": 1.5187844381365836,
+      "grad_norm": 0.41879175195622614,
+      "learning_rate": 5.759803284064181e-06,
+      "loss": 0.1974,
+      "step": 4548
+    },
+    {
+      "epoch": 1.5191183837034563,
+      "grad_norm": 0.41661555641918735,
+      "learning_rate": 5.757882462216299e-06,
+      "loss": 0.2028,
+      "step": 4549
+    },
+    {
+      "epoch": 1.519452329270329,
+      "grad_norm": 0.5008947195462263,
+      "learning_rate": 5.755961525881345e-06,
+      "loss": 0.211,
+      "step": 4550
+    },
+    {
+      "epoch": 1.5197862748372015,
+      "grad_norm": 0.43921728419522166,
+      "learning_rate": 5.7540404753495034e-06,
+      "loss": 0.2032,
+      "step": 4551
+    },
+    {
+      "epoch": 1.520120220404074,
+      "grad_norm": 0.44693077234088574,
+      "learning_rate": 5.75211931091097e-06,
+      "loss": 0.2265,
+      "step": 4552
+    },
+    {
+      "epoch": 1.5204541659709467,
+      "grad_norm": 0.4001694902689412,
+      "learning_rate": 5.750198032855956e-06,
+      "loss": 0.2,
+      "step": 4553
+    },
+    {
+      "epoch": 1.5207881115378195,
+      "grad_norm": 0.4045351162020478,
+      "learning_rate": 5.748276641474698e-06,
+      "loss": 0.1983,
+      "step": 4554
+    },
+    {
+      "epoch": 1.521122057104692,
+      "grad_norm": 0.3977245043542055,
+      "learning_rate": 5.746355137057442e-06,
+      "loss": 0.199,
+      "step": 4555
+    },
+    {
+      "epoch": 1.5214560026715644,
+      "grad_norm": 0.4176634955723099,
+      "learning_rate": 5.7444335198944555e-06,
+      "loss": 0.2065,
+      "step": 4556
+    },
+    {
+      "epoch": 1.5217899482384372,
+      "grad_norm": 0.4086844831410891,
+      "learning_rate": 5.7425117902760195e-06,
+      "loss": 0.2004,
+      "step": 4557
+    },
+    {
+      "epoch": 1.5221238938053099,
+      "grad_norm": 0.441434511195485,
+      "learning_rate": 5.7405899484924346e-06,
+      "loss": 0.2116,
+      "step": 4558
+    },
+    {
+      "epoch": 1.5224578393721824,
+      "grad_norm": 0.41164529074446093,
+      "learning_rate": 5.738667994834019e-06,
+      "loss": 0.1939,
+      "step": 4559
+    },
+    {
+      "epoch": 1.5227917849390549,
+      "grad_norm": 0.4069363998020001,
+      "learning_rate": 5.736745929591103e-06,
+      "loss": 0.1885,
+      "step": 4560
+    },
+    {
+      "epoch": 1.5231257305059276,
+      "grad_norm": 0.40733152134950845,
+      "learning_rate": 5.734823753054042e-06,
+      "loss": 0.2001,
+      "step": 4561
+    },
+    {
+      "epoch": 1.5234596760728,
+      "grad_norm": 0.3828618387327402,
+      "learning_rate": 5.732901465513199e-06,
+      "loss": 0.1982,
+      "step": 4562
+    },
+    {
+      "epoch": 1.5237936216396726,
+      "grad_norm": 0.4305349510985267,
+      "learning_rate": 5.73097906725896e-06,
+      "loss": 0.1974,
+      "step": 4563
+    },
+    {
+      "epoch": 1.5241275672065453,
+      "grad_norm": 0.40465587338482667,
+      "learning_rate": 5.729056558581727e-06,
+      "loss": 0.1986,
+      "step": 4564
+    },
+    {
+      "epoch": 1.524461512773418,
+      "grad_norm": 0.471282991054306,
+      "learning_rate": 5.727133939771915e-06,
+      "loss": 0.2081,
+      "step": 4565
+    },
+    {
+      "epoch": 1.5247954583402905,
+      "grad_norm": 0.4088188439945464,
+      "learning_rate": 5.725211211119961e-06,
+      "loss": 0.1952,
+      "step": 4566
+    },
+    {
+      "epoch": 1.525129403907163,
+      "grad_norm": 0.3934161375353832,
+      "learning_rate": 5.723288372916315e-06,
+      "loss": 0.1901,
+      "step": 4567
+    },
+    {
+      "epoch": 1.5254633494740357,
+      "grad_norm": 0.39417936401264786,
+      "learning_rate": 5.721365425451442e-06,
+      "loss": 0.195,
+      "step": 4568
+    },
+    {
+      "epoch": 1.5257972950409084,
+      "grad_norm": 0.41813761847460523,
+      "learning_rate": 5.719442369015828e-06,
+      "loss": 0.2002,
+      "step": 4569
+    },
+    {
+      "epoch": 1.526131240607781,
+      "grad_norm": 0.3928817442742676,
+      "learning_rate": 5.717519203899975e-06,
+      "loss": 0.1821,
+      "step": 4570
+    },
+    {
+      "epoch": 1.5264651861746534,
+      "grad_norm": 0.42060022944156716,
+      "learning_rate": 5.715595930394396e-06,
+      "loss": 0.197,
+      "step": 4571
+    },
+    {
+      "epoch": 1.5267991317415261,
+      "grad_norm": 0.414538136672965,
+      "learning_rate": 5.713672548789626e-06,
+      "loss": 0.2109,
+      "step": 4572
+    },
+    {
+      "epoch": 1.5271330773083989,
+      "grad_norm": 0.40300175281434797,
+      "learning_rate": 5.711749059376215e-06,
+      "loss": 0.194,
+      "step": 4573
+    },
+    {
+      "epoch": 1.5274670228752714,
+      "grad_norm": 0.4202683530403047,
+      "learning_rate": 5.7098254624447255e-06,
+      "loss": 0.1891,
+      "step": 4574
+    },
+    {
+      "epoch": 1.5278009684421439,
+      "grad_norm": 0.41121947338511855,
+      "learning_rate": 5.707901758285745e-06,
+      "loss": 0.1911,
+      "step": 4575
+    },
+    {
+      "epoch": 1.5281349140090166,
+      "grad_norm": 0.3973014662751213,
+      "learning_rate": 5.705977947189868e-06,
+      "loss": 0.1994,
+      "step": 4576
+    },
+    {
+      "epoch": 1.5284688595758893,
+      "grad_norm": 0.43891931027961867,
+      "learning_rate": 5.704054029447708e-06,
+      "loss": 0.2007,
+      "step": 4577
+    },
+    {
+      "epoch": 1.5288028051427616,
+      "grad_norm": 0.4502833076199682,
+      "learning_rate": 5.702130005349899e-06,
+      "loss": 0.214,
+      "step": 4578
+    },
+    {
+      "epoch": 1.5291367507096343,
+      "grad_norm": 0.44959299233942557,
+      "learning_rate": 5.700205875187084e-06,
+      "loss": 0.2153,
+      "step": 4579
+    },
+    {
+      "epoch": 1.529470696276507,
+      "grad_norm": 0.4305977208026748,
+      "learning_rate": 5.698281639249927e-06,
+      "loss": 0.2123,
+      "step": 4580
+    },
+    {
+      "epoch": 1.5298046418433795,
+      "grad_norm": 0.43603488906842863,
+      "learning_rate": 5.696357297829106e-06,
+      "loss": 0.2089,
+      "step": 4581
+    },
+    {
+      "epoch": 1.530138587410252,
+      "grad_norm": 0.4250706270087593,
+      "learning_rate": 5.6944328512153165e-06,
+      "loss": 0.1977,
+      "step": 4582
+    },
+    {
+      "epoch": 1.5304725329771247,
+      "grad_norm": 0.4198326353777762,
+      "learning_rate": 5.692508299699269e-06,
+      "loss": 0.2009,
+      "step": 4583
+    },
+    {
+      "epoch": 1.5308064785439974,
+      "grad_norm": 0.395956572492155,
+      "learning_rate": 5.690583643571687e-06,
+      "loss": 0.198,
+      "step": 4584
+    },
+    {
+      "epoch": 1.53114042411087,
+      "grad_norm": 0.43947140761901465,
+      "learning_rate": 5.688658883123315e-06,
+      "loss": 0.2027,
+      "step": 4585
+    },
+    {
+      "epoch": 1.5314743696777424,
+      "grad_norm": 0.4108183015582965,
+      "learning_rate": 5.68673401864491e-06,
+      "loss": 0.2118,
+      "step": 4586
+    },
+    {
+      "epoch": 1.5318083152446151,
+      "grad_norm": 0.41182034693533875,
+      "learning_rate": 5.684809050427247e-06,
+      "loss": 0.1987,
+      "step": 4587
+    },
+    {
+      "epoch": 1.5321422608114879,
+      "grad_norm": 0.4349586140015333,
+      "learning_rate": 5.682883978761111e-06,
+      "loss": 0.2002,
+      "step": 4588
+    },
+    {
+      "epoch": 1.5324762063783604,
+      "grad_norm": 0.39453978972461545,
+      "learning_rate": 5.680958803937311e-06,
+      "loss": 0.1962,
+      "step": 4589
+    },
+    {
+      "epoch": 1.5328101519452328,
+      "grad_norm": 0.4430655520347526,
+      "learning_rate": 5.6790335262466645e-06,
+      "loss": 0.2243,
+      "step": 4590
+    },
+    {
+      "epoch": 1.5331440975121056,
+      "grad_norm": 0.38426560740617555,
+      "learning_rate": 5.677108145980008e-06,
+      "loss": 0.1949,
+      "step": 4591
+    },
+    {
+      "epoch": 1.5334780430789783,
+      "grad_norm": 0.41412598049147914,
+      "learning_rate": 5.675182663428196e-06,
+      "loss": 0.2016,
+      "step": 4592
+    },
+    {
+      "epoch": 1.5338119886458508,
+      "grad_norm": 0.3952364153216541,
+      "learning_rate": 5.673257078882091e-06,
+      "loss": 0.1957,
+      "step": 4593
+    },
+    {
+      "epoch": 1.5341459342127233,
+      "grad_norm": 0.4347041827090324,
+      "learning_rate": 5.671331392632577e-06,
+      "loss": 0.2062,
+      "step": 4594
+    },
+    {
+      "epoch": 1.534479879779596,
+      "grad_norm": 0.4243751095481007,
+      "learning_rate": 5.6694056049705506e-06,
+      "loss": 0.2011,
+      "step": 4595
+    },
+    {
+      "epoch": 1.5348138253464685,
+      "grad_norm": 0.41160762842006665,
+      "learning_rate": 5.667479716186927e-06,
+      "loss": 0.2053,
+      "step": 4596
+    },
+    {
+      "epoch": 1.535147770913341,
+      "grad_norm": 0.45726125048358807,
+      "learning_rate": 5.665553726572631e-06,
+      "loss": 0.206,
+      "step": 4597
+    },
+    {
+      "epoch": 1.5354817164802137,
+      "grad_norm": 0.395285907262752,
+      "learning_rate": 5.663627636418611e-06,
+      "loss": 0.1837,
+      "step": 4598
+    },
+    {
+      "epoch": 1.5358156620470864,
+      "grad_norm": 0.3940866152777301,
+      "learning_rate": 5.661701446015821e-06,
+      "loss": 0.1945,
+      "step": 4599
+    },
+    {
+      "epoch": 1.536149607613959,
+      "grad_norm": 0.39866982139324986,
+      "learning_rate": 5.659775155655235e-06,
+      "loss": 0.1928,
+      "step": 4600
+    },
+    {
+      "epoch": 1.5364835531808314,
+      "grad_norm": 0.4186774736236097,
+      "learning_rate": 5.6578487656278446e-06,
+      "loss": 0.1981,
+      "step": 4601
+    },
+    {
+      "epoch": 1.5368174987477041,
+      "grad_norm": 0.41470326955184655,
+      "learning_rate": 5.655922276224652e-06,
+      "loss": 0.2064,
+      "step": 4602
+    },
+    {
+      "epoch": 1.5371514443145768,
+      "grad_norm": 0.4092155718896311,
+      "learning_rate": 5.653995687736676e-06,
+      "loss": 0.2006,
+      "step": 4603
+    },
+    {
+      "epoch": 1.5374853898814493,
+      "grad_norm": 0.4106065112514508,
+      "learning_rate": 5.652069000454951e-06,
+      "loss": 0.2022,
+      "step": 4604
+    },
+    {
+      "epoch": 1.5378193354483218,
+      "grad_norm": 0.4858555749514037,
+      "learning_rate": 5.650142214670527e-06,
+      "loss": 0.1805,
+      "step": 4605
+    },
+    {
+      "epoch": 1.5381532810151946,
+      "grad_norm": 0.4812494451824753,
+      "learning_rate": 5.648215330674464e-06,
+      "loss": 0.1882,
+      "step": 4606
+    },
+    {
+      "epoch": 1.5384872265820673,
+      "grad_norm": 0.4261470703710842,
+      "learning_rate": 5.646288348757845e-06,
+      "loss": 0.1926,
+      "step": 4607
+    },
+    {
+      "epoch": 1.5388211721489398,
+      "grad_norm": 0.3952136708762545,
+      "learning_rate": 5.64436126921176e-06,
+      "loss": 0.1975,
+      "step": 4608
+    },
+    {
+      "epoch": 1.5391551177158123,
+      "grad_norm": 0.3970827209666597,
+      "learning_rate": 5.642434092327318e-06,
+      "loss": 0.1999,
+      "step": 4609
+    },
+    {
+      "epoch": 1.539489063282685,
+      "grad_norm": 0.40998763479801215,
+      "learning_rate": 5.640506818395643e-06,
+      "loss": 0.1986,
+      "step": 4610
+    },
+    {
+      "epoch": 1.5398230088495575,
+      "grad_norm": 0.47084062645452907,
+      "learning_rate": 5.638579447707871e-06,
+      "loss": 0.193,
+      "step": 4611
+    },
+    {
+      "epoch": 1.54015695441643,
+      "grad_norm": 0.3860268369786831,
+      "learning_rate": 5.636651980555153e-06,
+      "loss": 0.1909,
+      "step": 4612
+    },
+    {
+      "epoch": 1.5404908999833027,
+      "grad_norm": 0.3942593539079482,
+      "learning_rate": 5.634724417228658e-06,
+      "loss": 0.1966,
+      "step": 4613
+    },
+    {
+      "epoch": 1.5408248455501754,
+      "grad_norm": 0.42574482432433103,
+      "learning_rate": 5.632796758019566e-06,
+      "loss": 0.1917,
+      "step": 4614
+    },
+    {
+      "epoch": 1.541158791117048,
+      "grad_norm": 0.39053070306639825,
+      "learning_rate": 5.630869003219072e-06,
+      "loss": 0.1977,
+      "step": 4615
+    },
+    {
+      "epoch": 1.5414927366839204,
+      "grad_norm": 0.43083185683956327,
+      "learning_rate": 5.628941153118388e-06,
+      "loss": 0.2072,
+      "step": 4616
+    },
+    {
+      "epoch": 1.5418266822507931,
+      "grad_norm": 0.36680215926548226,
+      "learning_rate": 5.627013208008737e-06,
+      "loss": 0.1708,
+      "step": 4617
+    },
+    {
+      "epoch": 1.5421606278176658,
+      "grad_norm": 0.4168771060059623,
+      "learning_rate": 5.625085168181357e-06,
+      "loss": 0.1957,
+      "step": 4618
+    },
+    {
+      "epoch": 1.5424945733845383,
+      "grad_norm": 0.4159514176383808,
+      "learning_rate": 5.623157033927503e-06,
+      "loss": 0.1933,
+      "step": 4619
+    },
+    {
+      "epoch": 1.5428285189514108,
+      "grad_norm": 0.45704579742530344,
+      "learning_rate": 5.621228805538443e-06,
+      "loss": 0.1997,
+      "step": 4620
+    },
+    {
+      "epoch": 1.5431624645182835,
+      "grad_norm": 0.428254804443801,
+      "learning_rate": 5.619300483305454e-06,
+      "loss": 0.1862,
+      "step": 4621
+    },
+    {
+      "epoch": 1.5434964100851563,
+      "grad_norm": 0.4203762697429112,
+      "learning_rate": 5.617372067519837e-06,
+      "loss": 0.2034,
+      "step": 4622
+    },
+    {
+      "epoch": 1.5438303556520288,
+      "grad_norm": 0.4200993106054955,
+      "learning_rate": 5.6154435584729e-06,
+      "loss": 0.208,
+      "step": 4623
+    },
+    {
+      "epoch": 1.5441643012189012,
+      "grad_norm": 0.39120604266589976,
+      "learning_rate": 5.6135149564559665e-06,
+      "loss": 0.1893,
+      "step": 4624
+    },
+    {
+      "epoch": 1.544498246785774,
+      "grad_norm": 0.4347411574630664,
+      "learning_rate": 5.611586261760375e-06,
+      "loss": 0.181,
+      "step": 4625
+    },
+    {
+      "epoch": 1.5448321923526467,
+      "grad_norm": 0.43018401156890745,
+      "learning_rate": 5.609657474677478e-06,
+      "loss": 0.1994,
+      "step": 4626
+    },
+    {
+      "epoch": 1.545166137919519,
+      "grad_norm": 0.4820806633908888,
+      "learning_rate": 5.607728595498641e-06,
+      "loss": 0.2073,
+      "step": 4627
+    },
+    {
+      "epoch": 1.5455000834863917,
+      "grad_norm": 0.40368430637167996,
+      "learning_rate": 5.6057996245152435e-06,
+      "loss": 0.1919,
+      "step": 4628
+    },
+    {
+      "epoch": 1.5458340290532644,
+      "grad_norm": 0.406341895380249,
+      "learning_rate": 5.603870562018679e-06,
+      "loss": 0.1957,
+      "step": 4629
+    },
+    {
+      "epoch": 1.5461679746201369,
+      "grad_norm": 0.4120838046068009,
+      "learning_rate": 5.601941408300358e-06,
+      "loss": 0.2073,
+      "step": 4630
+    },
+    {
+      "epoch": 1.5465019201870094,
+      "grad_norm": 0.40592326676295587,
+      "learning_rate": 5.600012163651698e-06,
+      "loss": 0.1984,
+      "step": 4631
+    },
+    {
+      "epoch": 1.546835865753882,
+      "grad_norm": 0.4363803396643685,
+      "learning_rate": 5.598082828364134e-06,
+      "loss": 0.1972,
+      "step": 4632
+    },
+    {
+      "epoch": 1.5471698113207548,
+      "grad_norm": 0.43058371265331713,
+      "learning_rate": 5.596153402729118e-06,
+      "loss": 0.196,
+      "step": 4633
+    },
+    {
+      "epoch": 1.5475037568876273,
+      "grad_norm": 0.39059793054857367,
+      "learning_rate": 5.594223887038113e-06,
+      "loss": 0.181,
+      "step": 4634
+    },
+    {
+      "epoch": 1.5478377024544998,
+      "grad_norm": 0.4261233357009109,
+      "learning_rate": 5.592294281582591e-06,
+      "loss": 0.1893,
+      "step": 4635
+    },
+    {
+      "epoch": 1.5481716480213725,
+      "grad_norm": 0.41289802429846045,
+      "learning_rate": 5.590364586654043e-06,
+      "loss": 0.2031,
+      "step": 4636
+    },
+    {
+      "epoch": 1.5485055935882452,
+      "grad_norm": 0.45183145302663547,
+      "learning_rate": 5.588434802543975e-06,
+      "loss": 0.1886,
+      "step": 4637
+    },
+    {
+      "epoch": 1.5488395391551177,
+      "grad_norm": 0.4250974981387959,
+      "learning_rate": 5.5865049295439e-06,
+      "loss": 0.1951,
+      "step": 4638
+    },
+    {
+      "epoch": 1.5491734847219902,
+      "grad_norm": 0.43975870182322807,
+      "learning_rate": 5.584574967945351e-06,
+      "loss": 0.2104,
+      "step": 4639
+    },
+    {
+      "epoch": 1.549507430288863,
+      "grad_norm": 0.44559355295469355,
+      "learning_rate": 5.582644918039869e-06,
+      "loss": 0.2092,
+      "step": 4640
+    },
+    {
+      "epoch": 1.5498413758557357,
+      "grad_norm": 0.41539145621347284,
+      "learning_rate": 5.580714780119011e-06,
+      "loss": 0.2034,
+      "step": 4641
+    },
+    {
+      "epoch": 1.5501753214226082,
+      "grad_norm": 0.40591271120703043,
+      "learning_rate": 5.578784554474348e-06,
+      "loss": 0.188,
+      "step": 4642
+    },
+    {
+      "epoch": 1.5505092669894807,
+      "grad_norm": 0.6532970191187181,
+      "learning_rate": 5.5768542413974645e-06,
+      "loss": 0.2037,
+      "step": 4643
+    },
+    {
+      "epoch": 1.5508432125563534,
+      "grad_norm": 0.4315096314647443,
+      "learning_rate": 5.574923841179953e-06,
+      "loss": 0.2007,
+      "step": 4644
+    },
+    {
+      "epoch": 1.5511771581232259,
+      "grad_norm": 0.38269782258347723,
+      "learning_rate": 5.572993354113429e-06,
+      "loss": 0.183,
+      "step": 4645
+    },
+    {
+      "epoch": 1.5515111036900984,
+      "grad_norm": 0.5076233296132883,
+      "learning_rate": 5.5710627804895105e-06,
+      "loss": 0.214,
+      "step": 4646
+    },
+    {
+      "epoch": 1.551845049256971,
+      "grad_norm": 0.41321256949804475,
+      "learning_rate": 5.569132120599834e-06,
+      "loss": 0.1927,
+      "step": 4647
+    },
+    {
+      "epoch": 1.5521789948238438,
+      "grad_norm": 0.39911707248154793,
+      "learning_rate": 5.567201374736051e-06,
+      "loss": 0.1849,
+      "step": 4648
+    },
+    {
+      "epoch": 1.5525129403907163,
+      "grad_norm": 0.44520911248754574,
+      "learning_rate": 5.565270543189821e-06,
+      "loss": 0.2107,
+      "step": 4649
+    },
+    {
+      "epoch": 1.5528468859575888,
+      "grad_norm": 0.3773856938490637,
+      "learning_rate": 5.563339626252819e-06,
+      "loss": 0.1834,
+      "step": 4650
+    },
+    {
+      "epoch": 1.5531808315244615,
+      "grad_norm": 0.39836943833180216,
+      "learning_rate": 5.561408624216734e-06,
+      "loss": 0.1834,
+      "step": 4651
+    },
+    {
+      "epoch": 1.5535147770913342,
+      "grad_norm": 0.39621414053364384,
+      "learning_rate": 5.559477537373267e-06,
+      "loss": 0.1918,
+      "step": 4652
+    },
+    {
+      "epoch": 1.5538487226582067,
+      "grad_norm": 0.4435868338139983,
+      "learning_rate": 5.557546366014129e-06,
+      "loss": 0.1853,
+      "step": 4653
+    },
+    {
+      "epoch": 1.5541826682250792,
+      "grad_norm": 0.3993984001133118,
+      "learning_rate": 5.555615110431049e-06,
+      "loss": 0.1993,
+      "step": 4654
+    },
+    {
+      "epoch": 1.554516613791952,
+      "grad_norm": 0.41979545755044656,
+      "learning_rate": 5.553683770915763e-06,
+      "loss": 0.2005,
+      "step": 4655
+    },
+    {
+      "epoch": 1.5548505593588247,
+      "grad_norm": 0.4128510095489835,
+      "learning_rate": 5.551752347760023e-06,
+      "loss": 0.1937,
+      "step": 4656
+    },
+    {
+      "epoch": 1.5551845049256972,
+      "grad_norm": 0.41228406880777296,
+      "learning_rate": 5.549820841255597e-06,
+      "loss": 0.1934,
+      "step": 4657
+    },
+    {
+      "epoch": 1.5555184504925696,
+      "grad_norm": 0.40502569693476115,
+      "learning_rate": 5.547889251694257e-06,
+      "loss": 0.1993,
+      "step": 4658
+    },
+    {
+      "epoch": 1.5558523960594424,
+      "grad_norm": 0.45309964934422253,
+      "learning_rate": 5.545957579367795e-06,
+      "loss": 0.2157,
+      "step": 4659
+    },
+    {
+      "epoch": 1.5561863416263149,
+      "grad_norm": 0.4075848451781985,
+      "learning_rate": 5.544025824568011e-06,
+      "loss": 0.1882,
+      "step": 4660
+    },
+    {
+      "epoch": 1.5565202871931874,
+      "grad_norm": 0.41425823932525097,
+      "learning_rate": 5.542093987586722e-06,
+      "loss": 0.1992,
+      "step": 4661
+    },
+    {
+      "epoch": 1.55685423276006,
+      "grad_norm": 0.40778580005587917,
+      "learning_rate": 5.540162068715752e-06,
+      "loss": 0.1909,
+      "step": 4662
+    },
+    {
+      "epoch": 1.5571881783269328,
+      "grad_norm": 0.40536508487131023,
+      "learning_rate": 5.538230068246942e-06,
+      "loss": 0.2013,
+      "step": 4663
+    },
+    {
+      "epoch": 1.5575221238938053,
+      "grad_norm": 0.45353755000496815,
+      "learning_rate": 5.536297986472142e-06,
+      "loss": 0.1986,
+      "step": 4664
+    },
+    {
+      "epoch": 1.5578560694606778,
+      "grad_norm": 0.38378580944056184,
+      "learning_rate": 5.534365823683219e-06,
+      "loss": 0.1869,
+      "step": 4665
+    },
+    {
+      "epoch": 1.5581900150275505,
+      "grad_norm": 0.5116844500041666,
+      "learning_rate": 5.532433580172044e-06,
+      "loss": 0.1968,
+      "step": 4666
+    },
+    {
+      "epoch": 1.5585239605944232,
+      "grad_norm": 0.40984005892330067,
+      "learning_rate": 5.5305012562305075e-06,
+      "loss": 0.1805,
+      "step": 4667
+    },
+    {
+      "epoch": 1.5588579061612957,
+      "grad_norm": 0.4695150357003394,
+      "learning_rate": 5.528568852150511e-06,
+      "loss": 0.2131,
+      "step": 4668
+    },
+    {
+      "epoch": 1.5591918517281682,
+      "grad_norm": 0.454406568295344,
+      "learning_rate": 5.526636368223965e-06,
+      "loss": 0.2069,
+      "step": 4669
+    },
+    {
+      "epoch": 1.559525797295041,
+      "grad_norm": 0.41198712780712915,
+      "learning_rate": 5.524703804742793e-06,
+      "loss": 0.2079,
+      "step": 4670
+    },
+    {
+      "epoch": 1.5598597428619136,
+      "grad_norm": 0.3930997269695332,
+      "learning_rate": 5.522771161998936e-06,
+      "loss": 0.2006,
+      "step": 4671
+    },
+    {
+      "epoch": 1.5601936884287861,
+      "grad_norm": 0.4132056747503601,
+      "learning_rate": 5.52083844028434e-06,
+      "loss": 0.1947,
+      "step": 4672
+    },
+    {
+      "epoch": 1.5605276339956586,
+      "grad_norm": 0.44323946531045155,
+      "learning_rate": 5.518905639890961e-06,
+      "loss": 0.2009,
+      "step": 4673
+    },
+    {
+      "epoch": 1.5608615795625314,
+      "grad_norm": 0.43128682055477907,
+      "learning_rate": 5.516972761110778e-06,
+      "loss": 0.2031,
+      "step": 4674
+    },
+    {
+      "epoch": 1.561195525129404,
+      "grad_norm": 0.4152986643855445,
+      "learning_rate": 5.515039804235772e-06,
+      "loss": 0.1924,
+      "step": 4675
+    },
+    {
+      "epoch": 1.5615294706962763,
+      "grad_norm": 0.43452047598785126,
+      "learning_rate": 5.51310676955794e-06,
+      "loss": 0.1928,
+      "step": 4676
+    },
+    {
+      "epoch": 1.561863416263149,
+      "grad_norm": 0.400631942313824,
+      "learning_rate": 5.511173657369287e-06,
+      "loss": 0.1914,
+      "step": 4677
+    },
+    {
+      "epoch": 1.5621973618300218,
+      "grad_norm": 0.45756926721328306,
+      "learning_rate": 5.509240467961835e-06,
+      "loss": 0.1983,
+      "step": 4678
+    },
+    {
+      "epoch": 1.5625313073968943,
+      "grad_norm": 0.37815230836023156,
+      "learning_rate": 5.507307201627614e-06,
+      "loss": 0.1795,
+      "step": 4679
+    },
+    {
+      "epoch": 1.5628652529637668,
+      "grad_norm": 0.42395660449614353,
+      "learning_rate": 5.505373858658668e-06,
+      "loss": 0.2012,
+      "step": 4680
+    },
+    {
+      "epoch": 1.5631991985306395,
+      "grad_norm": 0.38723972997180905,
+      "learning_rate": 5.503440439347048e-06,
+      "loss": 0.186,
+      "step": 4681
+    },
+    {
+      "epoch": 1.5635331440975122,
+      "grad_norm": 0.3784601820153379,
+      "learning_rate": 5.501506943984823e-06,
+      "loss": 0.183,
+      "step": 4682
+    },
+    {
+      "epoch": 1.5638670896643847,
+      "grad_norm": 0.4147470025596834,
+      "learning_rate": 5.4995733728640695e-06,
+      "loss": 0.1917,
+      "step": 4683
+    },
+    {
+      "epoch": 1.5642010352312572,
+      "grad_norm": 0.43067484554889957,
+      "learning_rate": 5.497639726276876e-06,
+      "loss": 0.2089,
+      "step": 4684
+    },
+    {
+      "epoch": 1.56453498079813,
+      "grad_norm": 0.4656315754036739,
+      "learning_rate": 5.49570600451534e-06,
+      "loss": 0.2105,
+      "step": 4685
+    },
+    {
+      "epoch": 1.5648689263650026,
+      "grad_norm": 0.4202462754676753,
+      "learning_rate": 5.493772207871577e-06,
+      "loss": 0.2008,
+      "step": 4686
+    },
+    {
+      "epoch": 1.5652028719318751,
+      "grad_norm": 0.45653138426163725,
+      "learning_rate": 5.491838336637708e-06,
+      "loss": 0.2045,
+      "step": 4687
+    },
+    {
+      "epoch": 1.5655368174987476,
+      "grad_norm": 0.41926970695465754,
+      "learning_rate": 5.4899043911058665e-06,
+      "loss": 0.202,
+      "step": 4688
+    },
+    {
+      "epoch": 1.5658707630656203,
+      "grad_norm": 0.41411270997069943,
+      "learning_rate": 5.487970371568199e-06,
+      "loss": 0.1866,
+      "step": 4689
+    },
+    {
+      "epoch": 1.566204708632493,
+      "grad_norm": 0.38632209581344107,
+      "learning_rate": 5.486036278316861e-06,
+      "loss": 0.1826,
+      "step": 4690
+    },
+    {
+      "epoch": 1.5665386541993656,
+      "grad_norm": 0.3689564655199529,
+      "learning_rate": 5.48410211164402e-06,
+      "loss": 0.1834,
+      "step": 4691
+    },
+    {
+      "epoch": 1.566872599766238,
+      "grad_norm": 0.4331808630136572,
+      "learning_rate": 5.482167871841855e-06,
+      "loss": 0.2066,
+      "step": 4692
+    },
+    {
+      "epoch": 1.5672065453331108,
+      "grad_norm": 0.41532024396149275,
+      "learning_rate": 5.480233559202556e-06,
+      "loss": 0.2061,
+      "step": 4693
+    },
+    {
+      "epoch": 1.5675404908999833,
+      "grad_norm": 0.41478007296781094,
+      "learning_rate": 5.4782991740183225e-06,
+      "loss": 0.2085,
+      "step": 4694
+    },
+    {
+      "epoch": 1.5678744364668558,
+      "grad_norm": 0.39634309936244644,
+      "learning_rate": 5.476364716581367e-06,
+      "loss": 0.1963,
+      "step": 4695
+    },
+    {
+      "epoch": 1.5682083820337285,
+      "grad_norm": 0.38011643978852666,
+      "learning_rate": 5.474430187183912e-06,
+      "loss": 0.1885,
+      "step": 4696
+    },
+    {
+      "epoch": 1.5685423276006012,
+      "grad_norm": 0.4050712387842652,
+      "learning_rate": 5.472495586118192e-06,
+      "loss": 0.2043,
+      "step": 4697
+    },
+    {
+      "epoch": 1.5688762731674737,
+      "grad_norm": 0.398212340908208,
+      "learning_rate": 5.47056091367645e-06,
+      "loss": 0.1949,
+      "step": 4698
+    },
+    {
+      "epoch": 1.5692102187343462,
+      "grad_norm": 0.43479513365376116,
+      "learning_rate": 5.468626170150942e-06,
+      "loss": 0.2019,
+      "step": 4699
+    },
+    {
+      "epoch": 1.569544164301219,
+      "grad_norm": 0.4110382738577834,
+      "learning_rate": 5.466691355833932e-06,
+      "loss": 0.189,
+      "step": 4700
+    },
+    {
+      "epoch": 1.5698781098680916,
+      "grad_norm": 0.4152423718197278,
+      "learning_rate": 5.464756471017696e-06,
+      "loss": 0.2,
+      "step": 4701
+    },
+    {
+      "epoch": 1.5702120554349641,
+      "grad_norm": 0.37504987375871496,
+      "learning_rate": 5.462821515994525e-06,
+      "loss": 0.1847,
+      "step": 4702
+    },
+    {
+      "epoch": 1.5705460010018366,
+      "grad_norm": 0.4007750379597849,
+      "learning_rate": 5.460886491056714e-06,
+      "loss": 0.1973,
+      "step": 4703
+    },
+    {
+      "epoch": 1.5708799465687093,
+      "grad_norm": 0.4162908356303437,
+      "learning_rate": 5.458951396496572e-06,
+      "loss": 0.2064,
+      "step": 4704
+    },
+    {
+      "epoch": 1.571213892135582,
+      "grad_norm": 0.3842156235774017,
+      "learning_rate": 5.457016232606417e-06,
+      "loss": 0.1938,
+      "step": 4705
+    },
+    {
+      "epoch": 1.5715478377024545,
+      "grad_norm": 0.3956066511286018,
+      "learning_rate": 5.455080999678579e-06,
+      "loss": 0.1965,
+      "step": 4706
+    },
+    {
+      "epoch": 1.571881783269327,
+      "grad_norm": 0.42278335408976025,
+      "learning_rate": 5.453145698005399e-06,
+      "loss": 0.203,
+      "step": 4707
+    },
+    {
+      "epoch": 1.5722157288361998,
+      "grad_norm": 0.3980754778029337,
+      "learning_rate": 5.451210327879223e-06,
+      "loss": 0.1885,
+      "step": 4708
+    },
+    {
+      "epoch": 1.5725496744030723,
+      "grad_norm": 0.41626036909895414,
+      "learning_rate": 5.449274889592416e-06,
+      "loss": 0.1933,
+      "step": 4709
+    },
+    {
+      "epoch": 1.5728836199699447,
+      "grad_norm": 0.45739490381726283,
+      "learning_rate": 5.4473393834373466e-06,
+      "loss": 0.2051,
+      "step": 4710
+    },
+    {
+      "epoch": 1.5732175655368175,
+      "grad_norm": 0.40969808362672155,
+      "learning_rate": 5.445403809706395e-06,
+      "loss": 0.1942,
+      "step": 4711
+    },
+    {
+      "epoch": 1.5735515111036902,
+      "grad_norm": 0.38965421516201565,
+      "learning_rate": 5.443468168691954e-06,
+      "loss": 0.1872,
+      "step": 4712
+    },
+    {
+      "epoch": 1.5738854566705627,
+      "grad_norm": 0.3802144601432017,
+      "learning_rate": 5.441532460686426e-06,
+      "loss": 0.1936,
+      "step": 4713
+    },
+    {
+      "epoch": 1.5742194022374352,
+      "grad_norm": 0.4282184037870892,
+      "learning_rate": 5.4395966859822195e-06,
+      "loss": 0.2006,
+      "step": 4714
+    },
+    {
+      "epoch": 1.574553347804308,
+      "grad_norm": 0.41529391109227304,
+      "learning_rate": 5.437660844871758e-06,
+      "loss": 0.2018,
+      "step": 4715
+    },
+    {
+      "epoch": 1.5748872933711806,
+      "grad_norm": 0.43028826862778147,
+      "learning_rate": 5.435724937647473e-06,
+      "loss": 0.2069,
+      "step": 4716
+    },
+    {
+      "epoch": 1.575221238938053,
+      "grad_norm": 0.42632722327309375,
+      "learning_rate": 5.433788964601804e-06,
+      "loss": 0.2035,
+      "step": 4717
+    },
+    {
+      "epoch": 1.5755551845049256,
+      "grad_norm": 0.42219790114922706,
+      "learning_rate": 5.431852926027206e-06,
+      "loss": 0.1933,
+      "step": 4718
+    },
+    {
+      "epoch": 1.5758891300717983,
+      "grad_norm": 0.41389183867680046,
+      "learning_rate": 5.429916822216138e-06,
+      "loss": 0.1997,
+      "step": 4719
+    },
+    {
+      "epoch": 1.576223075638671,
+      "grad_norm": 0.42005326382754093,
+      "learning_rate": 5.42798065346107e-06,
+      "loss": 0.1881,
+      "step": 4720
+    },
+    {
+      "epoch": 1.5765570212055435,
+      "grad_norm": 0.39753444162905593,
+      "learning_rate": 5.426044420054488e-06,
+      "loss": 0.1929,
+      "step": 4721
+    },
+    {
+      "epoch": 1.576890966772416,
+      "grad_norm": 0.41691063344852947,
+      "learning_rate": 5.424108122288878e-06,
+      "loss": 0.2058,
+      "step": 4722
+    },
+    {
+      "epoch": 1.5772249123392887,
+      "grad_norm": 0.4523145688376783,
+      "learning_rate": 5.4221717604567435e-06,
+      "loss": 0.2156,
+      "step": 4723
+    },
+    {
+      "epoch": 1.5775588579061615,
+      "grad_norm": 0.39679608858354004,
+      "learning_rate": 5.420235334850593e-06,
+      "loss": 0.1923,
+      "step": 4724
+    },
+    {
+      "epoch": 1.5778928034730337,
+      "grad_norm": 0.3958825165224562,
+      "learning_rate": 5.418298845762947e-06,
+      "loss": 0.1885,
+      "step": 4725
+    },
+    {
+      "epoch": 1.5782267490399065,
+      "grad_norm": 0.42180340518034565,
+      "learning_rate": 5.416362293486336e-06,
+      "loss": 0.1985,
+      "step": 4726
+    },
+    {
+      "epoch": 1.5785606946067792,
+      "grad_norm": 0.43679318826263547,
+      "learning_rate": 5.4144256783132975e-06,
+      "loss": 0.2089,
+      "step": 4727
+    },
+    {
+      "epoch": 1.5788946401736517,
+      "grad_norm": 0.38015308316695867,
+      "learning_rate": 5.41248900053638e-06,
+      "loss": 0.1811,
+      "step": 4728
+    },
+    {
+      "epoch": 1.5792285857405242,
+      "grad_norm": 0.41432535867088516,
+      "learning_rate": 5.4105522604481435e-06,
+      "loss": 0.1965,
+      "step": 4729
+    },
+    {
+      "epoch": 1.5795625313073969,
+      "grad_norm": 0.36628677647353347,
+      "learning_rate": 5.408615458341152e-06,
+      "loss": 0.1796,
+      "step": 4730
+    },
+    {
+      "epoch": 1.5798964768742696,
+      "grad_norm": 0.3869654366325382,
+      "learning_rate": 5.4066785945079855e-06,
+      "loss": 0.1973,
+      "step": 4731
+    },
+    {
+      "epoch": 1.580230422441142,
+      "grad_norm": 0.3878145330599434,
+      "learning_rate": 5.404741669241228e-06,
+      "loss": 0.1883,
+      "step": 4732
+    },
+    {
+      "epoch": 1.5805643680080146,
+      "grad_norm": 0.4021115624829742,
+      "learning_rate": 5.402804682833477e-06,
+      "loss": 0.193,
+      "step": 4733
+    },
+    {
+      "epoch": 1.5808983135748873,
+      "grad_norm": 0.49296268950806166,
+      "learning_rate": 5.400867635577335e-06,
+      "loss": 0.1977,
+      "step": 4734
+    },
+    {
+      "epoch": 1.58123225914176,
+      "grad_norm": 0.4345599581839175,
+      "learning_rate": 5.398930527765416e-06,
+      "loss": 0.1982,
+      "step": 4735
+    },
+    {
+      "epoch": 1.5815662047086325,
+      "grad_norm": 0.413964619926734,
+      "learning_rate": 5.396993359690345e-06,
+      "loss": 0.2034,
+      "step": 4736
+    },
+    {
+      "epoch": 1.581900150275505,
+      "grad_norm": 0.43552047249198406,
+      "learning_rate": 5.395056131644752e-06,
+      "loss": 0.1958,
+      "step": 4737
+    },
+    {
+      "epoch": 1.5822340958423777,
+      "grad_norm": 0.4109376138019141,
+      "learning_rate": 5.393118843921277e-06,
+      "loss": 0.1922,
+      "step": 4738
+    },
+    {
+      "epoch": 1.5825680414092504,
+      "grad_norm": 0.5086676990134947,
+      "learning_rate": 5.391181496812573e-06,
+      "loss": 0.1883,
+      "step": 4739
+    },
+    {
+      "epoch": 1.582901986976123,
+      "grad_norm": 0.39573926156137973,
+      "learning_rate": 5.389244090611298e-06,
+      "loss": 0.1994,
+      "step": 4740
+    },
+    {
+      "epoch": 1.5832359325429954,
+      "grad_norm": 0.4607457839079693,
+      "learning_rate": 5.38730662561012e-06,
+      "loss": 0.205,
+      "step": 4741
+    },
+    {
+      "epoch": 1.5835698781098682,
+      "grad_norm": 0.40835357222366736,
+      "learning_rate": 5.385369102101716e-06,
+      "loss": 0.1872,
+      "step": 4742
+    },
+    {
+      "epoch": 1.5839038236767407,
+      "grad_norm": 0.48689078188687296,
+      "learning_rate": 5.38343152037877e-06,
+      "loss": 0.1961,
+      "step": 4743
+    },
+    {
+      "epoch": 1.5842377692436131,
+      "grad_norm": 0.4066030628039563,
+      "learning_rate": 5.38149388073398e-06,
+      "loss": 0.1945,
+      "step": 4744
+    },
+    {
+      "epoch": 1.5845717148104859,
+      "grad_norm": 0.366013738429132,
+      "learning_rate": 5.379556183460047e-06,
+      "loss": 0.1856,
+      "step": 4745
+    },
+    {
+      "epoch": 1.5849056603773586,
+      "grad_norm": 0.43200895906800973,
+      "learning_rate": 5.377618428849683e-06,
+      "loss": 0.2068,
+      "step": 4746
+    },
+    {
+      "epoch": 1.585239605944231,
+      "grad_norm": 0.49242292623459655,
+      "learning_rate": 5.375680617195609e-06,
+      "loss": 0.2096,
+      "step": 4747
+    },
+    {
+      "epoch": 1.5855735515111036,
+      "grad_norm": 0.4132744338380143,
+      "learning_rate": 5.373742748790555e-06,
+      "loss": 0.1864,
+      "step": 4748
+    },
+    {
+      "epoch": 1.5859074970779763,
+      "grad_norm": 0.39706760556922854,
+      "learning_rate": 5.371804823927258e-06,
+      "loss": 0.1926,
+      "step": 4749
+    },
+    {
+      "epoch": 1.586241442644849,
+      "grad_norm": 0.4238854670573685,
+      "learning_rate": 5.369866842898465e-06,
+      "loss": 0.2077,
+      "step": 4750
+    },
+    {
+      "epoch": 1.5865753882117215,
+      "grad_norm": 0.42640493614431735,
+      "learning_rate": 5.367928805996929e-06,
+      "loss": 0.1972,
+      "step": 4751
+    },
+    {
+      "epoch": 1.586909333778594,
+      "grad_norm": 0.37013179888742664,
+      "learning_rate": 5.365990713515414e-06,
+      "loss": 0.1729,
+      "step": 4752
+    },
+    {
+      "epoch": 1.5872432793454667,
+      "grad_norm": 0.45907590421110495,
+      "learning_rate": 5.364052565746693e-06,
+      "loss": 0.2073,
+      "step": 4753
+    },
+    {
+      "epoch": 1.5875772249123394,
+      "grad_norm": 0.43304177739828936,
+      "learning_rate": 5.362114362983547e-06,
+      "loss": 0.2037,
+      "step": 4754
+    },
+    {
+      "epoch": 1.587911170479212,
+      "grad_norm": 0.43747626863293665,
+      "learning_rate": 5.360176105518761e-06,
+      "loss": 0.1922,
+      "step": 4755
+    },
+    {
+      "epoch": 1.5882451160460844,
+      "grad_norm": 0.5704783919256339,
+      "learning_rate": 5.358237793645133e-06,
+      "loss": 0.191,
+      "step": 4756
+    },
+    {
+      "epoch": 1.5885790616129571,
+      "grad_norm": 0.41877733795764044,
+      "learning_rate": 5.356299427655469e-06,
+      "loss": 0.1945,
+      "step": 4757
+    },
+    {
+      "epoch": 1.5889130071798296,
+      "grad_norm": 0.4566114150645618,
+      "learning_rate": 5.354361007842581e-06,
+      "loss": 0.1982,
+      "step": 4758
+    },
+    {
+      "epoch": 1.5892469527467021,
+      "grad_norm": 0.38852790236519913,
+      "learning_rate": 5.352422534499291e-06,
+      "loss": 0.1853,
+      "step": 4759
+    },
+    {
+      "epoch": 1.5895808983135749,
+      "grad_norm": 0.40708531373941814,
+      "learning_rate": 5.350484007918428e-06,
+      "loss": 0.2064,
+      "step": 4760
+    },
+    {
+      "epoch": 1.5899148438804476,
+      "grad_norm": 0.4253185652044943,
+      "learning_rate": 5.3485454283928265e-06,
+      "loss": 0.192,
+      "step": 4761
+    },
+    {
+      "epoch": 1.59024878944732,
+      "grad_norm": 0.41539393746379444,
+      "learning_rate": 5.346606796215335e-06,
+      "loss": 0.1835,
+      "step": 4762
+    },
+    {
+      "epoch": 1.5905827350141926,
+      "grad_norm": 0.41820448778568065,
+      "learning_rate": 5.344668111678805e-06,
+      "loss": 0.2028,
+      "step": 4763
+    },
+    {
+      "epoch": 1.5909166805810653,
+      "grad_norm": 0.41281987451744845,
+      "learning_rate": 5.3427293750761e-06,
+      "loss": 0.2046,
+      "step": 4764
+    },
+    {
+      "epoch": 1.591250626147938,
+      "grad_norm": 0.49736587210070504,
+      "learning_rate": 5.340790586700086e-06,
+      "loss": 0.2222,
+      "step": 4765
+    },
+    {
+      "epoch": 1.5915845717148105,
+      "grad_norm": 0.46561471285392064,
+      "learning_rate": 5.338851746843643e-06,
+      "loss": 0.2092,
+      "step": 4766
+    },
+    {
+      "epoch": 1.591918517281683,
+      "grad_norm": 0.4281307653749809,
+      "learning_rate": 5.336912855799652e-06,
+      "loss": 0.2,
+      "step": 4767
+    },
+    {
+      "epoch": 1.5922524628485557,
+      "grad_norm": 0.39860709481846124,
+      "learning_rate": 5.334973913861008e-06,
+      "loss": 0.1945,
+      "step": 4768
+    },
+    {
+      "epoch": 1.5925864084154284,
+      "grad_norm": 0.4217558786649774,
+      "learning_rate": 5.33303492132061e-06,
+      "loss": 0.196,
+      "step": 4769
+    },
+    {
+      "epoch": 1.592920353982301,
+      "grad_norm": 0.4309860097814307,
+      "learning_rate": 5.3310958784713655e-06,
+      "loss": 0.1979,
+      "step": 4770
+    },
+    {
+      "epoch": 1.5932542995491734,
+      "grad_norm": 0.42902743725005,
+      "learning_rate": 5.329156785606191e-06,
+      "loss": 0.2034,
+      "step": 4771
+    },
+    {
+      "epoch": 1.5935882451160461,
+      "grad_norm": 0.38985978116145,
+      "learning_rate": 5.327217643018008e-06,
+      "loss": 0.2006,
+      "step": 4772
+    },
+    {
+      "epoch": 1.5939221906829188,
+      "grad_norm": 0.4347966967493165,
+      "learning_rate": 5.325278450999747e-06,
+      "loss": 0.2114,
+      "step": 4773
+    },
+    {
+      "epoch": 1.5942561362497911,
+      "grad_norm": 0.39474589860208664,
+      "learning_rate": 5.323339209844346e-06,
+      "loss": 0.1921,
+      "step": 4774
+    },
+    {
+      "epoch": 1.5945900818166638,
+      "grad_norm": 0.38256053396845374,
+      "learning_rate": 5.32139991984475e-06,
+      "loss": 0.1939,
+      "step": 4775
+    },
+    {
+      "epoch": 1.5949240273835366,
+      "grad_norm": 0.3729647811843787,
+      "learning_rate": 5.319460581293911e-06,
+      "loss": 0.1874,
+      "step": 4776
+    },
+    {
+      "epoch": 1.595257972950409,
+      "grad_norm": 0.45375740690251337,
+      "learning_rate": 5.317521194484791e-06,
+      "loss": 0.2095,
+      "step": 4777
+    },
+    {
+      "epoch": 1.5955919185172815,
+      "grad_norm": 0.3939370561880113,
+      "learning_rate": 5.315581759710356e-06,
+      "loss": 0.1933,
+      "step": 4778
+    },
+    {
+      "epoch": 1.5959258640841543,
+      "grad_norm": 0.41746052414225593,
+      "learning_rate": 5.313642277263577e-06,
+      "loss": 0.189,
+      "step": 4779
+    },
+    {
+      "epoch": 1.596259809651027,
+      "grad_norm": 0.4516550553035685,
+      "learning_rate": 5.311702747437443e-06,
+      "loss": 0.1925,
+      "step": 4780
+    },
+    {
+      "epoch": 1.5965937552178995,
+      "grad_norm": 0.41772684280109756,
+      "learning_rate": 5.309763170524937e-06,
+      "loss": 0.192,
+      "step": 4781
+    },
+    {
+      "epoch": 1.596927700784772,
+      "grad_norm": 0.41226393435164765,
+      "learning_rate": 5.307823546819056e-06,
+      "loss": 0.1922,
+      "step": 4782
+    },
+    {
+      "epoch": 1.5972616463516447,
+      "grad_norm": 0.39301899832094617,
+      "learning_rate": 5.305883876612805e-06,
+      "loss": 0.1864,
+      "step": 4783
+    },
+    {
+      "epoch": 1.5975955919185174,
+      "grad_norm": 0.4344890610361969,
+      "learning_rate": 5.303944160199193e-06,
+      "loss": 0.1933,
+      "step": 4784
+    },
+    {
+      "epoch": 1.59792953748539,
+      "grad_norm": 0.4204667147166973,
+      "learning_rate": 5.302004397871237e-06,
+      "loss": 0.2018,
+      "step": 4785
+    },
+    {
+      "epoch": 1.5982634830522624,
+      "grad_norm": 0.4025801727517493,
+      "learning_rate": 5.3000645899219594e-06,
+      "loss": 0.1944,
+      "step": 4786
+    },
+    {
+      "epoch": 1.5985974286191351,
+      "grad_norm": 0.4300003951448309,
+      "learning_rate": 5.298124736644392e-06,
+      "loss": 0.1997,
+      "step": 4787
+    },
+    {
+      "epoch": 1.5989313741860078,
+      "grad_norm": 0.4322841928213523,
+      "learning_rate": 5.296184838331575e-06,
+      "loss": 0.1934,
+      "step": 4788
+    },
+    {
+      "epoch": 1.5992653197528803,
+      "grad_norm": 0.4268719199609731,
+      "learning_rate": 5.2942448952765495e-06,
+      "loss": 0.1878,
+      "step": 4789
+    },
+    {
+      "epoch": 1.5995992653197528,
+      "grad_norm": 0.4096006704992976,
+      "learning_rate": 5.292304907772367e-06,
+      "loss": 0.1984,
+      "step": 4790
+    },
+    {
+      "epoch": 1.5999332108866255,
+      "grad_norm": 0.44146154991878817,
+      "learning_rate": 5.290364876112088e-06,
+      "loss": 0.2028,
+      "step": 4791
+    },
+    {
+      "epoch": 1.600267156453498,
+      "grad_norm": 0.4127132107934224,
+      "learning_rate": 5.288424800588775e-06,
+      "loss": 0.1985,
+      "step": 4792
+    },
+    {
+      "epoch": 1.6006011020203705,
+      "grad_norm": 0.4089302263887995,
+      "learning_rate": 5.2864846814955e-06,
+      "loss": 0.201,
+      "step": 4793
+    },
+    {
+      "epoch": 1.6009350475872433,
+      "grad_norm": 0.4057658128571867,
+      "learning_rate": 5.28454451912534e-06,
+      "loss": 0.1919,
+      "step": 4794
+    },
+    {
+      "epoch": 1.601268993154116,
+      "grad_norm": 0.41550864844996815,
+      "learning_rate": 5.28260431377138e-06,
+      "loss": 0.193,
+      "step": 4795
+    },
+    {
+      "epoch": 1.6016029387209885,
+      "grad_norm": 0.4214643164993601,
+      "learning_rate": 5.280664065726712e-06,
+      "loss": 0.2037,
+      "step": 4796
+    },
+    {
+      "epoch": 1.601936884287861,
+      "grad_norm": 0.41574763678512067,
+      "learning_rate": 5.278723775284432e-06,
+      "loss": 0.1955,
+      "step": 4797
+    },
+    {
+      "epoch": 1.6022708298547337,
+      "grad_norm": 0.4156915810039953,
+      "learning_rate": 5.276783442737642e-06,
+      "loss": 0.1951,
+      "step": 4798
+    },
+    {
+      "epoch": 1.6026047754216064,
+      "grad_norm": 0.4504952363847089,
+      "learning_rate": 5.274843068379456e-06,
+      "loss": 0.2121,
+      "step": 4799
+    },
+    {
+      "epoch": 1.602938720988479,
+      "grad_norm": 0.4141410647043094,
+      "learning_rate": 5.272902652502988e-06,
+      "loss": 0.1908,
+      "step": 4800
+    },
+    {
+      "epoch": 1.6032726665553514,
+      "grad_norm": 0.4172694674527439,
+      "learning_rate": 5.27096219540136e-06,
+      "loss": 0.2045,
+      "step": 4801
+    },
+    {
+      "epoch": 1.603606612122224,
+      "grad_norm": 0.39101651996056197,
+      "learning_rate": 5.269021697367702e-06,
+      "loss": 0.1979,
+      "step": 4802
+    },
+    {
+      "epoch": 1.6039405576890968,
+      "grad_norm": 0.4008081930642851,
+      "learning_rate": 5.26708115869515e-06,
+      "loss": 0.1828,
+      "step": 4803
+    },
+    {
+      "epoch": 1.6042745032559693,
+      "grad_norm": 0.40199523795553577,
+      "learning_rate": 5.265140579676844e-06,
+      "loss": 0.1946,
+      "step": 4804
+    },
+    {
+      "epoch": 1.6046084488228418,
+      "grad_norm": 0.36475147506022104,
+      "learning_rate": 5.263199960605931e-06,
+      "loss": 0.186,
+      "step": 4805
+    },
+    {
+      "epoch": 1.6049423943897145,
+      "grad_norm": 0.44832636055257613,
+      "learning_rate": 5.261259301775564e-06,
+      "loss": 0.2019,
+      "step": 4806
+    },
+    {
+      "epoch": 1.605276339956587,
+      "grad_norm": 0.3988676531453092,
+      "learning_rate": 5.259318603478904e-06,
+      "loss": 0.1845,
+      "step": 4807
+    },
+    {
+      "epoch": 1.6056102855234595,
+      "grad_norm": 0.4394954645416581,
+      "learning_rate": 5.2573778660091156e-06,
+      "loss": 0.2012,
+      "step": 4808
+    },
+    {
+      "epoch": 1.6059442310903322,
+      "grad_norm": 0.4090850456456306,
+      "learning_rate": 5.255437089659371e-06,
+      "loss": 0.1831,
+      "step": 4809
+    },
+    {
+      "epoch": 1.606278176657205,
+      "grad_norm": 0.4802815380853867,
+      "learning_rate": 5.253496274722846e-06,
+      "loss": 0.1978,
+      "step": 4810
+    },
+    {
+      "epoch": 1.6066121222240775,
+      "grad_norm": 0.38421296402882255,
+      "learning_rate": 5.251555421492722e-06,
+      "loss": 0.1815,
+      "step": 4811
+    },
+    {
+      "epoch": 1.60694606779095,
+      "grad_norm": 0.40180199874469,
+      "learning_rate": 5.249614530262191e-06,
+      "loss": 0.1883,
+      "step": 4812
+    },
+    {
+      "epoch": 1.6072800133578227,
+      "grad_norm": 0.3883383467412946,
+      "learning_rate": 5.2476736013244475e-06,
+      "loss": 0.1819,
+      "step": 4813
+    },
+    {
+      "epoch": 1.6076139589246954,
+      "grad_norm": 0.4086793201262821,
+      "learning_rate": 5.245732634972688e-06,
+      "loss": 0.1958,
+      "step": 4814
+    },
+    {
+      "epoch": 1.6079479044915679,
+      "grad_norm": 0.4224168613202054,
+      "learning_rate": 5.243791631500122e-06,
+      "loss": 0.1926,
+      "step": 4815
+    },
+    {
+      "epoch": 1.6082818500584404,
+      "grad_norm": 0.42613150503363995,
+      "learning_rate": 5.24185059119996e-06,
+      "loss": 0.1975,
+      "step": 4816
+    },
+    {
+      "epoch": 1.608615795625313,
+      "grad_norm": 0.4415148575023946,
+      "learning_rate": 5.239909514365415e-06,
+      "loss": 0.2004,
+      "step": 4817
+    },
+    {
+      "epoch": 1.6089497411921858,
+      "grad_norm": 0.40970747948408875,
+      "learning_rate": 5.237968401289717e-06,
+      "loss": 0.1971,
+      "step": 4818
+    },
+    {
+      "epoch": 1.6092836867590583,
+      "grad_norm": 0.4112895680407661,
+      "learning_rate": 5.236027252266088e-06,
+      "loss": 0.1972,
+      "step": 4819
+    },
+    {
+      "epoch": 1.6096176323259308,
+      "grad_norm": 0.40428437979576776,
+      "learning_rate": 5.234086067587765e-06,
+      "loss": 0.1942,
+      "step": 4820
+    },
+    {
+      "epoch": 1.6099515778928035,
+      "grad_norm": 0.419338867445877,
+      "learning_rate": 5.232144847547983e-06,
+      "loss": 0.1785,
+      "step": 4821
+    },
+    {
+      "epoch": 1.6102855234596762,
+      "grad_norm": 0.4699954998213189,
+      "learning_rate": 5.230203592439989e-06,
+      "loss": 0.2027,
+      "step": 4822
+    },
+    {
+      "epoch": 1.6106194690265485,
+      "grad_norm": 0.38579142330375554,
+      "learning_rate": 5.228262302557034e-06,
+      "loss": 0.1933,
+      "step": 4823
+    },
+    {
+      "epoch": 1.6109534145934212,
+      "grad_norm": 0.44120355162315,
+      "learning_rate": 5.226320978192369e-06,
+      "loss": 0.2033,
+      "step": 4824
+    },
+    {
+      "epoch": 1.611287360160294,
+      "grad_norm": 0.40448539512862475,
+      "learning_rate": 5.224379619639253e-06,
+      "loss": 0.1851,
+      "step": 4825
+    },
+    {
+      "epoch": 1.6116213057271664,
+      "grad_norm": 0.4454285188019959,
+      "learning_rate": 5.222438227190957e-06,
+      "loss": 0.2021,
+      "step": 4826
+    },
+    {
+      "epoch": 1.611955251294039,
+      "grad_norm": 0.414760600019116,
+      "learning_rate": 5.220496801140746e-06,
+      "loss": 0.1996,
+      "step": 4827
+    },
+    {
+      "epoch": 1.6122891968609117,
+      "grad_norm": 0.42718840912720824,
+      "learning_rate": 5.218555341781897e-06,
+      "loss": 0.189,
+      "step": 4828
+    },
+    {
+      "epoch": 1.6126231424277844,
+      "grad_norm": 0.4267525421289148,
+      "learning_rate": 5.216613849407691e-06,
+      "loss": 0.197,
+      "step": 4829
+    },
+    {
+      "epoch": 1.6129570879946569,
+      "grad_norm": 0.4493424424103547,
+      "learning_rate": 5.214672324311412e-06,
+      "loss": 0.1982,
+      "step": 4830
+    },
+    {
+      "epoch": 1.6132910335615294,
+      "grad_norm": 0.4308674096556119,
+      "learning_rate": 5.21273076678635e-06,
+      "loss": 0.1943,
+      "step": 4831
+    },
+    {
+      "epoch": 1.613624979128402,
+      "grad_norm": 0.3925998201907294,
+      "learning_rate": 5.210789177125802e-06,
+      "loss": 0.1881,
+      "step": 4832
+    },
+    {
+      "epoch": 1.6139589246952748,
+      "grad_norm": 0.4254383389522204,
+      "learning_rate": 5.208847555623066e-06,
+      "loss": 0.195,
+      "step": 4833
+    },
+    {
+      "epoch": 1.6142928702621473,
+      "grad_norm": 0.3784003115153938,
+      "learning_rate": 5.206905902571447e-06,
+      "loss": 0.1845,
+      "step": 4834
+    },
+    {
+      "epoch": 1.6146268158290198,
+      "grad_norm": 0.4259770207037785,
+      "learning_rate": 5.204964218264258e-06,
+      "loss": 0.1833,
+      "step": 4835
+    },
+    {
+      "epoch": 1.6149607613958925,
+      "grad_norm": 0.39738746452523543,
+      "learning_rate": 5.203022502994808e-06,
+      "loss": 0.1976,
+      "step": 4836
+    },
+    {
+      "epoch": 1.6152947069627652,
+      "grad_norm": 0.39030271717351783,
+      "learning_rate": 5.201080757056418e-06,
+      "loss": 0.1999,
+      "step": 4837
+    },
+    {
+      "epoch": 1.6156286525296377,
+      "grad_norm": 0.4263335380461042,
+      "learning_rate": 5.1991389807424145e-06,
+      "loss": 0.2086,
+      "step": 4838
+    },
+    {
+      "epoch": 1.6159625980965102,
+      "grad_norm": 0.398745944842516,
+      "learning_rate": 5.1971971743461215e-06,
+      "loss": 0.1932,
+      "step": 4839
+    },
+    {
+      "epoch": 1.616296543663383,
+      "grad_norm": 0.42880316783788475,
+      "learning_rate": 5.195255338160873e-06,
+      "loss": 0.2033,
+      "step": 4840
+    },
+    {
+      "epoch": 1.6166304892302554,
+      "grad_norm": 0.4503242539328835,
+      "learning_rate": 5.193313472480007e-06,
+      "loss": 0.216,
+      "step": 4841
+    },
+    {
+      "epoch": 1.616964434797128,
+      "grad_norm": 0.40831941532509597,
+      "learning_rate": 5.191371577596866e-06,
+      "loss": 0.1872,
+      "step": 4842
+    },
+    {
+      "epoch": 1.6172983803640006,
+      "grad_norm": 0.41791124388464684,
+      "learning_rate": 5.189429653804794e-06,
+      "loss": 0.194,
+      "step": 4843
+    },
+    {
+      "epoch": 1.6176323259308734,
+      "grad_norm": 0.4292909114375624,
+      "learning_rate": 5.187487701397142e-06,
+      "loss": 0.1999,
+      "step": 4844
+    },
+    {
+      "epoch": 1.6179662714977459,
+      "grad_norm": 0.4136962868331666,
+      "learning_rate": 5.185545720667266e-06,
+      "loss": 0.1884,
+      "step": 4845
+    },
+    {
+      "epoch": 1.6183002170646184,
+      "grad_norm": 0.39331667168777334,
+      "learning_rate": 5.183603711908523e-06,
+      "loss": 0.1932,
+      "step": 4846
+    },
+    {
+      "epoch": 1.618634162631491,
+      "grad_norm": 0.380147314511648,
+      "learning_rate": 5.181661675414278e-06,
+      "loss": 0.1895,
+      "step": 4847
+    },
+    {
+      "epoch": 1.6189681081983638,
+      "grad_norm": 0.4066449462044048,
+      "learning_rate": 5.179719611477898e-06,
+      "loss": 0.1937,
+      "step": 4848
+    },
+    {
+      "epoch": 1.6193020537652363,
+      "grad_norm": 0.4258446181103049,
+      "learning_rate": 5.1777775203927535e-06,
+      "loss": 0.2068,
+      "step": 4849
+    },
+    {
+      "epoch": 1.6196359993321088,
+      "grad_norm": 0.4113383272205944,
+      "learning_rate": 5.175835402452223e-06,
+      "loss": 0.2044,
+      "step": 4850
+    },
+    {
+      "epoch": 1.6199699448989815,
+      "grad_norm": 0.413380097915521,
+      "learning_rate": 5.173893257949683e-06,
+      "loss": 0.2014,
+      "step": 4851
+    },
+    {
+      "epoch": 1.6203038904658542,
+      "grad_norm": 0.42517246513422285,
+      "learning_rate": 5.17195108717852e-06,
+      "loss": 0.2054,
+      "step": 4852
+    },
+    {
+      "epoch": 1.6206378360327267,
+      "grad_norm": 0.41524625901706064,
+      "learning_rate": 5.170008890432121e-06,
+      "loss": 0.1873,
+      "step": 4853
+    },
+    {
+      "epoch": 1.6209717815995992,
+      "grad_norm": 0.43808341311644106,
+      "learning_rate": 5.168066668003876e-06,
+      "loss": 0.2091,
+      "step": 4854
+    },
+    {
+      "epoch": 1.621305727166472,
+      "grad_norm": 0.4252977935303211,
+      "learning_rate": 5.166124420187182e-06,
+      "loss": 0.2007,
+      "step": 4855
+    },
+    {
+      "epoch": 1.6216396727333444,
+      "grad_norm": 0.4319004455143065,
+      "learning_rate": 5.164182147275439e-06,
+      "loss": 0.2144,
+      "step": 4856
+    },
+    {
+      "epoch": 1.621973618300217,
+      "grad_norm": 0.40036208771754045,
+      "learning_rate": 5.16223984956205e-06,
+      "loss": 0.1875,
+      "step": 4857
+    },
+    {
+      "epoch": 1.6223075638670896,
+      "grad_norm": 0.418519003967928,
+      "learning_rate": 5.1602975273404196e-06,
+      "loss": 0.1976,
+      "step": 4858
+    },
+    {
+      "epoch": 1.6226415094339623,
+      "grad_norm": 0.40350253708517614,
+      "learning_rate": 5.158355180903961e-06,
+      "loss": 0.1989,
+      "step": 4859
+    },
+    {
+      "epoch": 1.6229754550008348,
+      "grad_norm": 0.41204369381140116,
+      "learning_rate": 5.156412810546089e-06,
+      "loss": 0.1949,
+      "step": 4860
+    },
+    {
+      "epoch": 1.6233094005677073,
+      "grad_norm": 0.4382744192834387,
+      "learning_rate": 5.154470416560219e-06,
+      "loss": 0.2089,
+      "step": 4861
+    },
+    {
+      "epoch": 1.62364334613458,
+      "grad_norm": 0.40586800209662394,
+      "learning_rate": 5.152527999239774e-06,
+      "loss": 0.1748,
+      "step": 4862
+    },
+    {
+      "epoch": 1.6239772917014528,
+      "grad_norm": 0.43817177004856733,
+      "learning_rate": 5.150585558878177e-06,
+      "loss": 0.2037,
+      "step": 4863
+    },
+    {
+      "epoch": 1.6243112372683253,
+      "grad_norm": 0.40381422303951253,
+      "learning_rate": 5.148643095768861e-06,
+      "loss": 0.1881,
+      "step": 4864
+    },
+    {
+      "epoch": 1.6246451828351978,
+      "grad_norm": 0.4223072439345843,
+      "learning_rate": 5.146700610205254e-06,
+      "loss": 0.2005,
+      "step": 4865
+    },
+    {
+      "epoch": 1.6249791284020705,
+      "grad_norm": 0.39611701923208614,
+      "learning_rate": 5.144758102480792e-06,
+      "loss": 0.1915,
+      "step": 4866
+    },
+    {
+      "epoch": 1.6253130739689432,
+      "grad_norm": 0.39897896763142227,
+      "learning_rate": 5.142815572888915e-06,
+      "loss": 0.1801,
+      "step": 4867
+    },
+    {
+      "epoch": 1.6256470195358157,
+      "grad_norm": 0.3942908284104242,
+      "learning_rate": 5.140873021723065e-06,
+      "loss": 0.1873,
+      "step": 4868
+    },
+    {
+      "epoch": 1.6259809651026882,
+      "grad_norm": 0.39322301183965197,
+      "learning_rate": 5.138930449276686e-06,
+      "loss": 0.1992,
+      "step": 4869
+    },
+    {
+      "epoch": 1.626314910669561,
+      "grad_norm": 0.4303711635184881,
+      "learning_rate": 5.136987855843226e-06,
+      "loss": 0.1995,
+      "step": 4870
+    },
+    {
+      "epoch": 1.6266488562364336,
+      "grad_norm": 0.4345354327414109,
+      "learning_rate": 5.135045241716138e-06,
+      "loss": 0.2136,
+      "step": 4871
+    },
+    {
+      "epoch": 1.626982801803306,
+      "grad_norm": 0.539742638729196,
+      "learning_rate": 5.133102607188875e-06,
+      "loss": 0.2132,
+      "step": 4872
+    },
+    {
+      "epoch": 1.6273167473701786,
+      "grad_norm": 0.4040192184352221,
+      "learning_rate": 5.131159952554896e-06,
+      "loss": 0.1896,
+      "step": 4873
+    },
+    {
+      "epoch": 1.6276506929370513,
+      "grad_norm": 0.4169971044202786,
+      "learning_rate": 5.129217278107663e-06,
+      "loss": 0.194,
+      "step": 4874
+    },
+    {
+      "epoch": 1.6279846385039238,
+      "grad_norm": 0.40503849753645316,
+      "learning_rate": 5.127274584140636e-06,
+      "loss": 0.1903,
+      "step": 4875
+    },
+    {
+      "epoch": 1.6283185840707963,
+      "grad_norm": 0.4237099685305266,
+      "learning_rate": 5.125331870947287e-06,
+      "loss": 0.2008,
+      "step": 4876
+    },
+    {
+      "epoch": 1.628652529637669,
+      "grad_norm": 0.4142662739561068,
+      "learning_rate": 5.123389138821084e-06,
+      "loss": 0.1995,
+      "step": 4877
+    },
+    {
+      "epoch": 1.6289864752045418,
+      "grad_norm": 0.45354343375917766,
+      "learning_rate": 5.121446388055497e-06,
+      "loss": 0.21,
+      "step": 4878
+    },
+    {
+      "epoch": 1.6293204207714143,
+      "grad_norm": 0.37425067796417305,
+      "learning_rate": 5.119503618944004e-06,
+      "loss": 0.1834,
+      "step": 4879
+    },
+    {
+      "epoch": 1.6296543663382868,
+      "grad_norm": 0.3788887242903905,
+      "learning_rate": 5.117560831780082e-06,
+      "loss": 0.1904,
+      "step": 4880
+    },
+    {
+      "epoch": 1.6299883119051595,
+      "grad_norm": 0.3898545052130911,
+      "learning_rate": 5.115618026857211e-06,
+      "loss": 0.1888,
+      "step": 4881
+    },
+    {
+      "epoch": 1.6303222574720322,
+      "grad_norm": 0.4105364824350985,
+      "learning_rate": 5.113675204468876e-06,
+      "loss": 0.1911,
+      "step": 4882
+    },
+    {
+      "epoch": 1.6306562030389047,
+      "grad_norm": 0.3962203176947265,
+      "learning_rate": 5.111732364908564e-06,
+      "loss": 0.1764,
+      "step": 4883
+    },
+    {
+      "epoch": 1.6309901486057772,
+      "grad_norm": 0.3507120420190793,
+      "learning_rate": 5.109789508469761e-06,
+      "loss": 0.1671,
+      "step": 4884
+    },
+    {
+      "epoch": 1.63132409417265,
+      "grad_norm": 0.3949491699443807,
+      "learning_rate": 5.107846635445962e-06,
+      "loss": 0.1958,
+      "step": 4885
+    },
+    {
+      "epoch": 1.6316580397395226,
+      "grad_norm": 0.44189276954645507,
+      "learning_rate": 5.1059037461306586e-06,
+      "loss": 0.2031,
+      "step": 4886
+    },
+    {
+      "epoch": 1.631991985306395,
+      "grad_norm": 0.42594578365840874,
+      "learning_rate": 5.103960840817346e-06,
+      "loss": 0.2029,
+      "step": 4887
+    },
+    {
+      "epoch": 1.6323259308732676,
+      "grad_norm": 0.4073041041534602,
+      "learning_rate": 5.1020179197995245e-06,
+      "loss": 0.1948,
+      "step": 4888
+    },
+    {
+      "epoch": 1.6326598764401403,
+      "grad_norm": 0.4197429915528991,
+      "learning_rate": 5.1000749833706964e-06,
+      "loss": 0.179,
+      "step": 4889
+    },
+    {
+      "epoch": 1.6329938220070128,
+      "grad_norm": 0.4231299956266433,
+      "learning_rate": 5.098132031824362e-06,
+      "loss": 0.2036,
+      "step": 4890
+    },
+    {
+      "epoch": 1.6333277675738853,
+      "grad_norm": 0.43474360913815285,
+      "learning_rate": 5.096189065454029e-06,
+      "loss": 0.2072,
+      "step": 4891
+    },
+    {
+      "epoch": 1.633661713140758,
+      "grad_norm": 0.42739722495251686,
+      "learning_rate": 5.094246084553206e-06,
+      "loss": 0.1998,
+      "step": 4892
+    },
+    {
+      "epoch": 1.6339956587076307,
+      "grad_norm": 0.4068442630703969,
+      "learning_rate": 5.092303089415403e-06,
+      "loss": 0.1965,
+      "step": 4893
+    },
+    {
+      "epoch": 1.6343296042745032,
+      "grad_norm": 0.42075102506762035,
+      "learning_rate": 5.09036008033413e-06,
+      "loss": 0.1958,
+      "step": 4894
+    },
+    {
+      "epoch": 1.6346635498413757,
+      "grad_norm": 0.4751211507782048,
+      "learning_rate": 5.0884170576029034e-06,
+      "loss": 0.2173,
+      "step": 4895
+    },
+    {
+      "epoch": 1.6349974954082485,
+      "grad_norm": 0.4414325188339375,
+      "learning_rate": 5.086474021515238e-06,
+      "loss": 0.2036,
+      "step": 4896
+    },
+    {
+      "epoch": 1.6353314409751212,
+      "grad_norm": 0.43981042916979307,
+      "learning_rate": 5.084530972364656e-06,
+      "loss": 0.1957,
+      "step": 4897
+    },
+    {
+      "epoch": 1.6356653865419937,
+      "grad_norm": 0.4447308583849977,
+      "learning_rate": 5.082587910444674e-06,
+      "loss": 0.1966,
+      "step": 4898
+    },
+    {
+      "epoch": 1.6359993321088662,
+      "grad_norm": 0.44669721483940267,
+      "learning_rate": 5.080644836048815e-06,
+      "loss": 0.2017,
+      "step": 4899
+    },
+    {
+      "epoch": 1.6363332776757389,
+      "grad_norm": 0.37911277492603973,
+      "learning_rate": 5.0787017494706035e-06,
+      "loss": 0.1885,
+      "step": 4900
+    },
+    {
+      "epoch": 1.6366672232426116,
+      "grad_norm": 0.42209764612725126,
+      "learning_rate": 5.076758651003567e-06,
+      "loss": 0.1904,
+      "step": 4901
+    },
+    {
+      "epoch": 1.637001168809484,
+      "grad_norm": 0.41618819886027797,
+      "learning_rate": 5.0748155409412325e-06,
+      "loss": 0.1971,
+      "step": 4902
+    },
+    {
+      "epoch": 1.6373351143763566,
+      "grad_norm": 0.37968963437992637,
+      "learning_rate": 5.0728724195771295e-06,
+      "loss": 0.1933,
+      "step": 4903
+    },
+    {
+      "epoch": 1.6376690599432293,
+      "grad_norm": 0.49016772674623965,
+      "learning_rate": 5.070929287204789e-06,
+      "loss": 0.2005,
+      "step": 4904
+    },
+    {
+      "epoch": 1.6380030055101018,
+      "grad_norm": 0.4179568582831211,
+      "learning_rate": 5.068986144117746e-06,
+      "loss": 0.2006,
+      "step": 4905
+    },
+    {
+      "epoch": 1.6383369510769743,
+      "grad_norm": 0.4255338314834005,
+      "learning_rate": 5.067042990609533e-06,
+      "loss": 0.194,
+      "step": 4906
+    },
+    {
+      "epoch": 1.638670896643847,
+      "grad_norm": 0.4489710292882698,
+      "learning_rate": 5.065099826973685e-06,
+      "loss": 0.2143,
+      "step": 4907
+    },
+    {
+      "epoch": 1.6390048422107197,
+      "grad_norm": 0.43882456276403864,
+      "learning_rate": 5.0631566535037435e-06,
+      "loss": 0.1977,
+      "step": 4908
+    },
+    {
+      "epoch": 1.6393387877775922,
+      "grad_norm": 0.38272437413780414,
+      "learning_rate": 5.061213470493246e-06,
+      "loss": 0.187,
+      "step": 4909
+    },
+    {
+      "epoch": 1.6396727333444647,
+      "grad_norm": 0.4223324994778571,
+      "learning_rate": 5.059270278235732e-06,
+      "loss": 0.1967,
+      "step": 4910
+    },
+    {
+      "epoch": 1.6400066789113374,
+      "grad_norm": 0.3976067064634602,
+      "learning_rate": 5.057327077024745e-06,
+      "loss": 0.1903,
+      "step": 4911
+    },
+    {
+      "epoch": 1.6403406244782102,
+      "grad_norm": 0.4209543105402407,
+      "learning_rate": 5.055383867153829e-06,
+      "loss": 0.2025,
+      "step": 4912
+    },
+    {
+      "epoch": 1.6406745700450827,
+      "grad_norm": 0.3947766936457843,
+      "learning_rate": 5.053440648916526e-06,
+      "loss": 0.1819,
+      "step": 4913
+    },
+    {
+      "epoch": 1.6410085156119552,
+      "grad_norm": 0.4026201573480306,
+      "learning_rate": 5.051497422606385e-06,
+      "loss": 0.1935,
+      "step": 4914
+    },
+    {
+      "epoch": 1.6413424611788279,
+      "grad_norm": 0.37785493785849833,
+      "learning_rate": 5.049554188516952e-06,
+      "loss": 0.1956,
+      "step": 4915
+    },
+    {
+      "epoch": 1.6416764067457006,
+      "grad_norm": 0.4080022366885912,
+      "learning_rate": 5.047610946941775e-06,
+      "loss": 0.1959,
+      "step": 4916
+    },
+    {
+      "epoch": 1.642010352312573,
+      "grad_norm": 0.39521347259779677,
+      "learning_rate": 5.045667698174403e-06,
+      "loss": 0.1841,
+      "step": 4917
+    },
+    {
+      "epoch": 1.6423442978794456,
+      "grad_norm": 0.4052000993449699,
+      "learning_rate": 5.043724442508388e-06,
+      "loss": 0.1873,
+      "step": 4918
+    },
+    {
+      "epoch": 1.6426782434463183,
+      "grad_norm": 0.4289851629916548,
+      "learning_rate": 5.0417811802372815e-06,
+      "loss": 0.2038,
+      "step": 4919
+    },
+    {
+      "epoch": 1.643012189013191,
+      "grad_norm": 0.42873602680133893,
+      "learning_rate": 5.039837911654637e-06,
+      "loss": 0.1924,
+      "step": 4920
+    },
+    {
+      "epoch": 1.6433461345800633,
+      "grad_norm": 0.4163011122982262,
+      "learning_rate": 5.037894637054005e-06,
+      "loss": 0.1887,
+      "step": 4921
+    },
+    {
+      "epoch": 1.643680080146936,
+      "grad_norm": 0.41256746568734526,
+      "learning_rate": 5.035951356728942e-06,
+      "loss": 0.1868,
+      "step": 4922
+    },
+    {
+      "epoch": 1.6440140257138087,
+      "grad_norm": 0.38452956569882735,
+      "learning_rate": 5.034008070973004e-06,
+      "loss": 0.1955,
+      "step": 4923
+    },
+    {
+      "epoch": 1.6443479712806812,
+      "grad_norm": 0.47266399939167203,
+      "learning_rate": 5.032064780079746e-06,
+      "loss": 0.2108,
+      "step": 4924
+    },
+    {
+      "epoch": 1.6446819168475537,
+      "grad_norm": 0.3661920769763338,
+      "learning_rate": 5.030121484342725e-06,
+      "loss": 0.1821,
+      "step": 4925
+    },
+    {
+      "epoch": 1.6450158624144264,
+      "grad_norm": 0.38288278251589564,
+      "learning_rate": 5.0281781840555e-06,
+      "loss": 0.1872,
+      "step": 4926
+    },
+    {
+      "epoch": 1.6453498079812992,
+      "grad_norm": 0.3951674051282928,
+      "learning_rate": 5.026234879511629e-06,
+      "loss": 0.1919,
+      "step": 4927
+    },
+    {
+      "epoch": 1.6456837535481716,
+      "grad_norm": 0.4446820667734045,
+      "learning_rate": 5.024291571004668e-06,
+      "loss": 0.2049,
+      "step": 4928
+    },
+    {
+      "epoch": 1.6460176991150441,
+      "grad_norm": 0.3847089437235357,
+      "learning_rate": 5.022348258828181e-06,
+      "loss": 0.1854,
+      "step": 4929
+    },
+    {
+      "epoch": 1.6463516446819169,
+      "grad_norm": 0.4465777995021089,
+      "learning_rate": 5.020404943275727e-06,
+      "loss": 0.2032,
+      "step": 4930
+    },
+    {
+      "epoch": 1.6466855902487896,
+      "grad_norm": 0.4209000333850444,
+      "learning_rate": 5.018461624640864e-06,
+      "loss": 0.1881,
+      "step": 4931
+    },
+    {
+      "epoch": 1.647019535815662,
+      "grad_norm": 0.43874264508637995,
+      "learning_rate": 5.016518303217157e-06,
+      "loss": 0.1959,
+      "step": 4932
+    },
+    {
+      "epoch": 1.6473534813825346,
+      "grad_norm": 0.40621965004843547,
+      "learning_rate": 5.014574979298166e-06,
+      "loss": 0.195,
+      "step": 4933
+    },
+    {
+      "epoch": 1.6476874269494073,
+      "grad_norm": 0.4112489831025004,
+      "learning_rate": 5.012631653177451e-06,
+      "loss": 0.1962,
+      "step": 4934
+    },
+    {
+      "epoch": 1.64802137251628,
+      "grad_norm": 0.38501149745631036,
+      "learning_rate": 5.010688325148577e-06,
+      "loss": 0.1827,
+      "step": 4935
+    },
+    {
+      "epoch": 1.6483553180831525,
+      "grad_norm": 0.382230170818548,
+      "learning_rate": 5.008744995505107e-06,
+      "loss": 0.1816,
+      "step": 4936
+    },
+    {
+      "epoch": 1.648689263650025,
+      "grad_norm": 0.4269969102329425,
+      "learning_rate": 5.0068016645406e-06,
+      "loss": 0.1912,
+      "step": 4937
+    },
+    {
+      "epoch": 1.6490232092168977,
+      "grad_norm": 0.3911253797830778,
+      "learning_rate": 5.0048583325486234e-06,
+      "loss": 0.1907,
+      "step": 4938
+    },
+    {
+      "epoch": 1.6493571547837702,
+      "grad_norm": 0.3973399530499921,
+      "learning_rate": 5.002914999822737e-06,
+      "loss": 0.1993,
+      "step": 4939
+    },
+    {
+      "epoch": 1.6496911003506427,
+      "grad_norm": 0.381805883336254,
+      "learning_rate": 5.000971666656508e-06,
+      "loss": 0.1866,
+      "step": 4940
+    },
+    {
+      "epoch": 1.6500250459175154,
+      "grad_norm": 0.44398075937862946,
+      "learning_rate": 4.999028333343494e-06,
+      "loss": 0.2153,
+      "step": 4941
+    },
+    {
+      "epoch": 1.6503589914843881,
+      "grad_norm": 0.40495075548352927,
+      "learning_rate": 4.9970850001772634e-06,
+      "loss": 0.2006,
+      "step": 4942
+    },
+    {
+      "epoch": 1.6506929370512606,
+      "grad_norm": 0.47323246207653963,
+      "learning_rate": 4.995141667451378e-06,
+      "loss": 0.2111,
+      "step": 4943
+    },
+    {
+      "epoch": 1.6510268826181331,
+      "grad_norm": 0.40689036499077386,
+      "learning_rate": 4.993198335459401e-06,
+      "loss": 0.1961,
+      "step": 4944
+    },
+    {
+      "epoch": 1.6513608281850058,
+      "grad_norm": 0.45363419907785657,
+      "learning_rate": 4.991255004494896e-06,
+      "loss": 0.2002,
+      "step": 4945
+    },
+    {
+      "epoch": 1.6516947737518786,
+      "grad_norm": 0.39079299321552946,
+      "learning_rate": 4.989311674851424e-06,
+      "loss": 0.1918,
+      "step": 4946
+    },
+    {
+      "epoch": 1.652028719318751,
+      "grad_norm": 0.4079537560804328,
+      "learning_rate": 4.9873683468225495e-06,
+      "loss": 0.1934,
+      "step": 4947
+    },
+    {
+      "epoch": 1.6523626648856236,
+      "grad_norm": 0.3895830948669782,
+      "learning_rate": 4.985425020701836e-06,
+      "loss": 0.1966,
+      "step": 4948
+    },
+    {
+      "epoch": 1.6526966104524963,
+      "grad_norm": 0.39586577460541733,
+      "learning_rate": 4.983481696782844e-06,
+      "loss": 0.2038,
+      "step": 4949
+    },
+    {
+      "epoch": 1.653030556019369,
+      "grad_norm": 0.4316130617272694,
+      "learning_rate": 4.9815383753591365e-06,
+      "loss": 0.2015,
+      "step": 4950
+    },
+    {
+      "epoch": 1.6533645015862415,
+      "grad_norm": 0.4491142890206108,
+      "learning_rate": 4.9795950567242754e-06,
+      "loss": 0.2137,
+      "step": 4951
+    },
+    {
+      "epoch": 1.653698447153114,
+      "grad_norm": 0.371221820736492,
+      "learning_rate": 4.9776517411718214e-06,
+      "loss": 0.1865,
+      "step": 4952
+    },
+    {
+      "epoch": 1.6540323927199867,
+      "grad_norm": 0.8306835153106146,
+      "learning_rate": 4.9757084289953325e-06,
+      "loss": 0.1926,
+      "step": 4953
+    },
+    {
+      "epoch": 1.6543663382868592,
+      "grad_norm": 0.4009920133756041,
+      "learning_rate": 4.973765120488373e-06,
+      "loss": 0.2038,
+      "step": 4954
+    },
+    {
+      "epoch": 1.6547002838537317,
+      "grad_norm": 0.4568623426752501,
+      "learning_rate": 4.9718218159445015e-06,
+      "loss": 0.1989,
+      "step": 4955
+    },
+    {
+      "epoch": 1.6550342294206044,
+      "grad_norm": 0.48262334533879275,
+      "learning_rate": 4.969878515657276e-06,
+      "loss": 0.1881,
+      "step": 4956
+    },
+    {
+      "epoch": 1.6553681749874771,
+      "grad_norm": 0.40522771927191203,
+      "learning_rate": 4.967935219920257e-06,
+      "loss": 0.1892,
+      "step": 4957
+    },
+    {
+      "epoch": 1.6557021205543496,
+      "grad_norm": 0.37521846679027776,
+      "learning_rate": 4.9659919290269986e-06,
+      "loss": 0.1844,
+      "step": 4958
+    },
+    {
+      "epoch": 1.6560360661212221,
+      "grad_norm": 0.4083183199052804,
+      "learning_rate": 4.964048643271058e-06,
+      "loss": 0.2098,
+      "step": 4959
+    },
+    {
+      "epoch": 1.6563700116880948,
+      "grad_norm": 0.4004642626569134,
+      "learning_rate": 4.962105362945996e-06,
+      "loss": 0.1891,
+      "step": 4960
+    },
+    {
+      "epoch": 1.6567039572549676,
+      "grad_norm": 0.40776367914606315,
+      "learning_rate": 4.960162088345365e-06,
+      "loss": 0.1911,
+      "step": 4961
+    },
+    {
+      "epoch": 1.65703790282184,
+      "grad_norm": 0.4170345590008578,
+      "learning_rate": 4.958218819762719e-06,
+      "loss": 0.1925,
+      "step": 4962
+    },
+    {
+      "epoch": 1.6573718483887125,
+      "grad_norm": 0.3933408020439283,
+      "learning_rate": 4.9562755574916125e-06,
+      "loss": 0.1841,
+      "step": 4963
+    },
+    {
+      "epoch": 1.6577057939555853,
+      "grad_norm": 0.38270569831932166,
+      "learning_rate": 4.954332301825597e-06,
+      "loss": 0.185,
+      "step": 4964
+    },
+    {
+      "epoch": 1.658039739522458,
+      "grad_norm": 0.4416397735911187,
+      "learning_rate": 4.952389053058226e-06,
+      "loss": 0.1998,
+      "step": 4965
+    },
+    {
+      "epoch": 1.6583736850893305,
+      "grad_norm": 0.4038672656766227,
+      "learning_rate": 4.95044581148305e-06,
+      "loss": 0.1876,
+      "step": 4966
+    },
+    {
+      "epoch": 1.658707630656203,
+      "grad_norm": 0.45268259547839906,
+      "learning_rate": 4.948502577393617e-06,
+      "loss": 0.214,
+      "step": 4967
+    },
+    {
+      "epoch": 1.6590415762230757,
+      "grad_norm": 0.40485817055381734,
+      "learning_rate": 4.946559351083475e-06,
+      "loss": 0.1994,
+      "step": 4968
+    },
+    {
+      "epoch": 1.6593755217899484,
+      "grad_norm": 0.39575079659822177,
+      "learning_rate": 4.944616132846174e-06,
+      "loss": 0.1897,
+      "step": 4969
+    },
+    {
+      "epoch": 1.6597094673568207,
+      "grad_norm": 0.4236243366509332,
+      "learning_rate": 4.942672922975255e-06,
+      "loss": 0.1893,
+      "step": 4970
+    },
+    {
+      "epoch": 1.6600434129236934,
+      "grad_norm": 0.42327398359903823,
+      "learning_rate": 4.940729721764268e-06,
+      "loss": 0.1849,
+      "step": 4971
+    },
+    {
+      "epoch": 1.6603773584905661,
+      "grad_norm": 0.41896021825786833,
+      "learning_rate": 4.938786529506755e-06,
+      "loss": 0.1968,
+      "step": 4972
+    },
+    {
+      "epoch": 1.6607113040574386,
+      "grad_norm": 0.44274001256771034,
+      "learning_rate": 4.936843346496257e-06,
+      "loss": 0.2005,
+      "step": 4973
+    },
+    {
+      "epoch": 1.661045249624311,
+      "grad_norm": 0.4440886178498309,
+      "learning_rate": 4.934900173026316e-06,
+      "loss": 0.2138,
+      "step": 4974
+    },
+    {
+      "epoch": 1.6613791951911838,
+      "grad_norm": 0.4494201652824032,
+      "learning_rate": 4.93295700939047e-06,
+      "loss": 0.2151,
+      "step": 4975
+    },
+    {
+      "epoch": 1.6617131407580565,
+      "grad_norm": 0.4008141428086073,
+      "learning_rate": 4.931013855882255e-06,
+      "loss": 0.1846,
+      "step": 4976
+    },
+    {
+      "epoch": 1.662047086324929,
+      "grad_norm": 0.4187124211937765,
+      "learning_rate": 4.929070712795211e-06,
+      "loss": 0.1974,
+      "step": 4977
+    },
+    {
+      "epoch": 1.6623810318918015,
+      "grad_norm": 0.3800461821906711,
+      "learning_rate": 4.927127580422871e-06,
+      "loss": 0.18,
+      "step": 4978
+    },
+    {
+      "epoch": 1.6627149774586742,
+      "grad_norm": 0.3790661455201971,
+      "learning_rate": 4.925184459058769e-06,
+      "loss": 0.1794,
+      "step": 4979
+    },
+    {
+      "epoch": 1.663048923025547,
+      "grad_norm": 0.4088327927350755,
+      "learning_rate": 4.9232413489964345e-06,
+      "loss": 0.1928,
+      "step": 4980
+    },
+    {
+      "epoch": 1.6633828685924195,
+      "grad_norm": 0.39844429506903506,
+      "learning_rate": 4.921298250529398e-06,
+      "loss": 0.1873,
+      "step": 4981
+    },
+    {
+      "epoch": 1.663716814159292,
+      "grad_norm": 0.3728433467890873,
+      "learning_rate": 4.919355163951186e-06,
+      "loss": 0.1875,
+      "step": 4982
+    },
+    {
+      "epoch": 1.6640507597261647,
+      "grad_norm": 0.4049596237809687,
+      "learning_rate": 4.917412089555328e-06,
+      "loss": 0.1834,
+      "step": 4983
+    },
+    {
+      "epoch": 1.6643847052930374,
+      "grad_norm": 0.43046575539496446,
+      "learning_rate": 4.915469027635345e-06,
+      "loss": 0.2045,
+      "step": 4984
+    },
+    {
+      "epoch": 1.6647186508599099,
+      "grad_norm": 0.34374023895710515,
+      "learning_rate": 4.9135259784847625e-06,
+      "loss": 0.1614,
+      "step": 4985
+    },
+    {
+      "epoch": 1.6650525964267824,
+      "grad_norm": 0.4120359522713758,
+      "learning_rate": 4.911582942397098e-06,
+      "loss": 0.1893,
+      "step": 4986
+    },
+    {
+      "epoch": 1.665386541993655,
+      "grad_norm": 0.4527687274785145,
+      "learning_rate": 4.909639919665872e-06,
+      "loss": 0.2074,
+      "step": 4987
+    },
+    {
+      "epoch": 1.6657204875605276,
+      "grad_norm": 0.42010451152241596,
+      "learning_rate": 4.907696910584599e-06,
+      "loss": 0.1902,
+      "step": 4988
+    },
+    {
+      "epoch": 1.6660544331274,
+      "grad_norm": 0.4080591968473613,
+      "learning_rate": 4.905753915446795e-06,
+      "loss": 0.1869,
+      "step": 4989
+    },
+    {
+      "epoch": 1.6663883786942728,
+      "grad_norm": 0.45688748641848076,
+      "learning_rate": 4.903810934545972e-06,
+      "loss": 0.1957,
+      "step": 4990
+    },
+    {
+      "epoch": 1.6667223242611455,
+      "grad_norm": 0.39479981393128605,
+      "learning_rate": 4.90186796817564e-06,
+      "loss": 0.1892,
+      "step": 4991
+    },
+    {
+      "epoch": 1.667056269828018,
+      "grad_norm": 0.3691965681647722,
+      "learning_rate": 4.899925016629307e-06,
+      "loss": 0.1811,
+      "step": 4992
+    },
+    {
+      "epoch": 1.6673902153948905,
+      "grad_norm": 0.45179774449253307,
+      "learning_rate": 4.897982080200477e-06,
+      "loss": 0.1988,
+      "step": 4993
+    },
+    {
+      "epoch": 1.6677241609617632,
+      "grad_norm": 0.3740310306629665,
+      "learning_rate": 4.896039159182655e-06,
+      "loss": 0.1808,
+      "step": 4994
+    },
+    {
+      "epoch": 1.668058106528636,
+      "grad_norm": 0.4089047480386049,
+      "learning_rate": 4.894096253869343e-06,
+      "loss": 0.1927,
+      "step": 4995
+    },
+    {
+      "epoch": 1.6683920520955084,
+      "grad_norm": 0.399676505442746,
+      "learning_rate": 4.89215336455404e-06,
+      "loss": 0.1931,
+      "step": 4996
+    },
+    {
+      "epoch": 1.668725997662381,
+      "grad_norm": 0.3895788933672829,
+      "learning_rate": 4.89021049153024e-06,
+      "loss": 0.1849,
+      "step": 4997
+    },
+    {
+      "epoch": 1.6690599432292537,
+      "grad_norm": 0.3595030957949447,
+      "learning_rate": 4.888267635091439e-06,
+      "loss": 0.173,
+      "step": 4998
+    },
+    {
+      "epoch": 1.6693938887961264,
+      "grad_norm": 0.41403286511225557,
+      "learning_rate": 4.886324795531126e-06,
+      "loss": 0.1847,
+      "step": 4999
+    },
+    {
+      "epoch": 1.6697278343629989,
+      "grad_norm": 0.4112481211288421,
+      "learning_rate": 4.88438197314279e-06,
+      "loss": 0.1868,
+      "step": 5000
+    },
+    {
+      "epoch": 1.6700617799298714,
+      "grad_norm": 0.41924546376727767,
+      "learning_rate": 4.88243916821992e-06,
+      "loss": 0.1997,
+      "step": 5001
+    },
+    {
+      "epoch": 1.670395725496744,
+      "grad_norm": 0.3888547499719048,
+      "learning_rate": 4.880496381055998e-06,
+      "loss": 0.1854,
+      "step": 5002
+    },
+    {
+      "epoch": 1.6707296710636166,
+      "grad_norm": 0.41588935541900535,
+      "learning_rate": 4.878553611944505e-06,
+      "loss": 0.1913,
+      "step": 5003
+    },
+    {
+      "epoch": 1.671063616630489,
+      "grad_norm": 0.37772961201560074,
+      "learning_rate": 4.876610861178918e-06,
+      "loss": 0.1902,
+      "step": 5004
+    },
+    {
+      "epoch": 1.6713975621973618,
+      "grad_norm": 0.39759117151567896,
+      "learning_rate": 4.874668129052712e-06,
+      "loss": 0.2015,
+      "step": 5005
+    },
+    {
+      "epoch": 1.6717315077642345,
+      "grad_norm": 0.40235692971912135,
+      "learning_rate": 4.872725415859363e-06,
+      "loss": 0.1863,
+      "step": 5006
+    },
+    {
+      "epoch": 1.672065453331107,
+      "grad_norm": 0.36928962349677913,
+      "learning_rate": 4.8707827218923385e-06,
+      "loss": 0.1808,
+      "step": 5007
+    },
+    {
+      "epoch": 1.6723993988979795,
+      "grad_norm": 0.4562272715321814,
+      "learning_rate": 4.868840047445106e-06,
+      "loss": 0.2151,
+      "step": 5008
+    },
+    {
+      "epoch": 1.6727333444648522,
+      "grad_norm": 0.43138486747940097,
+      "learning_rate": 4.866897392811127e-06,
+      "loss": 0.213,
+      "step": 5009
+    },
+    {
+      "epoch": 1.673067290031725,
+      "grad_norm": 0.39105820437363875,
+      "learning_rate": 4.864954758283865e-06,
+      "loss": 0.1844,
+      "step": 5010
+    },
+    {
+      "epoch": 1.6734012355985974,
+      "grad_norm": 0.39624876568230644,
+      "learning_rate": 4.8630121441567755e-06,
+      "loss": 0.1976,
+      "step": 5011
+    },
+    {
+      "epoch": 1.67373518116547,
+      "grad_norm": 0.38463611162536465,
+      "learning_rate": 4.861069550723316e-06,
+      "loss": 0.1955,
+      "step": 5012
+    },
+    {
+      "epoch": 1.6740691267323426,
+      "grad_norm": 0.4218274037026937,
+      "learning_rate": 4.859126978276937e-06,
+      "loss": 0.2063,
+      "step": 5013
+    },
+    {
+      "epoch": 1.6744030722992154,
+      "grad_norm": 0.4055270643811191,
+      "learning_rate": 4.857184427111086e-06,
+      "loss": 0.2086,
+      "step": 5014
+    },
+    {
+      "epoch": 1.6747370178660879,
+      "grad_norm": 0.3660260268828606,
+      "learning_rate": 4.855241897519209e-06,
+      "loss": 0.1744,
+      "step": 5015
+    },
+    {
+      "epoch": 1.6750709634329604,
+      "grad_norm": 0.42399611512123225,
+      "learning_rate": 4.8532993897947464e-06,
+      "loss": 0.1971,
+      "step": 5016
+    },
+    {
+      "epoch": 1.675404908999833,
+      "grad_norm": 0.4052742336770543,
+      "learning_rate": 4.851356904231139e-06,
+      "loss": 0.1851,
+      "step": 5017
+    },
+    {
+      "epoch": 1.6757388545667058,
+      "grad_norm": 0.4318706621172862,
+      "learning_rate": 4.849414441121823e-06,
+      "loss": 0.2056,
+      "step": 5018
+    },
+    {
+      "epoch": 1.676072800133578,
+      "grad_norm": 0.4183164683620372,
+      "learning_rate": 4.847472000760228e-06,
+      "loss": 0.2033,
+      "step": 5019
+    },
+    {
+      "epoch": 1.6764067457004508,
+      "grad_norm": 0.41452196588285173,
+      "learning_rate": 4.845529583439783e-06,
+      "loss": 0.1937,
+      "step": 5020
+    },
+    {
+      "epoch": 1.6767406912673235,
+      "grad_norm": 0.44161534649268036,
+      "learning_rate": 4.843587189453914e-06,
+      "loss": 0.2104,
+      "step": 5021
+    },
+    {
+      "epoch": 1.677074636834196,
+      "grad_norm": 0.41291953906054923,
+      "learning_rate": 4.84164481909604e-06,
+      "loss": 0.1964,
+      "step": 5022
+    },
+    {
+      "epoch": 1.6774085824010685,
+      "grad_norm": 0.4107451131477934,
+      "learning_rate": 4.839702472659581e-06,
+      "loss": 0.1848,
+      "step": 5023
+    },
+    {
+      "epoch": 1.6777425279679412,
+      "grad_norm": 0.40593954437157626,
+      "learning_rate": 4.837760150437952e-06,
+      "loss": 0.1927,
+      "step": 5024
+    },
+    {
+      "epoch": 1.678076473534814,
+      "grad_norm": 0.41888894475916594,
+      "learning_rate": 4.8358178527245625e-06,
+      "loss": 0.2006,
+      "step": 5025
+    },
+    {
+      "epoch": 1.6784104191016864,
+      "grad_norm": 0.43080741370800135,
+      "learning_rate": 4.83387557981282e-06,
+      "loss": 0.196,
+      "step": 5026
+    },
+    {
+      "epoch": 1.678744364668559,
+      "grad_norm": 0.40951343834625403,
+      "learning_rate": 4.831933331996126e-06,
+      "loss": 0.1998,
+      "step": 5027
+    },
+    {
+      "epoch": 1.6790783102354316,
+      "grad_norm": 0.3875819486249406,
+      "learning_rate": 4.8299911095678816e-06,
+      "loss": 0.1833,
+      "step": 5028
+    },
+    {
+      "epoch": 1.6794122558023044,
+      "grad_norm": 0.4300987227755639,
+      "learning_rate": 4.82804891282148e-06,
+      "loss": 0.2069,
+      "step": 5029
+    },
+    {
+      "epoch": 1.6797462013691769,
+      "grad_norm": 0.37297544793032655,
+      "learning_rate": 4.8261067420503175e-06,
+      "loss": 0.188,
+      "step": 5030
+    },
+    {
+      "epoch": 1.6800801469360493,
+      "grad_norm": 0.39096748045381796,
+      "learning_rate": 4.8241645975477785e-06,
+      "loss": 0.1835,
+      "step": 5031
+    },
+    {
+      "epoch": 1.680414092502922,
+      "grad_norm": 0.3996077225322656,
+      "learning_rate": 4.822222479607247e-06,
+      "loss": 0.1943,
+      "step": 5032
+    },
+    {
+      "epoch": 1.6807480380697948,
+      "grad_norm": 0.4304762784482779,
+      "learning_rate": 4.820280388522104e-06,
+      "loss": 0.1948,
+      "step": 5033
+    },
+    {
+      "epoch": 1.6810819836366673,
+      "grad_norm": 0.3805492053734514,
+      "learning_rate": 4.818338324585725e-06,
+      "loss": 0.1838,
+      "step": 5034
+    },
+    {
+      "epoch": 1.6814159292035398,
+      "grad_norm": 0.37585210563270655,
+      "learning_rate": 4.816396288091478e-06,
+      "loss": 0.1854,
+      "step": 5035
+    },
+    {
+      "epoch": 1.6817498747704125,
+      "grad_norm": 0.42351245196254,
+      "learning_rate": 4.814454279332737e-06,
+      "loss": 0.2025,
+      "step": 5036
+    },
+    {
+      "epoch": 1.682083820337285,
+      "grad_norm": 0.41695687567784445,
+      "learning_rate": 4.81251229860286e-06,
+      "loss": 0.193,
+      "step": 5037
+    },
+    {
+      "epoch": 1.6824177659041575,
+      "grad_norm": 0.537261725808297,
+      "learning_rate": 4.810570346195207e-06,
+      "loss": 0.2114,
+      "step": 5038
+    },
+    {
+      "epoch": 1.6827517114710302,
+      "grad_norm": 0.4113025086849026,
+      "learning_rate": 4.808628422403135e-06,
+      "loss": 0.1948,
+      "step": 5039
+    },
+    {
+      "epoch": 1.683085657037903,
+      "grad_norm": 0.4076342773958272,
+      "learning_rate": 4.806686527519994e-06,
+      "loss": 0.1986,
+      "step": 5040
+    },
+    {
+      "epoch": 1.6834196026047754,
+      "grad_norm": 0.41025435360946494,
+      "learning_rate": 4.804744661839128e-06,
+      "loss": 0.2066,
+      "step": 5041
+    },
+    {
+      "epoch": 1.683753548171648,
+      "grad_norm": 0.3636218816609904,
+      "learning_rate": 4.80280282565388e-06,
+      "loss": 0.1872,
+      "step": 5042
+    },
+    {
+      "epoch": 1.6840874937385206,
+      "grad_norm": 0.3950953132425706,
+      "learning_rate": 4.800861019257587e-06,
+      "loss": 0.1984,
+      "step": 5043
+    },
+    {
+      "epoch": 1.6844214393053933,
+      "grad_norm": 0.42371769285822725,
+      "learning_rate": 4.798919242943583e-06,
+      "loss": 0.2037,
+      "step": 5044
+    },
+    {
+      "epoch": 1.6847553848722658,
+      "grad_norm": 0.392433056196255,
+      "learning_rate": 4.796977497005194e-06,
+      "loss": 0.1871,
+      "step": 5045
+    },
+    {
+      "epoch": 1.6850893304391383,
+      "grad_norm": 0.46155417346154815,
+      "learning_rate": 4.795035781735743e-06,
+      "loss": 0.1978,
+      "step": 5046
+    },
+    {
+      "epoch": 1.685423276006011,
+      "grad_norm": 0.440921144287475,
+      "learning_rate": 4.793094097428552e-06,
+      "loss": 0.2139,
+      "step": 5047
+    },
+    {
+      "epoch": 1.6857572215728838,
+      "grad_norm": 0.3794880760617637,
+      "learning_rate": 4.7911524443769346e-06,
+      "loss": 0.1911,
+      "step": 5048
+    },
+    {
+      "epoch": 1.6860911671397563,
+      "grad_norm": 0.395057876244426,
+      "learning_rate": 4.789210822874199e-06,
+      "loss": 0.1938,
+      "step": 5049
+    },
+    {
+      "epoch": 1.6864251127066288,
+      "grad_norm": 0.3775844552491621,
+      "learning_rate": 4.787269233213651e-06,
+      "loss": 0.188,
+      "step": 5050
+    },
+    {
+      "epoch": 1.6867590582735015,
+      "grad_norm": 0.3603621487839833,
+      "learning_rate": 4.785327675688591e-06,
+      "loss": 0.1885,
+      "step": 5051
+    },
+    {
+      "epoch": 1.687093003840374,
+      "grad_norm": 0.4027004483451677,
+      "learning_rate": 4.7833861505923096e-06,
+      "loss": 0.1893,
+      "step": 5052
+    },
+    {
+      "epoch": 1.6874269494072465,
+      "grad_norm": 0.4765942958527092,
+      "learning_rate": 4.781444658218103e-06,
+      "loss": 0.1933,
+      "step": 5053
+    },
+    {
+      "epoch": 1.6877608949741192,
+      "grad_norm": 0.41479948198174693,
+      "learning_rate": 4.779503198859255e-06,
+      "loss": 0.1885,
+      "step": 5054
+    },
+    {
+      "epoch": 1.688094840540992,
+      "grad_norm": 0.39411292837696527,
+      "learning_rate": 4.777561772809045e-06,
+      "loss": 0.1867,
+      "step": 5055
+    },
+    {
+      "epoch": 1.6884287861078644,
+      "grad_norm": 0.3932933436171054,
+      "learning_rate": 4.775620380360747e-06,
+      "loss": 0.1996,
+      "step": 5056
+    },
+    {
+      "epoch": 1.688762731674737,
+      "grad_norm": 0.5515039821445362,
+      "learning_rate": 4.773679021807634e-06,
+      "loss": 0.2012,
+      "step": 5057
+    },
+    {
+      "epoch": 1.6890966772416096,
+      "grad_norm": 0.4175879612749353,
+      "learning_rate": 4.771737697442968e-06,
+      "loss": 0.1885,
+      "step": 5058
+    },
+    {
+      "epoch": 1.6894306228084823,
+      "grad_norm": 0.4073593770421272,
+      "learning_rate": 4.7697964075600114e-06,
+      "loss": 0.1963,
+      "step": 5059
+    },
+    {
+      "epoch": 1.6897645683753548,
+      "grad_norm": 0.41602326668910145,
+      "learning_rate": 4.767855152452019e-06,
+      "loss": 0.1945,
+      "step": 5060
+    },
+    {
+      "epoch": 1.6900985139422273,
+      "grad_norm": 0.444598225558259,
+      "learning_rate": 4.765913932412237e-06,
+      "loss": 0.2069,
+      "step": 5061
+    },
+    {
+      "epoch": 1.6904324595091,
+      "grad_norm": 0.383752685178437,
+      "learning_rate": 4.763972747733913e-06,
+      "loss": 0.1897,
+      "step": 5062
+    },
+    {
+      "epoch": 1.6907664050759728,
+      "grad_norm": 0.42119196388224206,
+      "learning_rate": 4.762031598710285e-06,
+      "loss": 0.1927,
+      "step": 5063
+    },
+    {
+      "epoch": 1.6911003506428453,
+      "grad_norm": 0.4190290794400875,
+      "learning_rate": 4.760090485634584e-06,
+      "loss": 0.1973,
+      "step": 5064
+    },
+    {
+      "epoch": 1.6914342962097177,
+      "grad_norm": 0.423228700372625,
+      "learning_rate": 4.758149408800042e-06,
+      "loss": 0.2059,
+      "step": 5065
+    },
+    {
+      "epoch": 1.6917682417765905,
+      "grad_norm": 0.3900317630440837,
+      "learning_rate": 4.756208368499879e-06,
+      "loss": 0.1885,
+      "step": 5066
+    },
+    {
+      "epoch": 1.692102187343463,
+      "grad_norm": 0.412549195493742,
+      "learning_rate": 4.754267365027314e-06,
+      "loss": 0.1965,
+      "step": 5067
+    },
+    {
+      "epoch": 1.6924361329103355,
+      "grad_norm": 0.3953914528356778,
+      "learning_rate": 4.752326398675555e-06,
+      "loss": 0.2009,
+      "step": 5068
+    },
+    {
+      "epoch": 1.6927700784772082,
+      "grad_norm": 0.42031070024553674,
+      "learning_rate": 4.750385469737811e-06,
+      "loss": 0.1878,
+      "step": 5069
+    },
+    {
+      "epoch": 1.693104024044081,
+      "grad_norm": 0.4127570219253499,
+      "learning_rate": 4.748444578507278e-06,
+      "loss": 0.207,
+      "step": 5070
+    },
+    {
+      "epoch": 1.6934379696109534,
+      "grad_norm": 0.41827513152551066,
+      "learning_rate": 4.746503725277156e-06,
+      "loss": 0.2002,
+      "step": 5071
+    },
+    {
+      "epoch": 1.6937719151778259,
+      "grad_norm": 0.48196867703987134,
+      "learning_rate": 4.744562910340631e-06,
+      "loss": 0.2071,
+      "step": 5072
+    },
+    {
+      "epoch": 1.6941058607446986,
+      "grad_norm": 0.38681708500322204,
+      "learning_rate": 4.742622133990885e-06,
+      "loss": 0.185,
+      "step": 5073
+    },
+    {
+      "epoch": 1.6944398063115713,
+      "grad_norm": 0.4062271054342476,
+      "learning_rate": 4.740681396521097e-06,
+      "loss": 0.1876,
+      "step": 5074
+    },
+    {
+      "epoch": 1.6947737518784438,
+      "grad_norm": 0.516392208571396,
+      "learning_rate": 4.738740698224438e-06,
+      "loss": 0.2053,
+      "step": 5075
+    },
+    {
+      "epoch": 1.6951076974453163,
+      "grad_norm": 0.4361598083182339,
+      "learning_rate": 4.73680003939407e-06,
+      "loss": 0.2121,
+      "step": 5076
+    },
+    {
+      "epoch": 1.695441643012189,
+      "grad_norm": 0.41160087711576554,
+      "learning_rate": 4.734859420323158e-06,
+      "loss": 0.2036,
+      "step": 5077
+    },
+    {
+      "epoch": 1.6957755885790617,
+      "grad_norm": 0.49946880177831654,
+      "learning_rate": 4.7329188413048515e-06,
+      "loss": 0.1999,
+      "step": 5078
+    },
+    {
+      "epoch": 1.6961095341459342,
+      "grad_norm": 0.3666729685844792,
+      "learning_rate": 4.7309783026322995e-06,
+      "loss": 0.1833,
+      "step": 5079
+    },
+    {
+      "epoch": 1.6964434797128067,
+      "grad_norm": 0.40908472431553156,
+      "learning_rate": 4.7290378045986425e-06,
+      "loss": 0.1904,
+      "step": 5080
+    },
+    {
+      "epoch": 1.6967774252796795,
+      "grad_norm": 0.4479324104960931,
+      "learning_rate": 4.727097347497014e-06,
+      "loss": 0.1967,
+      "step": 5081
+    },
+    {
+      "epoch": 1.6971113708465522,
+      "grad_norm": 0.43959080462229055,
+      "learning_rate": 4.7251569316205455e-06,
+      "loss": 0.2051,
+      "step": 5082
+    },
+    {
+      "epoch": 1.6974453164134247,
+      "grad_norm": 0.38971023600971766,
+      "learning_rate": 4.723216557262359e-06,
+      "loss": 0.1886,
+      "step": 5083
+    },
+    {
+      "epoch": 1.6977792619802972,
+      "grad_norm": 0.4760777142908489,
+      "learning_rate": 4.721276224715569e-06,
+      "loss": 0.201,
+      "step": 5084
+    },
+    {
+      "epoch": 1.6981132075471699,
+      "grad_norm": 0.38860098507630997,
+      "learning_rate": 4.719335934273289e-06,
+      "loss": 0.1859,
+      "step": 5085
+    },
+    {
+      "epoch": 1.6984471531140424,
+      "grad_norm": 0.40103252102303055,
+      "learning_rate": 4.717395686228621e-06,
+      "loss": 0.1884,
+      "step": 5086
+    },
+    {
+      "epoch": 1.6987810986809149,
+      "grad_norm": 0.36902419962259236,
+      "learning_rate": 4.715455480874661e-06,
+      "loss": 0.1814,
+      "step": 5087
+    },
+    {
+      "epoch": 1.6991150442477876,
+      "grad_norm": 0.412125534705266,
+      "learning_rate": 4.713515318504501e-06,
+      "loss": 0.1913,
+      "step": 5088
+    },
+    {
+      "epoch": 1.6994489898146603,
+      "grad_norm": 0.39157748575603113,
+      "learning_rate": 4.711575199411226e-06,
+      "loss": 0.1829,
+      "step": 5089
+    },
+    {
+      "epoch": 1.6997829353815328,
+      "grad_norm": 0.41802116475269446,
+      "learning_rate": 4.7096351238879135e-06,
+      "loss": 0.1978,
+      "step": 5090
+    },
+    {
+      "epoch": 1.7001168809484053,
+      "grad_norm": 0.6279979025769422,
+      "learning_rate": 4.707695092227634e-06,
+      "loss": 0.1937,
+      "step": 5091
+    },
+    {
+      "epoch": 1.700450826515278,
+      "grad_norm": 0.4114760298428927,
+      "learning_rate": 4.705755104723453e-06,
+      "loss": 0.199,
+      "step": 5092
+    },
+    {
+      "epoch": 1.7007847720821507,
+      "grad_norm": 0.4395547405428648,
+      "learning_rate": 4.703815161668426e-06,
+      "loss": 0.1931,
+      "step": 5093
+    },
+    {
+      "epoch": 1.7011187176490232,
+      "grad_norm": 0.41225308183331766,
+      "learning_rate": 4.701875263355608e-06,
+      "loss": 0.191,
+      "step": 5094
+    },
+    {
+      "epoch": 1.7014526632158957,
+      "grad_norm": 0.3958758087990041,
+      "learning_rate": 4.699935410078042e-06,
+      "loss": 0.1941,
+      "step": 5095
+    },
+    {
+      "epoch": 1.7017866087827684,
+      "grad_norm": 0.41315215933016763,
+      "learning_rate": 4.697995602128766e-06,
+      "loss": 0.1947,
+      "step": 5096
+    },
+    {
+      "epoch": 1.7021205543496412,
+      "grad_norm": 0.3955180863149731,
+      "learning_rate": 4.696055839800809e-06,
+      "loss": 0.1991,
+      "step": 5097
+    },
+    {
+      "epoch": 1.7024544999165137,
+      "grad_norm": 0.43154636895066495,
+      "learning_rate": 4.694116123387197e-06,
+      "loss": 0.2104,
+      "step": 5098
+    },
+    {
+      "epoch": 1.7027884454833861,
+      "grad_norm": 0.4007233166701319,
+      "learning_rate": 4.692176453180944e-06,
+      "loss": 0.1902,
+      "step": 5099
+    },
+    {
+      "epoch": 1.7031223910502589,
+      "grad_norm": 0.40361257750068574,
+      "learning_rate": 4.6902368294750644e-06,
+      "loss": 0.2028,
+      "step": 5100
+    },
+    {
+      "epoch": 1.7034563366171314,
+      "grad_norm": 0.40699361291263014,
+      "learning_rate": 4.688297252562559e-06,
+      "loss": 0.1989,
+      "step": 5101
+    },
+    {
+      "epoch": 1.7037902821840039,
+      "grad_norm": 0.4079101725879806,
+      "learning_rate": 4.6863577227364235e-06,
+      "loss": 0.1909,
+      "step": 5102
+    },
+    {
+      "epoch": 1.7041242277508766,
+      "grad_norm": 0.38587148305952435,
+      "learning_rate": 4.684418240289648e-06,
+      "loss": 0.1861,
+      "step": 5103
+    },
+    {
+      "epoch": 1.7044581733177493,
+      "grad_norm": 0.4625478571203875,
+      "learning_rate": 4.682478805515212e-06,
+      "loss": 0.202,
+      "step": 5104
+    },
+    {
+      "epoch": 1.7047921188846218,
+      "grad_norm": 0.4358990786312495,
+      "learning_rate": 4.680539418706091e-06,
+      "loss": 0.1927,
+      "step": 5105
+    },
+    {
+      "epoch": 1.7051260644514943,
+      "grad_norm": 0.41133770169381584,
+      "learning_rate": 4.678600080155252e-06,
+      "loss": 0.1775,
+      "step": 5106
+    },
+    {
+      "epoch": 1.705460010018367,
+      "grad_norm": 0.3962294559870226,
+      "learning_rate": 4.676660790155656e-06,
+      "loss": 0.1981,
+      "step": 5107
+    },
+    {
+      "epoch": 1.7057939555852397,
+      "grad_norm": 0.40931109170500446,
+      "learning_rate": 4.674721549000255e-06,
+      "loss": 0.2042,
+      "step": 5108
+    },
+    {
+      "epoch": 1.7061279011521122,
+      "grad_norm": 0.4107315070526611,
+      "learning_rate": 4.6727823569819944e-06,
+      "loss": 0.1878,
+      "step": 5109
+    },
+    {
+      "epoch": 1.7064618467189847,
+      "grad_norm": 0.3752388358432499,
+      "learning_rate": 4.670843214393811e-06,
+      "loss": 0.1813,
+      "step": 5110
+    },
+    {
+      "epoch": 1.7067957922858574,
+      "grad_norm": 0.43157554986961477,
+      "learning_rate": 4.6689041215286344e-06,
+      "loss": 0.2003,
+      "step": 5111
+    },
+    {
+      "epoch": 1.7071297378527301,
+      "grad_norm": 0.42486623254964884,
+      "learning_rate": 4.666965078679391e-06,
+      "loss": 0.1971,
+      "step": 5112
+    },
+    {
+      "epoch": 1.7074636834196026,
+      "grad_norm": 0.405339605614925,
+      "learning_rate": 4.665026086138993e-06,
+      "loss": 0.2053,
+      "step": 5113
+    },
+    {
+      "epoch": 1.7077976289864751,
+      "grad_norm": 0.4002710266542765,
+      "learning_rate": 4.66308714420035e-06,
+      "loss": 0.186,
+      "step": 5114
+    },
+    {
+      "epoch": 1.7081315745533479,
+      "grad_norm": 0.4126660129106094,
+      "learning_rate": 4.6611482531563595e-06,
+      "loss": 0.1961,
+      "step": 5115
+    },
+    {
+      "epoch": 1.7084655201202203,
+      "grad_norm": 0.45263061369654795,
+      "learning_rate": 4.659209413299916e-06,
+      "loss": 0.2105,
+      "step": 5116
+    },
+    {
+      "epoch": 1.7087994656870928,
+      "grad_norm": 0.39704063413819723,
+      "learning_rate": 4.657270624923901e-06,
+      "loss": 0.1866,
+      "step": 5117
+    },
+    {
+      "epoch": 1.7091334112539656,
+      "grad_norm": 0.456588695870717,
+      "learning_rate": 4.6553318883211955e-06,
+      "loss": 0.1986,
+      "step": 5118
+    },
+    {
+      "epoch": 1.7094673568208383,
+      "grad_norm": 0.4056701874158402,
+      "learning_rate": 4.653393203784667e-06,
+      "loss": 0.1937,
+      "step": 5119
+    },
+    {
+      "epoch": 1.7098013023877108,
+      "grad_norm": 0.37801865981225746,
+      "learning_rate": 4.651454571607176e-06,
+      "loss": 0.1828,
+      "step": 5120
+    },
+    {
+      "epoch": 1.7101352479545833,
+      "grad_norm": 0.4340941520725217,
+      "learning_rate": 4.649515992081576e-06,
+      "loss": 0.1945,
+      "step": 5121
+    },
+    {
+      "epoch": 1.710469193521456,
+      "grad_norm": 0.41411971112825,
+      "learning_rate": 4.64757746550071e-06,
+      "loss": 0.2097,
+      "step": 5122
+    },
+    {
+      "epoch": 1.7108031390883287,
+      "grad_norm": 0.4212025495479919,
+      "learning_rate": 4.645638992157419e-06,
+      "loss": 0.1967,
+      "step": 5123
+    },
+    {
+      "epoch": 1.7111370846552012,
+      "grad_norm": 0.40864256762488815,
+      "learning_rate": 4.6437005723445316e-06,
+      "loss": 0.1974,
+      "step": 5124
+    },
+    {
+      "epoch": 1.7114710302220737,
+      "grad_norm": 0.4028769432832656,
+      "learning_rate": 4.6417622063548675e-06,
+      "loss": 0.1991,
+      "step": 5125
+    },
+    {
+      "epoch": 1.7118049757889464,
+      "grad_norm": 0.6276560821172299,
+      "learning_rate": 4.6398238944812414e-06,
+      "loss": 0.1923,
+      "step": 5126
+    },
+    {
+      "epoch": 1.7121389213558191,
+      "grad_norm": 0.4154958288569983,
+      "learning_rate": 4.637885637016456e-06,
+      "loss": 0.1945,
+      "step": 5127
+    },
+    {
+      "epoch": 1.7124728669226916,
+      "grad_norm": 0.3784544159799719,
+      "learning_rate": 4.635947434253308e-06,
+      "loss": 0.174,
+      "step": 5128
+    },
+    {
+      "epoch": 1.7128068124895641,
+      "grad_norm": 0.4121407590389507,
+      "learning_rate": 4.634009286484586e-06,
+      "loss": 0.195,
+      "step": 5129
+    },
+    {
+      "epoch": 1.7131407580564368,
+      "grad_norm": 0.3936720216642308,
+      "learning_rate": 4.632071194003073e-06,
+      "loss": 0.1804,
+      "step": 5130
+    },
+    {
+      "epoch": 1.7134747036233096,
+      "grad_norm": 0.41086226040341123,
+      "learning_rate": 4.630133157101537e-06,
+      "loss": 0.1921,
+      "step": 5131
+    },
+    {
+      "epoch": 1.713808649190182,
+      "grad_norm": 0.3884513041579706,
+      "learning_rate": 4.6281951760727435e-06,
+      "loss": 0.1786,
+      "step": 5132
+    },
+    {
+      "epoch": 1.7141425947570545,
+      "grad_norm": 0.39582994223607226,
+      "learning_rate": 4.626257251209446e-06,
+      "loss": 0.1835,
+      "step": 5133
+    },
+    {
+      "epoch": 1.7144765403239273,
+      "grad_norm": 0.5027499487784188,
+      "learning_rate": 4.624319382804391e-06,
+      "loss": 0.205,
+      "step": 5134
+    },
+    {
+      "epoch": 1.7148104858907998,
+      "grad_norm": 0.3865108788544779,
+      "learning_rate": 4.622381571150317e-06,
+      "loss": 0.1791,
+      "step": 5135
+    },
+    {
+      "epoch": 1.7151444314576723,
+      "grad_norm": 0.41736826198301175,
+      "learning_rate": 4.620443816539954e-06,
+      "loss": 0.1845,
+      "step": 5136
+    },
+    {
+      "epoch": 1.715478377024545,
+      "grad_norm": 0.4076416332201311,
+      "learning_rate": 4.618506119266021e-06,
+      "loss": 0.2023,
+      "step": 5137
+    },
+    {
+      "epoch": 1.7158123225914177,
+      "grad_norm": 0.3940577448952557,
+      "learning_rate": 4.6165684796212306e-06,
+      "loss": 0.1971,
+      "step": 5138
+    },
+    {
+      "epoch": 1.7161462681582902,
+      "grad_norm": 0.6699527350793285,
+      "learning_rate": 4.6146308978982865e-06,
+      "loss": 0.2015,
+      "step": 5139
+    },
+    {
+      "epoch": 1.7164802137251627,
+      "grad_norm": 0.41792420253823254,
+      "learning_rate": 4.612693374389881e-06,
+      "loss": 0.1996,
+      "step": 5140
+    },
+    {
+      "epoch": 1.7168141592920354,
+      "grad_norm": 0.486254656967708,
+      "learning_rate": 4.610755909388703e-06,
+      "loss": 0.2017,
+      "step": 5141
+    },
+    {
+      "epoch": 1.7171481048589081,
+      "grad_norm": 0.43032663564827006,
+      "learning_rate": 4.608818503187428e-06,
+      "loss": 0.2077,
+      "step": 5142
+    },
+    {
+      "epoch": 1.7174820504257806,
+      "grad_norm": 0.4197636003363948,
+      "learning_rate": 4.606881156078725e-06,
+      "loss": 0.2063,
+      "step": 5143
+    },
+    {
+      "epoch": 1.717815995992653,
+      "grad_norm": 0.41927575757137314,
+      "learning_rate": 4.604943868355251e-06,
+      "loss": 0.1971,
+      "step": 5144
+    },
+    {
+      "epoch": 1.7181499415595258,
+      "grad_norm": 0.4364890921897206,
+      "learning_rate": 4.603006640309658e-06,
+      "loss": 0.1992,
+      "step": 5145
+    },
+    {
+      "epoch": 1.7184838871263985,
+      "grad_norm": 0.40568220551742395,
+      "learning_rate": 4.601069472234584e-06,
+      "loss": 0.2016,
+      "step": 5146
+    },
+    {
+      "epoch": 1.718817832693271,
+      "grad_norm": 0.41925099662175647,
+      "learning_rate": 4.599132364422666e-06,
+      "loss": 0.2029,
+      "step": 5147
+    },
+    {
+      "epoch": 1.7191517782601435,
+      "grad_norm": 0.4064241820068898,
+      "learning_rate": 4.597195317166525e-06,
+      "loss": 0.1941,
+      "step": 5148
+    },
+    {
+      "epoch": 1.7194857238270163,
+      "grad_norm": 0.40433828939310057,
+      "learning_rate": 4.595258330758773e-06,
+      "loss": 0.1928,
+      "step": 5149
+    },
+    {
+      "epoch": 1.7198196693938888,
+      "grad_norm": 0.42000296483673916,
+      "learning_rate": 4.593321405492017e-06,
+      "loss": 0.1987,
+      "step": 5150
+    },
+    {
+      "epoch": 1.7201536149607612,
+      "grad_norm": 0.4031145987386355,
+      "learning_rate": 4.59138454165885e-06,
+      "loss": 0.1867,
+      "step": 5151
+    },
+    {
+      "epoch": 1.720487560527634,
+      "grad_norm": 0.3927642852721298,
+      "learning_rate": 4.589447739551857e-06,
+      "loss": 0.1886,
+      "step": 5152
+    },
+    {
+      "epoch": 1.7208215060945067,
+      "grad_norm": 0.4365198371307918,
+      "learning_rate": 4.58751099946362e-06,
+      "loss": 0.2067,
+      "step": 5153
+    },
+    {
+      "epoch": 1.7211554516613792,
+      "grad_norm": 0.47754776245942315,
+      "learning_rate": 4.585574321686704e-06,
+      "loss": 0.2044,
+      "step": 5154
+    },
+    {
+      "epoch": 1.7214893972282517,
+      "grad_norm": 0.41403702099913486,
+      "learning_rate": 4.583637706513665e-06,
+      "loss": 0.2045,
+      "step": 5155
+    },
+    {
+      "epoch": 1.7218233427951244,
+      "grad_norm": 0.4129816326090922,
+      "learning_rate": 4.5817011542370535e-06,
+      "loss": 0.1912,
+      "step": 5156
+    },
+    {
+      "epoch": 1.722157288361997,
+      "grad_norm": 0.42004603602811036,
+      "learning_rate": 4.579764665149409e-06,
+      "loss": 0.2005,
+      "step": 5157
+    },
+    {
+      "epoch": 1.7224912339288696,
+      "grad_norm": 0.4321750576374955,
+      "learning_rate": 4.577828239543257e-06,
+      "loss": 0.1997,
+      "step": 5158
+    },
+    {
+      "epoch": 1.722825179495742,
+      "grad_norm": 0.405430374187034,
+      "learning_rate": 4.575891877711123e-06,
+      "loss": 0.2003,
+      "step": 5159
+    },
+    {
+      "epoch": 1.7231591250626148,
+      "grad_norm": 0.3853641882041907,
+      "learning_rate": 4.573955579945514e-06,
+      "loss": 0.1809,
+      "step": 5160
+    },
+    {
+      "epoch": 1.7234930706294875,
+      "grad_norm": 0.39519841050530224,
+      "learning_rate": 4.572019346538931e-06,
+      "loss": 0.1826,
+      "step": 5161
+    },
+    {
+      "epoch": 1.72382701619636,
+      "grad_norm": 0.39830083588979237,
+      "learning_rate": 4.570083177783865e-06,
+      "loss": 0.1985,
+      "step": 5162
+    },
+    {
+      "epoch": 1.7241609617632325,
+      "grad_norm": 0.3912144272420254,
+      "learning_rate": 4.568147073972795e-06,
+      "loss": 0.1791,
+      "step": 5163
+    },
+    {
+      "epoch": 1.7244949073301052,
+      "grad_norm": 0.438731828398516,
+      "learning_rate": 4.566211035398196e-06,
+      "loss": 0.2143,
+      "step": 5164
+    },
+    {
+      "epoch": 1.7248288528969777,
+      "grad_norm": 0.38952165500853425,
+      "learning_rate": 4.564275062352529e-06,
+      "loss": 0.1898,
+      "step": 5165
+    },
+    {
+      "epoch": 1.7251627984638502,
+      "grad_norm": 0.46702184416618503,
+      "learning_rate": 4.5623391551282435e-06,
+      "loss": 0.2063,
+      "step": 5166
+    },
+    {
+      "epoch": 1.725496744030723,
+      "grad_norm": 0.3862830352258704,
+      "learning_rate": 4.560403314017782e-06,
+      "loss": 0.1919,
+      "step": 5167
+    },
+    {
+      "epoch": 1.7258306895975957,
+      "grad_norm": 0.3939486702748393,
+      "learning_rate": 4.558467539313576e-06,
+      "loss": 0.185,
+      "step": 5168
+    },
+    {
+      "epoch": 1.7261646351644682,
+      "grad_norm": 0.4475298224185486,
+      "learning_rate": 4.556531831308045e-06,
+      "loss": 0.2084,
+      "step": 5169
+    },
+    {
+      "epoch": 1.7264985807313407,
+      "grad_norm": 0.3740950936131439,
+      "learning_rate": 4.554596190293606e-06,
+      "loss": 0.1861,
+      "step": 5170
+    },
+    {
+      "epoch": 1.7268325262982134,
+      "grad_norm": 0.4279837775356034,
+      "learning_rate": 4.552660616562655e-06,
+      "loss": 0.1956,
+      "step": 5171
+    },
+    {
+      "epoch": 1.727166471865086,
+      "grad_norm": 0.38629605489548957,
+      "learning_rate": 4.550725110407586e-06,
+      "loss": 0.1882,
+      "step": 5172
+    },
+    {
+      "epoch": 1.7275004174319586,
+      "grad_norm": 0.40903630345175285,
+      "learning_rate": 4.548789672120779e-06,
+      "loss": 0.1947,
+      "step": 5173
+    },
+    {
+      "epoch": 1.727834362998831,
+      "grad_norm": 0.40754681344598553,
+      "learning_rate": 4.5468543019946045e-06,
+      "loss": 0.1895,
+      "step": 5174
+    },
+    {
+      "epoch": 1.7281683085657038,
+      "grad_norm": 0.4154808397286242,
+      "learning_rate": 4.544919000321421e-06,
+      "loss": 0.1979,
+      "step": 5175
+    },
+    {
+      "epoch": 1.7285022541325765,
+      "grad_norm": 0.3927989079761158,
+      "learning_rate": 4.542983767393584e-06,
+      "loss": 0.187,
+      "step": 5176
+    },
+    {
+      "epoch": 1.728836199699449,
+      "grad_norm": 0.4477597515047998,
+      "learning_rate": 4.541048603503429e-06,
+      "loss": 0.1982,
+      "step": 5177
+    },
+    {
+      "epoch": 1.7291701452663215,
+      "grad_norm": 0.4033600052314757,
+      "learning_rate": 4.539113508943287e-06,
+      "loss": 0.1842,
+      "step": 5178
+    },
+    {
+      "epoch": 1.7295040908331942,
+      "grad_norm": 0.38530844939756476,
+      "learning_rate": 4.537178484005476e-06,
+      "loss": 0.1906,
+      "step": 5179
+    },
+    {
+      "epoch": 1.729838036400067,
+      "grad_norm": 0.4345227695242709,
+      "learning_rate": 4.535243528982305e-06,
+      "loss": 0.2003,
+      "step": 5180
+    },
+    {
+      "epoch": 1.7301719819669394,
+      "grad_norm": 0.38337505336535044,
+      "learning_rate": 4.53330864416607e-06,
+      "loss": 0.1841,
+      "step": 5181
+    },
+    {
+      "epoch": 1.730505927533812,
+      "grad_norm": 0.38990039540591903,
+      "learning_rate": 4.531373829849061e-06,
+      "loss": 0.1947,
+      "step": 5182
+    },
+    {
+      "epoch": 1.7308398731006847,
+      "grad_norm": 0.3929342954019198,
+      "learning_rate": 4.529439086323552e-06,
+      "loss": 0.1867,
+      "step": 5183
+    },
+    {
+      "epoch": 1.7311738186675572,
+      "grad_norm": 0.42907331015138545,
+      "learning_rate": 4.52750441388181e-06,
+      "loss": 0.2045,
+      "step": 5184
+    },
+    {
+      "epoch": 1.7315077642344296,
+      "grad_norm": 0.3957492762261916,
+      "learning_rate": 4.52556981281609e-06,
+      "loss": 0.1965,
+      "step": 5185
+    },
+    {
+      "epoch": 1.7318417098013024,
+      "grad_norm": 0.3985478985780018,
+      "learning_rate": 4.523635283418635e-06,
+      "loss": 0.1836,
+      "step": 5186
+    },
+    {
+      "epoch": 1.732175655368175,
+      "grad_norm": 0.4100016585452721,
+      "learning_rate": 4.521700825981678e-06,
+      "loss": 0.2057,
+      "step": 5187
+    },
+    {
+      "epoch": 1.7325096009350476,
+      "grad_norm": 0.38648895139654443,
+      "learning_rate": 4.519766440797446e-06,
+      "loss": 0.1998,
+      "step": 5188
+    },
+    {
+      "epoch": 1.73284354650192,
+      "grad_norm": 0.3950585537509392,
+      "learning_rate": 4.517832128158147e-06,
+      "loss": 0.1954,
+      "step": 5189
+    },
+    {
+      "epoch": 1.7331774920687928,
+      "grad_norm": 0.351655257356069,
+      "learning_rate": 4.515897888355982e-06,
+      "loss": 0.1701,
+      "step": 5190
+    },
+    {
+      "epoch": 1.7335114376356655,
+      "grad_norm": 0.39426784941225645,
+      "learning_rate": 4.513963721683142e-06,
+      "loss": 0.1933,
+      "step": 5191
+    },
+    {
+      "epoch": 1.733845383202538,
+      "grad_norm": 0.4069564879031356,
+      "learning_rate": 4.5120296284318035e-06,
+      "loss": 0.1839,
+      "step": 5192
+    },
+    {
+      "epoch": 1.7341793287694105,
+      "grad_norm": 0.3679036421194138,
+      "learning_rate": 4.510095608894134e-06,
+      "loss": 0.1784,
+      "step": 5193
+    },
+    {
+      "epoch": 1.7345132743362832,
+      "grad_norm": 0.41535645901622537,
+      "learning_rate": 4.508161663362294e-06,
+      "loss": 0.2009,
+      "step": 5194
+    },
+    {
+      "epoch": 1.734847219903156,
+      "grad_norm": 0.4320307397620132,
+      "learning_rate": 4.506227792128424e-06,
+      "loss": 0.2062,
+      "step": 5195
+    },
+    {
+      "epoch": 1.7351811654700284,
+      "grad_norm": 0.42447532292063683,
+      "learning_rate": 4.504293995484662e-06,
+      "loss": 0.2063,
+      "step": 5196
+    },
+    {
+      "epoch": 1.735515111036901,
+      "grad_norm": 0.4315851980720862,
+      "learning_rate": 4.502360273723127e-06,
+      "loss": 0.2087,
+      "step": 5197
+    },
+    {
+      "epoch": 1.7358490566037736,
+      "grad_norm": 0.40956272992571285,
+      "learning_rate": 4.500426627135933e-06,
+      "loss": 0.187,
+      "step": 5198
+    },
+    {
+      "epoch": 1.7361830021706461,
+      "grad_norm": 0.4529666235929845,
+      "learning_rate": 4.4984930560151776e-06,
+      "loss": 0.1871,
+      "step": 5199
+    },
+    {
+      "epoch": 1.7365169477375186,
+      "grad_norm": 0.38754524741812935,
+      "learning_rate": 4.496559560652952e-06,
+      "loss": 0.1914,
+      "step": 5200
+    },
+    {
+      "epoch": 1.7368508933043914,
+      "grad_norm": 0.43465984631819043,
+      "learning_rate": 4.494626141341334e-06,
+      "loss": 0.1997,
+      "step": 5201
+    },
+    {
+      "epoch": 1.737184838871264,
+      "grad_norm": 0.383172124051344,
+      "learning_rate": 4.4926927983723876e-06,
+      "loss": 0.1932,
+      "step": 5202
+    },
+    {
+      "epoch": 1.7375187844381366,
+      "grad_norm": 0.4201088959332936,
+      "learning_rate": 4.490759532038166e-06,
+      "loss": 0.1946,
+      "step": 5203
+    },
+    {
+      "epoch": 1.737852730005009,
+      "grad_norm": 0.4688579896439726,
+      "learning_rate": 4.488826342630714e-06,
+      "loss": 0.2246,
+      "step": 5204
+    },
+    {
+      "epoch": 1.7381866755718818,
+      "grad_norm": 0.4749846175120122,
+      "learning_rate": 4.486893230442062e-06,
+      "loss": 0.2024,
+      "step": 5205
+    },
+    {
+      "epoch": 1.7385206211387545,
+      "grad_norm": 0.38787853638285946,
+      "learning_rate": 4.4849601957642295e-06,
+      "loss": 0.186,
+      "step": 5206
+    },
+    {
+      "epoch": 1.738854566705627,
+      "grad_norm": 0.38594915412595865,
+      "learning_rate": 4.483027238889223e-06,
+      "loss": 0.193,
+      "step": 5207
+    },
+    {
+      "epoch": 1.7391885122724995,
+      "grad_norm": 0.4113339936370685,
+      "learning_rate": 4.48109436010904e-06,
+      "loss": 0.1912,
+      "step": 5208
+    },
+    {
+      "epoch": 1.7395224578393722,
+      "grad_norm": 0.4094567096226062,
+      "learning_rate": 4.4791615597156635e-06,
+      "loss": 0.1948,
+      "step": 5209
+    },
+    {
+      "epoch": 1.739856403406245,
+      "grad_norm": 0.41900250037341225,
+      "learning_rate": 4.477228838001065e-06,
+      "loss": 0.1891,
+      "step": 5210
+    },
+    {
+      "epoch": 1.7401903489731174,
+      "grad_norm": 0.38308421909389073,
+      "learning_rate": 4.475296195257206e-06,
+      "loss": 0.1855,
+      "step": 5211
+    },
+    {
+      "epoch": 1.74052429453999,
+      "grad_norm": 0.38927592333160743,
+      "learning_rate": 4.4733636317760365e-06,
+      "loss": 0.1929,
+      "step": 5212
+    },
+    {
+      "epoch": 1.7408582401068626,
+      "grad_norm": 0.3896513984545134,
+      "learning_rate": 4.471431147849491e-06,
+      "loss": 0.1808,
+      "step": 5213
+    },
+    {
+      "epoch": 1.7411921856737351,
+      "grad_norm": 0.4339601478317658,
+      "learning_rate": 4.469498743769493e-06,
+      "loss": 0.2101,
+      "step": 5214
+    },
+    {
+      "epoch": 1.7415261312406076,
+      "grad_norm": 0.4446663564844156,
+      "learning_rate": 4.467566419827958e-06,
+      "loss": 0.1946,
+      "step": 5215
+    },
+    {
+      "epoch": 1.7418600768074803,
+      "grad_norm": 0.42455586753823066,
+      "learning_rate": 4.465634176316782e-06,
+      "loss": 0.1935,
+      "step": 5216
+    },
+    {
+      "epoch": 1.742194022374353,
+      "grad_norm": 0.39397487594186725,
+      "learning_rate": 4.463702013527857e-06,
+      "loss": 0.1798,
+      "step": 5217
+    },
+    {
+      "epoch": 1.7425279679412256,
+      "grad_norm": 0.44564863513382924,
+      "learning_rate": 4.4617699317530585e-06,
+      "loss": 0.209,
+      "step": 5218
+    },
+    {
+      "epoch": 1.742861913508098,
+      "grad_norm": 0.3815517649159005,
+      "learning_rate": 4.459837931284249e-06,
+      "loss": 0.1835,
+      "step": 5219
+    },
+    {
+      "epoch": 1.7431958590749708,
+      "grad_norm": 0.37966203750616717,
+      "learning_rate": 4.45790601241328e-06,
+      "loss": 0.1842,
+      "step": 5220
+    },
+    {
+      "epoch": 1.7435298046418435,
+      "grad_norm": 0.3877934659273845,
+      "learning_rate": 4.45597417543199e-06,
+      "loss": 0.1852,
+      "step": 5221
+    },
+    {
+      "epoch": 1.743863750208716,
+      "grad_norm": 0.41155888783147637,
+      "learning_rate": 4.454042420632206e-06,
+      "loss": 0.1863,
+      "step": 5222
+    },
+    {
+      "epoch": 1.7441976957755885,
+      "grad_norm": 0.405549495330632,
+      "learning_rate": 4.452110748305744e-06,
+      "loss": 0.1925,
+      "step": 5223
+    },
+    {
+      "epoch": 1.7445316413424612,
+      "grad_norm": 0.37776272889626006,
+      "learning_rate": 4.450179158744405e-06,
+      "loss": 0.1897,
+      "step": 5224
+    },
+    {
+      "epoch": 1.744865586909334,
+      "grad_norm": 0.41244283018850775,
+      "learning_rate": 4.448247652239978e-06,
+      "loss": 0.1966,
+      "step": 5225
+    },
+    {
+      "epoch": 1.7451995324762064,
+      "grad_norm": 0.37070906739068943,
+      "learning_rate": 4.4463162290842395e-06,
+      "loss": 0.185,
+      "step": 5226
+    },
+    {
+      "epoch": 1.745533478043079,
+      "grad_norm": 0.38778365798313136,
+      "learning_rate": 4.444384889568954e-06,
+      "loss": 0.1885,
+      "step": 5227
+    },
+    {
+      "epoch": 1.7458674236099516,
+      "grad_norm": 0.42132640207261324,
+      "learning_rate": 4.442453633985872e-06,
+      "loss": 0.1996,
+      "step": 5228
+    },
+    {
+      "epoch": 1.7462013691768243,
+      "grad_norm": 0.4647696066279634,
+      "learning_rate": 4.4405224626267345e-06,
+      "loss": 0.2121,
+      "step": 5229
+    },
+    {
+      "epoch": 1.7465353147436968,
+      "grad_norm": 0.40648419947551384,
+      "learning_rate": 4.438591375783267e-06,
+      "loss": 0.1898,
+      "step": 5230
+    },
+    {
+      "epoch": 1.7468692603105693,
+      "grad_norm": 0.48398101512288694,
+      "learning_rate": 4.4366603737471825e-06,
+      "loss": 0.2041,
+      "step": 5231
+    },
+    {
+      "epoch": 1.747203205877442,
+      "grad_norm": 0.3757616353194717,
+      "learning_rate": 4.434729456810182e-06,
+      "loss": 0.1725,
+      "step": 5232
+    },
+    {
+      "epoch": 1.7475371514443145,
+      "grad_norm": 0.439178547747858,
+      "learning_rate": 4.432798625263951e-06,
+      "loss": 0.2103,
+      "step": 5233
+    },
+    {
+      "epoch": 1.747871097011187,
+      "grad_norm": 0.3795864733687818,
+      "learning_rate": 4.430867879400167e-06,
+      "loss": 0.1806,
+      "step": 5234
+    },
+    {
+      "epoch": 1.7482050425780598,
+      "grad_norm": 0.41932040849791336,
+      "learning_rate": 4.428937219510491e-06,
+      "loss": 0.1955,
+      "step": 5235
+    },
+    {
+      "epoch": 1.7485389881449325,
+      "grad_norm": 0.4871060941331514,
+      "learning_rate": 4.427006645886573e-06,
+      "loss": 0.1906,
+      "step": 5236
+    },
+    {
+      "epoch": 1.748872933711805,
+      "grad_norm": 0.35978985995864865,
+      "learning_rate": 4.425076158820048e-06,
+      "loss": 0.1763,
+      "step": 5237
+    },
+    {
+      "epoch": 1.7492068792786775,
+      "grad_norm": 0.4361636932415916,
+      "learning_rate": 4.423145758602538e-06,
+      "loss": 0.1901,
+      "step": 5238
+    },
+    {
+      "epoch": 1.7495408248455502,
+      "grad_norm": 0.3897242187992088,
+      "learning_rate": 4.4212154455256535e-06,
+      "loss": 0.1854,
+      "step": 5239
+    },
+    {
+      "epoch": 1.749874770412423,
+      "grad_norm": 0.44303703845267467,
+      "learning_rate": 4.41928521988099e-06,
+      "loss": 0.2072,
+      "step": 5240
+    },
+    {
+      "epoch": 1.7502087159792954,
+      "grad_norm": 0.38297608026027435,
+      "learning_rate": 4.417355081960133e-06,
+      "loss": 0.1852,
+      "step": 5241
+    },
+    {
+      "epoch": 1.7505426615461679,
+      "grad_norm": 0.4035320423312304,
+      "learning_rate": 4.415425032054651e-06,
+      "loss": 0.1974,
+      "step": 5242
+    },
+    {
+      "epoch": 1.7508766071130406,
+      "grad_norm": 0.4211405489724584,
+      "learning_rate": 4.413495070456101e-06,
+      "loss": 0.2007,
+      "step": 5243
+    },
+    {
+      "epoch": 1.7512105526799133,
+      "grad_norm": 0.3643499101190018,
+      "learning_rate": 4.411565197456027e-06,
+      "loss": 0.185,
+      "step": 5244
+    },
+    {
+      "epoch": 1.7515444982467858,
+      "grad_norm": 0.411108277805822,
+      "learning_rate": 4.409635413345956e-06,
+      "loss": 0.2048,
+      "step": 5245
+    },
+    {
+      "epoch": 1.7518784438136583,
+      "grad_norm": 0.4194151678676051,
+      "learning_rate": 4.40770571841741e-06,
+      "loss": 0.1999,
+      "step": 5246
+    },
+    {
+      "epoch": 1.752212389380531,
+      "grad_norm": 0.40680894494308045,
+      "learning_rate": 4.405776112961889e-06,
+      "loss": 0.2074,
+      "step": 5247
+    },
+    {
+      "epoch": 1.7525463349474035,
+      "grad_norm": 0.41101451985867804,
+      "learning_rate": 4.4038465972708824e-06,
+      "loss": 0.2035,
+      "step": 5248
+    },
+    {
+      "epoch": 1.752880280514276,
+      "grad_norm": 0.4094941833906081,
+      "learning_rate": 4.4019171716358675e-06,
+      "loss": 0.2125,
+      "step": 5249
+    },
+    {
+      "epoch": 1.7532142260811487,
+      "grad_norm": 0.47029820090748287,
+      "learning_rate": 4.399987836348305e-06,
+      "loss": 0.2013,
+      "step": 5250
+    },
+    {
+      "epoch": 1.7535481716480215,
+      "grad_norm": 0.436547364677873,
+      "learning_rate": 4.398058591699645e-06,
+      "loss": 0.2054,
+      "step": 5251
+    },
+    {
+      "epoch": 1.753882117214894,
+      "grad_norm": 0.4223790978332405,
+      "learning_rate": 4.396129437981322e-06,
+      "loss": 0.2078,
+      "step": 5252
+    },
+    {
+      "epoch": 1.7542160627817664,
+      "grad_norm": 0.3751825180830145,
+      "learning_rate": 4.394200375484758e-06,
+      "loss": 0.1807,
+      "step": 5253
+    },
+    {
+      "epoch": 1.7545500083486392,
+      "grad_norm": 0.41355637546032376,
+      "learning_rate": 4.392271404501361e-06,
+      "loss": 0.1946,
+      "step": 5254
+    },
+    {
+      "epoch": 1.7548839539155119,
+      "grad_norm": 0.4283757551832914,
+      "learning_rate": 4.390342525322524e-06,
+      "loss": 0.1935,
+      "step": 5255
+    },
+    {
+      "epoch": 1.7552178994823844,
+      "grad_norm": 0.3614554875916809,
+      "learning_rate": 4.3884137382396255e-06,
+      "loss": 0.1699,
+      "step": 5256
+    },
+    {
+      "epoch": 1.7555518450492569,
+      "grad_norm": 0.4100290180136812,
+      "learning_rate": 4.3864850435440335e-06,
+      "loss": 0.1949,
+      "step": 5257
+    },
+    {
+      "epoch": 1.7558857906161296,
+      "grad_norm": 0.4447808535436557,
+      "learning_rate": 4.3845564415271e-06,
+      "loss": 0.1973,
+      "step": 5258
+    },
+    {
+      "epoch": 1.7562197361830023,
+      "grad_norm": 0.38742859585948136,
+      "learning_rate": 4.382627932480164e-06,
+      "loss": 0.1864,
+      "step": 5259
+    },
+    {
+      "epoch": 1.7565536817498748,
+      "grad_norm": 0.4042603667108564,
+      "learning_rate": 4.380699516694547e-06,
+      "loss": 0.1927,
+      "step": 5260
+    },
+    {
+      "epoch": 1.7568876273167473,
+      "grad_norm": 0.5355154040253798,
+      "learning_rate": 4.37877119446156e-06,
+      "loss": 0.2128,
+      "step": 5261
+    },
+    {
+      "epoch": 1.75722157288362,
+      "grad_norm": 0.41543428507131214,
+      "learning_rate": 4.3768429660725e-06,
+      "loss": 0.1852,
+      "step": 5262
+    },
+    {
+      "epoch": 1.7575555184504925,
+      "grad_norm": 0.39428337573752037,
+      "learning_rate": 4.374914831818643e-06,
+      "loss": 0.1965,
+      "step": 5263
+    },
+    {
+      "epoch": 1.757889464017365,
+      "grad_norm": 0.4158653114188019,
+      "learning_rate": 4.372986791991265e-06,
+      "loss": 0.1917,
+      "step": 5264
+    },
+    {
+      "epoch": 1.7582234095842377,
+      "grad_norm": 0.3700405425476926,
+      "learning_rate": 4.371058846881614e-06,
+      "loss": 0.1747,
+      "step": 5265
+    },
+    {
+      "epoch": 1.7585573551511104,
+      "grad_norm": 0.41343759916271333,
+      "learning_rate": 4.36913099678093e-06,
+      "loss": 0.2013,
+      "step": 5266
+    },
+    {
+      "epoch": 1.758891300717983,
+      "grad_norm": 0.42108912822588007,
+      "learning_rate": 4.367203241980437e-06,
+      "loss": 0.1909,
+      "step": 5267
+    },
+    {
+      "epoch": 1.7592252462848554,
+      "grad_norm": 0.3808692844158381,
+      "learning_rate": 4.3652755827713456e-06,
+      "loss": 0.1827,
+      "step": 5268
+    },
+    {
+      "epoch": 1.7595591918517282,
+      "grad_norm": 0.394990213378632,
+      "learning_rate": 4.363348019444848e-06,
+      "loss": 0.1926,
+      "step": 5269
+    },
+    {
+      "epoch": 1.7598931374186009,
+      "grad_norm": 0.38402085666635427,
+      "learning_rate": 4.361420552292132e-06,
+      "loss": 0.1904,
+      "step": 5270
+    },
+    {
+      "epoch": 1.7602270829854734,
+      "grad_norm": 0.34637884058258056,
+      "learning_rate": 4.35949318160436e-06,
+      "loss": 0.1812,
+      "step": 5271
+    },
+    {
+      "epoch": 1.7605610285523459,
+      "grad_norm": 0.3776613811676968,
+      "learning_rate": 4.357565907672684e-06,
+      "loss": 0.1903,
+      "step": 5272
+    },
+    {
+      "epoch": 1.7608949741192186,
+      "grad_norm": 0.4087198495663909,
+      "learning_rate": 4.355638730788242e-06,
+      "loss": 0.1876,
+      "step": 5273
+    },
+    {
+      "epoch": 1.7612289196860913,
+      "grad_norm": 0.38778910053833804,
+      "learning_rate": 4.353711651242157e-06,
+      "loss": 0.1813,
+      "step": 5274
+    },
+    {
+      "epoch": 1.7615628652529638,
+      "grad_norm": 0.37325883749294486,
+      "learning_rate": 4.3517846693255365e-06,
+      "loss": 0.1777,
+      "step": 5275
+    },
+    {
+      "epoch": 1.7618968108198363,
+      "grad_norm": 0.39984608759219054,
+      "learning_rate": 4.349857785329475e-06,
+      "loss": 0.1923,
+      "step": 5276
+    },
+    {
+      "epoch": 1.762230756386709,
+      "grad_norm": 0.3743519628081095,
+      "learning_rate": 4.34793099954505e-06,
+      "loss": 0.1841,
+      "step": 5277
+    },
+    {
+      "epoch": 1.7625647019535817,
+      "grad_norm": 0.3955250843428749,
+      "learning_rate": 4.3460043122633256e-06,
+      "loss": 0.1899,
+      "step": 5278
+    },
+    {
+      "epoch": 1.7628986475204542,
+      "grad_norm": 0.37322793088236206,
+      "learning_rate": 4.344077723775349e-06,
+      "loss": 0.1707,
+      "step": 5279
+    },
+    {
+      "epoch": 1.7632325930873267,
+      "grad_norm": 0.42561651349981755,
+      "learning_rate": 4.342151234372155e-06,
+      "loss": 0.2092,
+      "step": 5280
+    },
+    {
+      "epoch": 1.7635665386541994,
+      "grad_norm": 0.42631843408788267,
+      "learning_rate": 4.340224844344766e-06,
+      "loss": 0.2021,
+      "step": 5281
+    },
+    {
+      "epoch": 1.763900484221072,
+      "grad_norm": 0.40817135112216674,
+      "learning_rate": 4.338298553984181e-06,
+      "loss": 0.2068,
+      "step": 5282
+    },
+    {
+      "epoch": 1.7642344297879444,
+      "grad_norm": 0.39472505278629344,
+      "learning_rate": 4.336372363581391e-06,
+      "loss": 0.1853,
+      "step": 5283
+    },
+    {
+      "epoch": 1.7645683753548171,
+      "grad_norm": 0.4000097483831296,
+      "learning_rate": 4.33444627342737e-06,
+      "loss": 0.1919,
+      "step": 5284
+    },
+    {
+      "epoch": 1.7649023209216899,
+      "grad_norm": 0.42574844971222253,
+      "learning_rate": 4.332520283813075e-06,
+      "loss": 0.2056,
+      "step": 5285
+    },
+    {
+      "epoch": 1.7652362664885624,
+      "grad_norm": 0.4176968980323195,
+      "learning_rate": 4.330594395029449e-06,
+      "loss": 0.2053,
+      "step": 5286
+    },
+    {
+      "epoch": 1.7655702120554349,
+      "grad_norm": 0.41190325772011954,
+      "learning_rate": 4.328668607367424e-06,
+      "loss": 0.1903,
+      "step": 5287
+    },
+    {
+      "epoch": 1.7659041576223076,
+      "grad_norm": 0.4500552732783806,
+      "learning_rate": 4.326742921117911e-06,
+      "loss": 0.1977,
+      "step": 5288
+    },
+    {
+      "epoch": 1.7662381031891803,
+      "grad_norm": 0.451776540064002,
+      "learning_rate": 4.324817336571806e-06,
+      "loss": 0.2023,
+      "step": 5289
+    },
+    {
+      "epoch": 1.7665720487560528,
+      "grad_norm": 0.4188013135031576,
+      "learning_rate": 4.3228918540199926e-06,
+      "loss": 0.1844,
+      "step": 5290
+    },
+    {
+      "epoch": 1.7669059943229253,
+      "grad_norm": 0.3839938487428742,
+      "learning_rate": 4.320966473753337e-06,
+      "loss": 0.1842,
+      "step": 5291
+    },
+    {
+      "epoch": 1.767239939889798,
+      "grad_norm": 0.41400365926676513,
+      "learning_rate": 4.31904119606269e-06,
+      "loss": 0.1974,
+      "step": 5292
+    },
+    {
+      "epoch": 1.7675738854566707,
+      "grad_norm": 0.420903637964543,
+      "learning_rate": 4.31711602123889e-06,
+      "loss": 0.1956,
+      "step": 5293
+    },
+    {
+      "epoch": 1.7679078310235432,
+      "grad_norm": 0.3933732864451317,
+      "learning_rate": 4.315190949572755e-06,
+      "loss": 0.181,
+      "step": 5294
+    },
+    {
+      "epoch": 1.7682417765904157,
+      "grad_norm": 0.4002834826973489,
+      "learning_rate": 4.313265981355091e-06,
+      "loss": 0.1897,
+      "step": 5295
+    },
+    {
+      "epoch": 1.7685757221572884,
+      "grad_norm": 0.4534917662589281,
+      "learning_rate": 4.311341116876687e-06,
+      "loss": 0.1912,
+      "step": 5296
+    },
+    {
+      "epoch": 1.768909667724161,
+      "grad_norm": 0.3953727715420872,
+      "learning_rate": 4.309416356428315e-06,
+      "loss": 0.192,
+      "step": 5297
+    },
+    {
+      "epoch": 1.7692436132910334,
+      "grad_norm": 0.4235110543403725,
+      "learning_rate": 4.307491700300733e-06,
+      "loss": 0.2054,
+      "step": 5298
+    },
+    {
+      "epoch": 1.7695775588579061,
+      "grad_norm": 0.39646132981498244,
+      "learning_rate": 4.305567148784685e-06,
+      "loss": 0.195,
+      "step": 5299
+    },
+    {
+      "epoch": 1.7699115044247788,
+      "grad_norm": 0.373965105972047,
+      "learning_rate": 4.3036427021708955e-06,
+      "loss": 0.1876,
+      "step": 5300
+    },
+    {
+      "epoch": 1.7702454499916513,
+      "grad_norm": 0.3879823479267765,
+      "learning_rate": 4.301718360750074e-06,
+      "loss": 0.181,
+      "step": 5301
+    },
+    {
+      "epoch": 1.7705793955585238,
+      "grad_norm": 0.3861311454658933,
+      "learning_rate": 4.299794124812918e-06,
+      "loss": 0.1786,
+      "step": 5302
+    },
+    {
+      "epoch": 1.7709133411253966,
+      "grad_norm": 0.41958134378619033,
+      "learning_rate": 4.297869994650103e-06,
+      "loss": 0.2081,
+      "step": 5303
+    },
+    {
+      "epoch": 1.7712472866922693,
+      "grad_norm": 0.39575198482226387,
+      "learning_rate": 4.295945970552293e-06,
+      "loss": 0.1834,
+      "step": 5304
+    },
+    {
+      "epoch": 1.7715812322591418,
+      "grad_norm": 0.3988296719287663,
+      "learning_rate": 4.294022052810134e-06,
+      "loss": 0.1858,
+      "step": 5305
+    },
+    {
+      "epoch": 1.7719151778260143,
+      "grad_norm": 0.4327878844010829,
+      "learning_rate": 4.292098241714256e-06,
+      "loss": 0.1995,
+      "step": 5306
+    },
+    {
+      "epoch": 1.772249123392887,
+      "grad_norm": 0.3883451289026303,
+      "learning_rate": 4.290174537555275e-06,
+      "loss": 0.1912,
+      "step": 5307
+    },
+    {
+      "epoch": 1.7725830689597597,
+      "grad_norm": 0.45477611232921417,
+      "learning_rate": 4.2882509406237885e-06,
+      "loss": 0.1849,
+      "step": 5308
+    },
+    {
+      "epoch": 1.7729170145266322,
+      "grad_norm": 0.38385530123430495,
+      "learning_rate": 4.286327451210377e-06,
+      "loss": 0.1814,
+      "step": 5309
+    },
+    {
+      "epoch": 1.7732509600935047,
+      "grad_norm": 0.4346218793838651,
+      "learning_rate": 4.284404069605605e-06,
+      "loss": 0.1986,
+      "step": 5310
+    },
+    {
+      "epoch": 1.7735849056603774,
+      "grad_norm": 0.36896709229326324,
+      "learning_rate": 4.282480796100027e-06,
+      "loss": 0.1848,
+      "step": 5311
+    },
+    {
+      "epoch": 1.77391885122725,
+      "grad_norm": 0.3921476392351792,
+      "learning_rate": 4.280557630984173e-06,
+      "loss": 0.1888,
+      "step": 5312
+    },
+    {
+      "epoch": 1.7742527967941224,
+      "grad_norm": 0.4206459707979227,
+      "learning_rate": 4.27863457454856e-06,
+      "loss": 0.2004,
+      "step": 5313
+    },
+    {
+      "epoch": 1.7745867423609951,
+      "grad_norm": 0.36808858668646643,
+      "learning_rate": 4.276711627083688e-06,
+      "loss": 0.177,
+      "step": 5314
+    },
+    {
+      "epoch": 1.7749206879278678,
+      "grad_norm": 0.4227954793051248,
+      "learning_rate": 4.274788788880041e-06,
+      "loss": 0.2021,
+      "step": 5315
+    },
+    {
+      "epoch": 1.7752546334947403,
+      "grad_norm": 0.4271760624815115,
+      "learning_rate": 4.272866060228084e-06,
+      "loss": 0.2035,
+      "step": 5316
+    },
+    {
+      "epoch": 1.7755885790616128,
+      "grad_norm": 0.38758325418151324,
+      "learning_rate": 4.270943441418275e-06,
+      "loss": 0.1882,
+      "step": 5317
+    },
+    {
+      "epoch": 1.7759225246284855,
+      "grad_norm": 0.38762943385431875,
+      "learning_rate": 4.2690209327410406e-06,
+      "loss": 0.1907,
+      "step": 5318
+    },
+    {
+      "epoch": 1.7762564701953583,
+      "grad_norm": 0.4139981860338518,
+      "learning_rate": 4.267098534486803e-06,
+      "loss": 0.2032,
+      "step": 5319
+    },
+    {
+      "epoch": 1.7765904157622308,
+      "grad_norm": 0.4167718701052845,
+      "learning_rate": 4.26517624694596e-06,
+      "loss": 0.1997,
+      "step": 5320
+    },
+    {
+      "epoch": 1.7769243613291033,
+      "grad_norm": 0.3827532636411078,
+      "learning_rate": 4.2632540704088975e-06,
+      "loss": 0.1938,
+      "step": 5321
+    },
+    {
+      "epoch": 1.777258306895976,
+      "grad_norm": 0.40170330200281723,
+      "learning_rate": 4.261332005165984e-06,
+      "loss": 0.1712,
+      "step": 5322
+    },
+    {
+      "epoch": 1.7775922524628487,
+      "grad_norm": 0.39802116363666706,
+      "learning_rate": 4.259410051507567e-06,
+      "loss": 0.1896,
+      "step": 5323
+    },
+    {
+      "epoch": 1.7779261980297212,
+      "grad_norm": 0.4222010197604959,
+      "learning_rate": 4.257488209723981e-06,
+      "loss": 0.207,
+      "step": 5324
+    },
+    {
+      "epoch": 1.7782601435965937,
+      "grad_norm": 0.40495532187068245,
+      "learning_rate": 4.255566480105546e-06,
+      "loss": 0.1886,
+      "step": 5325
+    },
+    {
+      "epoch": 1.7785940891634664,
+      "grad_norm": 0.4075549108054149,
+      "learning_rate": 4.2536448629425585e-06,
+      "loss": 0.1796,
+      "step": 5326
+    },
+    {
+      "epoch": 1.7789280347303391,
+      "grad_norm": 0.37713030151299365,
+      "learning_rate": 4.2517233585253024e-06,
+      "loss": 0.1787,
+      "step": 5327
+    },
+    {
+      "epoch": 1.7792619802972116,
+      "grad_norm": 0.4343255349139648,
+      "learning_rate": 4.2498019671440435e-06,
+      "loss": 0.2084,
+      "step": 5328
+    },
+    {
+      "epoch": 1.779595925864084,
+      "grad_norm": 0.39420471849926414,
+      "learning_rate": 4.247880689089033e-06,
+      "loss": 0.1913,
+      "step": 5329
+    },
+    {
+      "epoch": 1.7799298714309568,
+      "grad_norm": 0.3672641930743383,
+      "learning_rate": 4.245959524650498e-06,
+      "loss": 0.1757,
+      "step": 5330
+    },
+    {
+      "epoch": 1.7802638169978293,
+      "grad_norm": 0.3904849887413555,
+      "learning_rate": 4.244038474118656e-06,
+      "loss": 0.1815,
+      "step": 5331
+    },
+    {
+      "epoch": 1.7805977625647018,
+      "grad_norm": 0.3761970196437007,
+      "learning_rate": 4.242117537783704e-06,
+      "loss": 0.1794,
+      "step": 5332
+    },
+    {
+      "epoch": 1.7809317081315745,
+      "grad_norm": 0.4138094357366333,
+      "learning_rate": 4.2401967159358195e-06,
+      "loss": 0.1933,
+      "step": 5333
+    },
+    {
+      "epoch": 1.7812656536984472,
+      "grad_norm": 0.4377054814288653,
+      "learning_rate": 4.2382760088651696e-06,
+      "loss": 0.2075,
+      "step": 5334
+    },
+    {
+      "epoch": 1.7815995992653197,
+      "grad_norm": 0.44167950165051495,
+      "learning_rate": 4.236355416861897e-06,
+      "loss": 0.2025,
+      "step": 5335
+    },
+    {
+      "epoch": 1.7819335448321922,
+      "grad_norm": 0.40362351633119864,
+      "learning_rate": 4.23443494021613e-06,
+      "loss": 0.192,
+      "step": 5336
+    },
+    {
+      "epoch": 1.782267490399065,
+      "grad_norm": 0.4000054483623267,
+      "learning_rate": 4.232514579217981e-06,
+      "loss": 0.1863,
+      "step": 5337
+    },
+    {
+      "epoch": 1.7826014359659377,
+      "grad_norm": 0.39822312077850025,
+      "learning_rate": 4.23059433415754e-06,
+      "loss": 0.1915,
+      "step": 5338
+    },
+    {
+      "epoch": 1.7829353815328102,
+      "grad_norm": 0.41333564470960643,
+      "learning_rate": 4.228674205324884e-06,
+      "loss": 0.1986,
+      "step": 5339
+    },
+    {
+      "epoch": 1.7832693270996827,
+      "grad_norm": 0.40541425631144945,
+      "learning_rate": 4.226754193010072e-06,
+      "loss": 0.1819,
+      "step": 5340
+    },
+    {
+      "epoch": 1.7836032726665554,
+      "grad_norm": 0.36616267473432135,
+      "learning_rate": 4.224834297503145e-06,
+      "loss": 0.1805,
+      "step": 5341
+    },
+    {
+      "epoch": 1.783937218233428,
+      "grad_norm": 0.4368241980696301,
+      "learning_rate": 4.222914519094124e-06,
+      "loss": 0.2082,
+      "step": 5342
+    },
+    {
+      "epoch": 1.7842711638003006,
+      "grad_norm": 0.4105369770801003,
+      "learning_rate": 4.220994858073014e-06,
+      "loss": 0.194,
+      "step": 5343
+    },
+    {
+      "epoch": 1.784605109367173,
+      "grad_norm": 0.3959702721339513,
+      "learning_rate": 4.2190753147298044e-06,
+      "loss": 0.1945,
+      "step": 5344
+    },
+    {
+      "epoch": 1.7849390549340458,
+      "grad_norm": 0.47707382396904957,
+      "learning_rate": 4.2171558893544626e-06,
+      "loss": 0.1815,
+      "step": 5345
+    },
+    {
+      "epoch": 1.7852730005009183,
+      "grad_norm": 0.42034793280157745,
+      "learning_rate": 4.215236582236941e-06,
+      "loss": 0.1976,
+      "step": 5346
+    },
+    {
+      "epoch": 1.7856069460677908,
+      "grad_norm": 0.3784248182395078,
+      "learning_rate": 4.213317393667175e-06,
+      "loss": 0.191,
+      "step": 5347
+    },
+    {
+      "epoch": 1.7859408916346635,
+      "grad_norm": 0.3947336613390132,
+      "learning_rate": 4.211398323935079e-06,
+      "loss": 0.1984,
+      "step": 5348
+    },
+    {
+      "epoch": 1.7862748372015362,
+      "grad_norm": 0.4745094890865065,
+      "learning_rate": 4.209479373330552e-06,
+      "loss": 0.2146,
+      "step": 5349
+    },
+    {
+      "epoch": 1.7866087827684087,
+      "grad_norm": 0.4115025463067994,
+      "learning_rate": 4.207560542143474e-06,
+      "loss": 0.2033,
+      "step": 5350
+    },
+    {
+      "epoch": 1.7869427283352812,
+      "grad_norm": 0.41096840682747465,
+      "learning_rate": 4.205641830663706e-06,
+      "loss": 0.1969,
+      "step": 5351
+    },
+    {
+      "epoch": 1.787276673902154,
+      "grad_norm": 0.37721407675737656,
+      "learning_rate": 4.2037232391810925e-06,
+      "loss": 0.1796,
+      "step": 5352
+    },
+    {
+      "epoch": 1.7876106194690267,
+      "grad_norm": 0.3712905961394812,
+      "learning_rate": 4.20180476798546e-06,
+      "loss": 0.1831,
+      "step": 5353
+    },
+    {
+      "epoch": 1.7879445650358992,
+      "grad_norm": 0.41330452273749596,
+      "learning_rate": 4.1998864173666174e-06,
+      "loss": 0.2077,
+      "step": 5354
+    },
+    {
+      "epoch": 1.7882785106027717,
+      "grad_norm": 0.4693570027309341,
+      "learning_rate": 4.197968187614351e-06,
+      "loss": 0.2115,
+      "step": 5355
+    },
+    {
+      "epoch": 1.7886124561696444,
+      "grad_norm": 0.4171457241557042,
+      "learning_rate": 4.196050079018433e-06,
+      "loss": 0.1869,
+      "step": 5356
+    },
+    {
+      "epoch": 1.788946401736517,
+      "grad_norm": 1.0151186082749302,
+      "learning_rate": 4.194132091868616e-06,
+      "loss": 0.2071,
+      "step": 5357
+    },
+    {
+      "epoch": 1.7892803473033896,
+      "grad_norm": 0.4187428989067555,
+      "learning_rate": 4.1922142264546365e-06,
+      "loss": 0.19,
+      "step": 5358
+    },
+    {
+      "epoch": 1.789614292870262,
+      "grad_norm": 0.3836619002670058,
+      "learning_rate": 4.1902964830662104e-06,
+      "loss": 0.1924,
+      "step": 5359
+    },
+    {
+      "epoch": 1.7899482384371348,
+      "grad_norm": 0.38919854186049957,
+      "learning_rate": 4.188378861993034e-06,
+      "loss": 0.1878,
+      "step": 5360
+    },
+    {
+      "epoch": 1.7902821840040073,
+      "grad_norm": 0.39768862778262004,
+      "learning_rate": 4.186461363524786e-06,
+      "loss": 0.198,
+      "step": 5361
+    },
+    {
+      "epoch": 1.7906161295708798,
+      "grad_norm": 0.39782728299988296,
+      "learning_rate": 4.184543987951127e-06,
+      "loss": 0.1801,
+      "step": 5362
+    },
+    {
+      "epoch": 1.7909500751377525,
+      "grad_norm": 0.495707344907115,
+      "learning_rate": 4.182626735561703e-06,
+      "loss": 0.181,
+      "step": 5363
+    },
+    {
+      "epoch": 1.7912840207046252,
+      "grad_norm": 0.44114708561246135,
+      "learning_rate": 4.180709606646134e-06,
+      "loss": 0.2039,
+      "step": 5364
+    },
+    {
+      "epoch": 1.7916179662714977,
+      "grad_norm": 0.3927854303581158,
+      "learning_rate": 4.178792601494026e-06,
+      "loss": 0.1896,
+      "step": 5365
+    },
+    {
+      "epoch": 1.7919519118383702,
+      "grad_norm": 0.421206797477163,
+      "learning_rate": 4.176875720394965e-06,
+      "loss": 0.2005,
+      "step": 5366
+    },
+    {
+      "epoch": 1.792285857405243,
+      "grad_norm": 0.5201455802658564,
+      "learning_rate": 4.174958963638518e-06,
+      "loss": 0.2143,
+      "step": 5367
+    },
+    {
+      "epoch": 1.7926198029721157,
+      "grad_norm": 0.4138358670591043,
+      "learning_rate": 4.173042331514234e-06,
+      "loss": 0.1827,
+      "step": 5368
+    },
+    {
+      "epoch": 1.7929537485389881,
+      "grad_norm": 0.44836688438557204,
+      "learning_rate": 4.171125824311642e-06,
+      "loss": 0.204,
+      "step": 5369
+    },
+    {
+      "epoch": 1.7932876941058606,
+      "grad_norm": 0.39838785544358446,
+      "learning_rate": 4.169209442320255e-06,
+      "loss": 0.1864,
+      "step": 5370
+    },
+    {
+      "epoch": 1.7936216396727334,
+      "grad_norm": 0.39065135717154886,
+      "learning_rate": 4.167293185829565e-06,
+      "loss": 0.1777,
+      "step": 5371
+    },
+    {
+      "epoch": 1.793955585239606,
+      "grad_norm": 0.3868324514391502,
+      "learning_rate": 4.165377055129043e-06,
+      "loss": 0.1898,
+      "step": 5372
+    },
+    {
+      "epoch": 1.7942895308064786,
+      "grad_norm": 0.4086998441668388,
+      "learning_rate": 4.163461050508144e-06,
+      "loss": 0.1874,
+      "step": 5373
+    },
+    {
+      "epoch": 1.794623476373351,
+      "grad_norm": 0.3777662978033055,
+      "learning_rate": 4.161545172256303e-06,
+      "loss": 0.1824,
+      "step": 5374
+    },
+    {
+      "epoch": 1.7949574219402238,
+      "grad_norm": 0.4382436302483726,
+      "learning_rate": 4.1596294206629375e-06,
+      "loss": 0.2086,
+      "step": 5375
+    },
+    {
+      "epoch": 1.7952913675070965,
+      "grad_norm": 0.41758627579949836,
+      "learning_rate": 4.157713796017442e-06,
+      "loss": 0.1874,
+      "step": 5376
+    },
+    {
+      "epoch": 1.795625313073969,
+      "grad_norm": 0.40993854049694717,
+      "learning_rate": 4.155798298609196e-06,
+      "loss": 0.2063,
+      "step": 5377
+    },
+    {
+      "epoch": 1.7959592586408415,
+      "grad_norm": 0.44807089311268883,
+      "learning_rate": 4.1538829287275565e-06,
+      "loss": 0.1874,
+      "step": 5378
+    },
+    {
+      "epoch": 1.7962932042077142,
+      "grad_norm": 0.41081845688238966,
+      "learning_rate": 4.151967686661864e-06,
+      "loss": 0.2055,
+      "step": 5379
+    },
+    {
+      "epoch": 1.7966271497745867,
+      "grad_norm": 0.3750452408616848,
+      "learning_rate": 4.150052572701435e-06,
+      "loss": 0.1923,
+      "step": 5380
+    },
+    {
+      "epoch": 1.7969610953414592,
+      "grad_norm": 0.3748777838197131,
+      "learning_rate": 4.148137587135575e-06,
+      "loss": 0.1856,
+      "step": 5381
+    },
+    {
+      "epoch": 1.797295040908332,
+      "grad_norm": 0.385221778710938,
+      "learning_rate": 4.146222730253563e-06,
+      "loss": 0.1847,
+      "step": 5382
+    },
+    {
+      "epoch": 1.7976289864752046,
+      "grad_norm": 0.4130555094657973,
+      "learning_rate": 4.1443080023446605e-06,
+      "loss": 0.2001,
+      "step": 5383
+    },
+    {
+      "epoch": 1.7979629320420771,
+      "grad_norm": 0.40847407861926593,
+      "learning_rate": 4.1423934036981096e-06,
+      "loss": 0.1968,
+      "step": 5384
+    },
+    {
+      "epoch": 1.7982968776089496,
+      "grad_norm": 0.45331715003457235,
+      "learning_rate": 4.140478934603133e-06,
+      "loss": 0.1954,
+      "step": 5385
+    },
+    {
+      "epoch": 1.7986308231758223,
+      "grad_norm": 0.3854167677351394,
+      "learning_rate": 4.138564595348932e-06,
+      "loss": 0.1847,
+      "step": 5386
+    },
+    {
+      "epoch": 1.798964768742695,
+      "grad_norm": 0.40379130860802437,
+      "learning_rate": 4.136650386224694e-06,
+      "loss": 0.1974,
+      "step": 5387
+    },
+    {
+      "epoch": 1.7992987143095676,
+      "grad_norm": 0.43616884197672473,
+      "learning_rate": 4.13473630751958e-06,
+      "loss": 0.2073,
+      "step": 5388
+    },
+    {
+      "epoch": 1.79963265987644,
+      "grad_norm": 0.39500763107731124,
+      "learning_rate": 4.132822359522735e-06,
+      "loss": 0.1964,
+      "step": 5389
+    },
+    {
+      "epoch": 1.7999666054433128,
+      "grad_norm": 0.5316276814151004,
+      "learning_rate": 4.130908542523285e-06,
+      "loss": 0.2012,
+      "step": 5390
+    },
+    {
+      "epoch": 1.8003005510101855,
+      "grad_norm": 0.4413915607742555,
+      "learning_rate": 4.128994856810332e-06,
+      "loss": 0.1929,
+      "step": 5391
+    },
+    {
+      "epoch": 1.800634496577058,
+      "grad_norm": 0.43359946931826776,
+      "learning_rate": 4.127081302672958e-06,
+      "loss": 0.1975,
+      "step": 5392
+    },
+    {
+      "epoch": 1.8009684421439305,
+      "grad_norm": 0.4326737933869161,
+      "learning_rate": 4.125167880400235e-06,
+      "loss": 0.1979,
+      "step": 5393
+    },
+    {
+      "epoch": 1.8013023877108032,
+      "grad_norm": 0.3999758339829761,
+      "learning_rate": 4.1232545902812046e-06,
+      "loss": 0.1782,
+      "step": 5394
+    },
+    {
+      "epoch": 1.8016363332776757,
+      "grad_norm": 0.39414865979855523,
+      "learning_rate": 4.121341432604892e-06,
+      "loss": 0.1908,
+      "step": 5395
+    },
+    {
+      "epoch": 1.8019702788445482,
+      "grad_norm": 0.40085607899323583,
+      "learning_rate": 4.1194284076603004e-06,
+      "loss": 0.1991,
+      "step": 5396
+    },
+    {
+      "epoch": 1.802304224411421,
+      "grad_norm": 0.3986732428896081,
+      "learning_rate": 4.117515515736418e-06,
+      "loss": 0.1892,
+      "step": 5397
+    },
+    {
+      "epoch": 1.8026381699782936,
+      "grad_norm": 0.38552180630358474,
+      "learning_rate": 4.1156027571222054e-06,
+      "loss": 0.1899,
+      "step": 5398
+    },
+    {
+      "epoch": 1.8029721155451661,
+      "grad_norm": 0.42639374577496114,
+      "learning_rate": 4.113690132106611e-06,
+      "loss": 0.2013,
+      "step": 5399
+    },
+    {
+      "epoch": 1.8033060611120386,
+      "grad_norm": 0.44076136474761535,
+      "learning_rate": 4.111777640978559e-06,
+      "loss": 0.1967,
+      "step": 5400
+    },
+    {
+      "epoch": 1.8036400066789113,
+      "grad_norm": 0.3715616165565932,
+      "learning_rate": 4.109865284026953e-06,
+      "loss": 0.1895,
+      "step": 5401
+    },
+    {
+      "epoch": 1.803973952245784,
+      "grad_norm": 0.413959384805662,
+      "learning_rate": 4.107953061540676e-06,
+      "loss": 0.1906,
+      "step": 5402
+    },
+    {
+      "epoch": 1.8043078978126565,
+      "grad_norm": 0.3986680001400406,
+      "learning_rate": 4.10604097380859e-06,
+      "loss": 0.1884,
+      "step": 5403
+    },
+    {
+      "epoch": 1.804641843379529,
+      "grad_norm": 0.39896994690686616,
+      "learning_rate": 4.104129021119543e-06,
+      "loss": 0.1864,
+      "step": 5404
+    },
+    {
+      "epoch": 1.8049757889464018,
+      "grad_norm": 0.41708702213753196,
+      "learning_rate": 4.102217203762357e-06,
+      "loss": 0.2042,
+      "step": 5405
+    },
+    {
+      "epoch": 1.8053097345132745,
+      "grad_norm": 0.4162521221231452,
+      "learning_rate": 4.1003055220258335e-06,
+      "loss": 0.1886,
+      "step": 5406
+    },
+    {
+      "epoch": 1.805643680080147,
+      "grad_norm": 0.388971195188673,
+      "learning_rate": 4.0983939761987535e-06,
+      "loss": 0.1804,
+      "step": 5407
+    },
+    {
+      "epoch": 1.8059776256470195,
+      "grad_norm": 0.3976488686723864,
+      "learning_rate": 4.09648256656988e-06,
+      "loss": 0.2036,
+      "step": 5408
+    },
+    {
+      "epoch": 1.8063115712138922,
+      "grad_norm": 0.43724410925090645,
+      "learning_rate": 4.094571293427951e-06,
+      "loss": 0.2063,
+      "step": 5409
+    },
+    {
+      "epoch": 1.8066455167807647,
+      "grad_norm": 0.42616248921069183,
+      "learning_rate": 4.092660157061691e-06,
+      "loss": 0.2003,
+      "step": 5410
+    },
+    {
+      "epoch": 1.8069794623476372,
+      "grad_norm": 0.4045940509370612,
+      "learning_rate": 4.090749157759799e-06,
+      "loss": 0.1894,
+      "step": 5411
+    },
+    {
+      "epoch": 1.80731340791451,
+      "grad_norm": 0.4173527749892698,
+      "learning_rate": 4.088838295810952e-06,
+      "loss": 0.2007,
+      "step": 5412
+    },
+    {
+      "epoch": 1.8076473534813826,
+      "grad_norm": 0.42528073955892315,
+      "learning_rate": 4.086927571503808e-06,
+      "loss": 0.2005,
+      "step": 5413
+    },
+    {
+      "epoch": 1.807981299048255,
+      "grad_norm": 0.41587887704528714,
+      "learning_rate": 4.0850169851270075e-06,
+      "loss": 0.2006,
+      "step": 5414
+    },
+    {
+      "epoch": 1.8083152446151276,
+      "grad_norm": 0.40738714104001256,
+      "learning_rate": 4.0831065369691615e-06,
+      "loss": 0.2008,
+      "step": 5415
+    },
+    {
+      "epoch": 1.8086491901820003,
+      "grad_norm": 0.44869021642959594,
+      "learning_rate": 4.0811962273188714e-06,
+      "loss": 0.2001,
+      "step": 5416
+    },
+    {
+      "epoch": 1.808983135748873,
+      "grad_norm": 0.4002553541976238,
+      "learning_rate": 4.0792860564647105e-06,
+      "loss": 0.1911,
+      "step": 5417
+    },
+    {
+      "epoch": 1.8093170813157455,
+      "grad_norm": 0.3786221254197682,
+      "learning_rate": 4.077376024695231e-06,
+      "loss": 0.1774,
+      "step": 5418
+    },
+    {
+      "epoch": 1.809651026882618,
+      "grad_norm": 0.3973451247400723,
+      "learning_rate": 4.075466132298967e-06,
+      "loss": 0.1918,
+      "step": 5419
+    },
+    {
+      "epoch": 1.8099849724494907,
+      "grad_norm": 0.40821244986487804,
+      "learning_rate": 4.073556379564429e-06,
+      "loss": 0.1921,
+      "step": 5420
+    },
+    {
+      "epoch": 1.8103189180163635,
+      "grad_norm": 0.4187248192126581,
+      "learning_rate": 4.071646766780109e-06,
+      "loss": 0.2055,
+      "step": 5421
+    },
+    {
+      "epoch": 1.810652863583236,
+      "grad_norm": 0.4009251034725688,
+      "learning_rate": 4.069737294234475e-06,
+      "loss": 0.1921,
+      "step": 5422
+    },
+    {
+      "epoch": 1.8109868091501085,
+      "grad_norm": 0.4153763020520176,
+      "learning_rate": 4.067827962215977e-06,
+      "loss": 0.1971,
+      "step": 5423
+    },
+    {
+      "epoch": 1.8113207547169812,
+      "grad_norm": 0.42059984630019226,
+      "learning_rate": 4.065918771013042e-06,
+      "loss": 0.1931,
+      "step": 5424
+    },
+    {
+      "epoch": 1.811654700283854,
+      "grad_norm": 0.38975550241965073,
+      "learning_rate": 4.064009720914074e-06,
+      "loss": 0.1939,
+      "step": 5425
+    },
+    {
+      "epoch": 1.8119886458507264,
+      "grad_norm": 0.4352813167969884,
+      "learning_rate": 4.062100812207459e-06,
+      "loss": 0.1984,
+      "step": 5426
+    },
+    {
+      "epoch": 1.8123225914175989,
+      "grad_norm": 0.41222394672549245,
+      "learning_rate": 4.060192045181558e-06,
+      "loss": 0.2003,
+      "step": 5427
+    },
+    {
+      "epoch": 1.8126565369844716,
+      "grad_norm": 0.4211144974261239,
+      "learning_rate": 4.058283420124716e-06,
+      "loss": 0.1946,
+      "step": 5428
+    },
+    {
+      "epoch": 1.812990482551344,
+      "grad_norm": 0.4033307519940219,
+      "learning_rate": 4.056374937325251e-06,
+      "loss": 0.1888,
+      "step": 5429
+    },
+    {
+      "epoch": 1.8133244281182166,
+      "grad_norm": 0.4234575464895144,
+      "learning_rate": 4.054466597071464e-06,
+      "loss": 0.1932,
+      "step": 5430
+    },
+    {
+      "epoch": 1.8136583736850893,
+      "grad_norm": 0.41837807879440336,
+      "learning_rate": 4.05255839965163e-06,
+      "loss": 0.2026,
+      "step": 5431
+    },
+    {
+      "epoch": 1.813992319251962,
+      "grad_norm": 0.43738094694833785,
+      "learning_rate": 4.050650345354006e-06,
+      "loss": 0.202,
+      "step": 5432
+    },
+    {
+      "epoch": 1.8143262648188345,
+      "grad_norm": 0.43660446022144006,
+      "learning_rate": 4.048742434466823e-06,
+      "loss": 0.2109,
+      "step": 5433
+    },
+    {
+      "epoch": 1.814660210385707,
+      "grad_norm": 0.3961921763065722,
+      "learning_rate": 4.046834667278298e-06,
+      "loss": 0.1926,
+      "step": 5434
+    },
+    {
+      "epoch": 1.8149941559525797,
+      "grad_norm": 0.39349069851042634,
+      "learning_rate": 4.04492704407662e-06,
+      "loss": 0.191,
+      "step": 5435
+    },
+    {
+      "epoch": 1.8153281015194525,
+      "grad_norm": 0.40519030905215747,
+      "learning_rate": 4.043019565149958e-06,
+      "loss": 0.1936,
+      "step": 5436
+    },
+    {
+      "epoch": 1.815662047086325,
+      "grad_norm": 0.37874729826316716,
+      "learning_rate": 4.041112230786458e-06,
+      "loss": 0.186,
+      "step": 5437
+    },
+    {
+      "epoch": 1.8159959926531974,
+      "grad_norm": 0.6081203465553392,
+      "learning_rate": 4.039205041274247e-06,
+      "loss": 0.2003,
+      "step": 5438
+    },
+    {
+      "epoch": 1.8163299382200702,
+      "grad_norm": 0.41622622386812613,
+      "learning_rate": 4.0372979969014245e-06,
+      "loss": 0.1904,
+      "step": 5439
+    },
+    {
+      "epoch": 1.8166638837869429,
+      "grad_norm": 0.4075713166770594,
+      "learning_rate": 4.035391097956077e-06,
+      "loss": 0.2012,
+      "step": 5440
+    },
+    {
+      "epoch": 1.8169978293538154,
+      "grad_norm": 0.4494439120686127,
+      "learning_rate": 4.0334843447262625e-06,
+      "loss": 0.2047,
+      "step": 5441
+    },
+    {
+      "epoch": 1.8173317749206879,
+      "grad_norm": 0.4137738402524847,
+      "learning_rate": 4.0315777375000185e-06,
+      "loss": 0.2019,
+      "step": 5442
+    },
+    {
+      "epoch": 1.8176657204875606,
+      "grad_norm": 0.393956419351833,
+      "learning_rate": 4.029671276565359e-06,
+      "loss": 0.1929,
+      "step": 5443
+    },
+    {
+      "epoch": 1.817999666054433,
+      "grad_norm": 0.3960846147046012,
+      "learning_rate": 4.027764962210278e-06,
+      "loss": 0.1828,
+      "step": 5444
+    },
+    {
+      "epoch": 1.8183336116213056,
+      "grad_norm": 0.3877136289033595,
+      "learning_rate": 4.025858794722749e-06,
+      "loss": 0.1908,
+      "step": 5445
+    },
+    {
+      "epoch": 1.8186675571881783,
+      "grad_norm": 0.38011780754177016,
+      "learning_rate": 4.0239527743907184e-06,
+      "loss": 0.1833,
+      "step": 5446
+    },
+    {
+      "epoch": 1.819001502755051,
+      "grad_norm": 0.41199492546924943,
+      "learning_rate": 4.022046901502114e-06,
+      "loss": 0.1817,
+      "step": 5447
+    },
+    {
+      "epoch": 1.8193354483219235,
+      "grad_norm": 0.401329738391802,
+      "learning_rate": 4.020141176344839e-06,
+      "loss": 0.1926,
+      "step": 5448
+    },
+    {
+      "epoch": 1.819669393888796,
+      "grad_norm": 0.41127280691800383,
+      "learning_rate": 4.018235599206778e-06,
+      "loss": 0.2009,
+      "step": 5449
+    },
+    {
+      "epoch": 1.8200033394556687,
+      "grad_norm": 0.4064474207715107,
+      "learning_rate": 4.016330170375787e-06,
+      "loss": 0.1935,
+      "step": 5450
+    },
+    {
+      "epoch": 1.8203372850225414,
+      "grad_norm": 0.4020536829609197,
+      "learning_rate": 4.014424890139709e-06,
+      "loss": 0.1837,
+      "step": 5451
+    },
+    {
+      "epoch": 1.820671230589414,
+      "grad_norm": 0.38407516465676206,
+      "learning_rate": 4.012519758786355e-06,
+      "loss": 0.1913,
+      "step": 5452
+    },
+    {
+      "epoch": 1.8210051761562864,
+      "grad_norm": 0.4123403170436455,
+      "learning_rate": 4.01061477660352e-06,
+      "loss": 0.1976,
+      "step": 5453
+    },
+    {
+      "epoch": 1.8213391217231591,
+      "grad_norm": 0.3839390900902544,
+      "learning_rate": 4.008709943878971e-06,
+      "loss": 0.1836,
+      "step": 5454
+    },
+    {
+      "epoch": 1.8216730672900319,
+      "grad_norm": 0.38868642954562516,
+      "learning_rate": 4.006805260900458e-06,
+      "loss": 0.1857,
+      "step": 5455
+    },
+    {
+      "epoch": 1.8220070128569044,
+      "grad_norm": 0.44277697324688925,
+      "learning_rate": 4.004900727955703e-06,
+      "loss": 0.203,
+      "step": 5456
+    },
+    {
+      "epoch": 1.8223409584237769,
+      "grad_norm": 0.41818991156707475,
+      "learning_rate": 4.0029963453324115e-06,
+      "loss": 0.1891,
+      "step": 5457
+    },
+    {
+      "epoch": 1.8226749039906496,
+      "grad_norm": 0.378807769453953,
+      "learning_rate": 4.001092113318261e-06,
+      "loss": 0.1809,
+      "step": 5458
+    },
+    {
+      "epoch": 1.823008849557522,
+      "grad_norm": 0.4201589268763546,
+      "learning_rate": 3.99918803220091e-06,
+      "loss": 0.2034,
+      "step": 5459
+    },
+    {
+      "epoch": 1.8233427951243946,
+      "grad_norm": 0.3896772152862828,
+      "learning_rate": 3.99728410226799e-06,
+      "loss": 0.1958,
+      "step": 5460
+    },
+    {
+      "epoch": 1.8236767406912673,
+      "grad_norm": 0.3922430722052448,
+      "learning_rate": 3.995380323807113e-06,
+      "loss": 0.1953,
+      "step": 5461
+    },
+    {
+      "epoch": 1.82401068625814,
+      "grad_norm": 0.35224053029208974,
+      "learning_rate": 3.993476697105864e-06,
+      "loss": 0.1756,
+      "step": 5462
+    },
+    {
+      "epoch": 1.8243446318250125,
+      "grad_norm": 0.4117629603867008,
+      "learning_rate": 3.991573222451815e-06,
+      "loss": 0.2055,
+      "step": 5463
+    },
+    {
+      "epoch": 1.824678577391885,
+      "grad_norm": 0.4119526641975354,
+      "learning_rate": 3.989669900132504e-06,
+      "loss": 0.1853,
+      "step": 5464
+    },
+    {
+      "epoch": 1.8250125229587577,
+      "grad_norm": 0.3944045127756812,
+      "learning_rate": 3.987766730435451e-06,
+      "loss": 0.1987,
+      "step": 5465
+    },
+    {
+      "epoch": 1.8253464685256304,
+      "grad_norm": 0.4015804217496364,
+      "learning_rate": 3.9858637136481515e-06,
+      "loss": 0.198,
+      "step": 5466
+    },
+    {
+      "epoch": 1.825680414092503,
+      "grad_norm": 0.37176769181449776,
+      "learning_rate": 3.98396085005808e-06,
+      "loss": 0.1862,
+      "step": 5467
+    },
+    {
+      "epoch": 1.8260143596593754,
+      "grad_norm": 0.40972865853666285,
+      "learning_rate": 3.982058139952684e-06,
+      "loss": 0.1886,
+      "step": 5468
+    },
+    {
+      "epoch": 1.8263483052262481,
+      "grad_norm": 0.3955918184023543,
+      "learning_rate": 3.980155583619392e-06,
+      "loss": 0.1969,
+      "step": 5469
+    },
+    {
+      "epoch": 1.8266822507931209,
+      "grad_norm": 0.3825992680611567,
+      "learning_rate": 3.978253181345609e-06,
+      "loss": 0.1867,
+      "step": 5470
+    },
+    {
+      "epoch": 1.8270161963599933,
+      "grad_norm": 0.38415661050520683,
+      "learning_rate": 3.9763509334187125e-06,
+      "loss": 0.1918,
+      "step": 5471
+    },
+    {
+      "epoch": 1.8273501419268658,
+      "grad_norm": 0.4392047476972209,
+      "learning_rate": 3.974448840126061e-06,
+      "loss": 0.1988,
+      "step": 5472
+    },
+    {
+      "epoch": 1.8276840874937386,
+      "grad_norm": 0.4115656594818337,
+      "learning_rate": 3.972546901754987e-06,
+      "loss": 0.1947,
+      "step": 5473
+    },
+    {
+      "epoch": 1.8280180330606113,
+      "grad_norm": 0.3900375859373049,
+      "learning_rate": 3.9706451185928e-06,
+      "loss": 0.1921,
+      "step": 5474
+    },
+    {
+      "epoch": 1.8283519786274836,
+      "grad_norm": 0.407212324644722,
+      "learning_rate": 3.968743490926791e-06,
+      "loss": 0.1978,
+      "step": 5475
+    },
+    {
+      "epoch": 1.8286859241943563,
+      "grad_norm": 0.3799461405429855,
+      "learning_rate": 3.966842019044219e-06,
+      "loss": 0.1794,
+      "step": 5476
+    },
+    {
+      "epoch": 1.829019869761229,
+      "grad_norm": 0.5359825297417488,
+      "learning_rate": 3.964940703232326e-06,
+      "loss": 0.2117,
+      "step": 5477
+    },
+    {
+      "epoch": 1.8293538153281015,
+      "grad_norm": 0.43912349342902457,
+      "learning_rate": 3.963039543778327e-06,
+      "loss": 0.1951,
+      "step": 5478
+    },
+    {
+      "epoch": 1.829687760894974,
+      "grad_norm": 0.35805460260955724,
+      "learning_rate": 3.961138540969411e-06,
+      "loss": 0.1778,
+      "step": 5479
+    },
+    {
+      "epoch": 1.8300217064618467,
+      "grad_norm": 0.3810672843512081,
+      "learning_rate": 3.9592376950927545e-06,
+      "loss": 0.187,
+      "step": 5480
+    },
+    {
+      "epoch": 1.8303556520287194,
+      "grad_norm": 0.4011484817855961,
+      "learning_rate": 3.957337006435499e-06,
+      "loss": 0.1785,
+      "step": 5481
+    },
+    {
+      "epoch": 1.830689597595592,
+      "grad_norm": 0.4789546571386921,
+      "learning_rate": 3.955436475284764e-06,
+      "loss": 0.2091,
+      "step": 5482
+    },
+    {
+      "epoch": 1.8310235431624644,
+      "grad_norm": 0.4063460886404937,
+      "learning_rate": 3.95353610192765e-06,
+      "loss": 0.1924,
+      "step": 5483
+    },
+    {
+      "epoch": 1.8313574887293371,
+      "grad_norm": 0.41526229249076346,
+      "learning_rate": 3.95163588665123e-06,
+      "loss": 0.1915,
+      "step": 5484
+    },
+    {
+      "epoch": 1.8316914342962098,
+      "grad_norm": 0.41928063093273427,
+      "learning_rate": 3.949735829742549e-06,
+      "loss": 0.2008,
+      "step": 5485
+    },
+    {
+      "epoch": 1.8320253798630823,
+      "grad_norm": 0.41388821065326953,
+      "learning_rate": 3.947835931488642e-06,
+      "loss": 0.1978,
+      "step": 5486
+    },
+    {
+      "epoch": 1.8323593254299548,
+      "grad_norm": 0.42225519497611547,
+      "learning_rate": 3.9459361921765045e-06,
+      "loss": 0.2089,
+      "step": 5487
+    },
+    {
+      "epoch": 1.8326932709968276,
+      "grad_norm": 0.3697474868956762,
+      "learning_rate": 3.944036612093117e-06,
+      "loss": 0.1834,
+      "step": 5488
+    },
+    {
+      "epoch": 1.8330272165637003,
+      "grad_norm": 0.40844338979490125,
+      "learning_rate": 3.942137191525434e-06,
+      "loss": 0.1815,
+      "step": 5489
+    },
+    {
+      "epoch": 1.8333611621305728,
+      "grad_norm": 0.37036000123683893,
+      "learning_rate": 3.9402379307603825e-06,
+      "loss": 0.1757,
+      "step": 5490
+    },
+    {
+      "epoch": 1.8336951076974453,
+      "grad_norm": 0.4416887840594564,
+      "learning_rate": 3.93833883008487e-06,
+      "loss": 0.1928,
+      "step": 5491
+    },
+    {
+      "epoch": 1.834029053264318,
+      "grad_norm": 0.4230338012823816,
+      "learning_rate": 3.936439889785778e-06,
+      "loss": 0.2049,
+      "step": 5492
+    },
+    {
+      "epoch": 1.8343629988311905,
+      "grad_norm": 0.4552356374064481,
+      "learning_rate": 3.934541110149964e-06,
+      "loss": 0.1859,
+      "step": 5493
+    },
+    {
+      "epoch": 1.834696944398063,
+      "grad_norm": 0.3996370964740024,
+      "learning_rate": 3.932642491464261e-06,
+      "loss": 0.1863,
+      "step": 5494
+    },
+    {
+      "epoch": 1.8350308899649357,
+      "grad_norm": 0.41349629542546434,
+      "learning_rate": 3.930744034015477e-06,
+      "loss": 0.2028,
+      "step": 5495
+    },
+    {
+      "epoch": 1.8353648355318084,
+      "grad_norm": 0.3874670370011834,
+      "learning_rate": 3.9288457380903954e-06,
+      "loss": 0.1946,
+      "step": 5496
+    },
+    {
+      "epoch": 1.835698781098681,
+      "grad_norm": 0.35173551946256226,
+      "learning_rate": 3.926947603975778e-06,
+      "loss": 0.1787,
+      "step": 5497
+    },
+    {
+      "epoch": 1.8360327266655534,
+      "grad_norm": 0.38153385633184733,
+      "learning_rate": 3.925049631958361e-06,
+      "loss": 0.1901,
+      "step": 5498
+    },
+    {
+      "epoch": 1.8363666722324261,
+      "grad_norm": 0.37437515777658054,
+      "learning_rate": 3.923151822324854e-06,
+      "loss": 0.1772,
+      "step": 5499
+    },
+    {
+      "epoch": 1.8367006177992988,
+      "grad_norm": 0.3666580407206624,
+      "learning_rate": 3.9212541753619435e-06,
+      "loss": 0.1753,
+      "step": 5500
+    },
+    {
+      "epoch": 1.8370345633661713,
+      "grad_norm": 0.41030532655032376,
+      "learning_rate": 3.9193566913562915e-06,
+      "loss": 0.2008,
+      "step": 5501
+    },
+    {
+      "epoch": 1.8373685089330438,
+      "grad_norm": 0.47705147398700304,
+      "learning_rate": 3.917459370594537e-06,
+      "loss": 0.2097,
+      "step": 5502
+    },
+    {
+      "epoch": 1.8377024544999165,
+      "grad_norm": 0.3711053442410675,
+      "learning_rate": 3.915562213363287e-06,
+      "loss": 0.1921,
+      "step": 5503
+    },
+    {
+      "epoch": 1.8380364000667893,
+      "grad_norm": 0.3925948723529021,
+      "learning_rate": 3.9136652199491365e-06,
+      "loss": 0.1895,
+      "step": 5504
+    },
+    {
+      "epoch": 1.8383703456336618,
+      "grad_norm": 0.44496015087618784,
+      "learning_rate": 3.911768390638645e-06,
+      "loss": 0.2036,
+      "step": 5505
+    },
+    {
+      "epoch": 1.8387042912005342,
+      "grad_norm": 0.4320342801300668,
+      "learning_rate": 3.909871725718353e-06,
+      "loss": 0.2007,
+      "step": 5506
+    },
+    {
+      "epoch": 1.839038236767407,
+      "grad_norm": 0.40009969508303217,
+      "learning_rate": 3.907975225474771e-06,
+      "loss": 0.1907,
+      "step": 5507
+    },
+    {
+      "epoch": 1.8393721823342795,
+      "grad_norm": 0.44359487303286543,
+      "learning_rate": 3.906078890194391e-06,
+      "loss": 0.2049,
+      "step": 5508
+    },
+    {
+      "epoch": 1.839706127901152,
+      "grad_norm": 0.38608393971174715,
+      "learning_rate": 3.904182720163672e-06,
+      "loss": 0.1898,
+      "step": 5509
+    },
+    {
+      "epoch": 1.8400400734680247,
+      "grad_norm": 0.3923128843531059,
+      "learning_rate": 3.902286715669058e-06,
+      "loss": 0.1837,
+      "step": 5510
+    },
+    {
+      "epoch": 1.8403740190348974,
+      "grad_norm": 0.4372944654236281,
+      "learning_rate": 3.9003908769969615e-06,
+      "loss": 0.1986,
+      "step": 5511
+    },
+    {
+      "epoch": 1.8407079646017699,
+      "grad_norm": 0.41187405317933395,
+      "learning_rate": 3.89849520443377e-06,
+      "loss": 0.1895,
+      "step": 5512
+    },
+    {
+      "epoch": 1.8410419101686424,
+      "grad_norm": 0.4036022582486776,
+      "learning_rate": 3.896599698265847e-06,
+      "loss": 0.1859,
+      "step": 5513
+    },
+    {
+      "epoch": 1.841375855735515,
+      "grad_norm": 0.39156933325917426,
+      "learning_rate": 3.894704358779533e-06,
+      "loss": 0.1933,
+      "step": 5514
+    },
+    {
+      "epoch": 1.8417098013023878,
+      "grad_norm": 0.41843329760041964,
+      "learning_rate": 3.892809186261138e-06,
+      "loss": 0.2076,
+      "step": 5515
+    },
+    {
+      "epoch": 1.8420437468692603,
+      "grad_norm": 0.456977264972126,
+      "learning_rate": 3.890914180996954e-06,
+      "loss": 0.2005,
+      "step": 5516
+    },
+    {
+      "epoch": 1.8423776924361328,
+      "grad_norm": 0.38080749341104936,
+      "learning_rate": 3.889019343273242e-06,
+      "loss": 0.1896,
+      "step": 5517
+    },
+    {
+      "epoch": 1.8427116380030055,
+      "grad_norm": 0.433536114286531,
+      "learning_rate": 3.887124673376239e-06,
+      "loss": 0.1972,
+      "step": 5518
+    },
+    {
+      "epoch": 1.8430455835698782,
+      "grad_norm": 0.4274587785341165,
+      "learning_rate": 3.885230171592157e-06,
+      "loss": 0.2031,
+      "step": 5519
+    },
+    {
+      "epoch": 1.8433795291367507,
+      "grad_norm": 0.4017626320601205,
+      "learning_rate": 3.883335838207183e-06,
+      "loss": 0.1903,
+      "step": 5520
+    },
+    {
+      "epoch": 1.8437134747036232,
+      "grad_norm": 0.40513921939830017,
+      "learning_rate": 3.881441673507481e-06,
+      "loss": 0.1864,
+      "step": 5521
+    },
+    {
+      "epoch": 1.844047420270496,
+      "grad_norm": 0.4158883732225394,
+      "learning_rate": 3.879547677779184e-06,
+      "loss": 0.186,
+      "step": 5522
+    },
+    {
+      "epoch": 1.8443813658373687,
+      "grad_norm": 0.39393567740605284,
+      "learning_rate": 3.8776538513084036e-06,
+      "loss": 0.1913,
+      "step": 5523
+    },
+    {
+      "epoch": 1.844715311404241,
+      "grad_norm": 0.3830752975691852,
+      "learning_rate": 3.875760194381224e-06,
+      "loss": 0.181,
+      "step": 5524
+    },
+    {
+      "epoch": 1.8450492569711137,
+      "grad_norm": 0.43236140310887133,
+      "learning_rate": 3.873866707283704e-06,
+      "loss": 0.195,
+      "step": 5525
+    },
+    {
+      "epoch": 1.8453832025379864,
+      "grad_norm": 0.3971639731949253,
+      "learning_rate": 3.871973390301876e-06,
+      "loss": 0.1908,
+      "step": 5526
+    },
+    {
+      "epoch": 1.8457171481048589,
+      "grad_norm": 0.4329975492794574,
+      "learning_rate": 3.8700802437217526e-06,
+      "loss": 0.1973,
+      "step": 5527
+    },
+    {
+      "epoch": 1.8460510936717314,
+      "grad_norm": 0.39079323807263294,
+      "learning_rate": 3.8681872678293115e-06,
+      "loss": 0.1805,
+      "step": 5528
+    },
+    {
+      "epoch": 1.846385039238604,
+      "grad_norm": 0.4575892417350222,
+      "learning_rate": 3.866294462910511e-06,
+      "loss": 0.1846,
+      "step": 5529
+    },
+    {
+      "epoch": 1.8467189848054768,
+      "grad_norm": 0.39421025346303856,
+      "learning_rate": 3.86440182925128e-06,
+      "loss": 0.1857,
+      "step": 5530
+    },
+    {
+      "epoch": 1.8470529303723493,
+      "grad_norm": 0.3889312076173911,
+      "learning_rate": 3.862509367137525e-06,
+      "loss": 0.1846,
+      "step": 5531
+    },
+    {
+      "epoch": 1.8473868759392218,
+      "grad_norm": 0.4334080310493205,
+      "learning_rate": 3.86061707685512e-06,
+      "loss": 0.2061,
+      "step": 5532
+    },
+    {
+      "epoch": 1.8477208215060945,
+      "grad_norm": 0.38246167969063505,
+      "learning_rate": 3.8587249586899245e-06,
+      "loss": 0.1822,
+      "step": 5533
+    },
+    {
+      "epoch": 1.8480547670729672,
+      "grad_norm": 0.38452054872558217,
+      "learning_rate": 3.856833012927762e-06,
+      "loss": 0.1843,
+      "step": 5534
+    },
+    {
+      "epoch": 1.8483887126398397,
+      "grad_norm": 0.36692188968122724,
+      "learning_rate": 3.854941239854433e-06,
+      "loss": 0.1859,
+      "step": 5535
+    },
+    {
+      "epoch": 1.8487226582067122,
+      "grad_norm": 0.4291853445087326,
+      "learning_rate": 3.853049639755713e-06,
+      "loss": 0.1927,
+      "step": 5536
+    },
+    {
+      "epoch": 1.849056603773585,
+      "grad_norm": 0.4024469130697564,
+      "learning_rate": 3.8511582129173495e-06,
+      "loss": 0.1853,
+      "step": 5537
+    },
+    {
+      "epoch": 1.8493905493404577,
+      "grad_norm": 0.4198881770666027,
+      "learning_rate": 3.8492669596250636e-06,
+      "loss": 0.1983,
+      "step": 5538
+    },
+    {
+      "epoch": 1.8497244949073302,
+      "grad_norm": 0.4290931844458051,
+      "learning_rate": 3.8473758801645535e-06,
+      "loss": 0.2179,
+      "step": 5539
+    },
+    {
+      "epoch": 1.8500584404742026,
+      "grad_norm": 0.4519053873864919,
+      "learning_rate": 3.84548497482149e-06,
+      "loss": 0.2008,
+      "step": 5540
+    },
+    {
+      "epoch": 1.8503923860410754,
+      "grad_norm": 0.39187413508386726,
+      "learning_rate": 3.843594243881513e-06,
+      "loss": 0.1898,
+      "step": 5541
+    },
+    {
+      "epoch": 1.8507263316079479,
+      "grad_norm": 0.4042759859011185,
+      "learning_rate": 3.841703687630243e-06,
+      "loss": 0.1902,
+      "step": 5542
+    },
+    {
+      "epoch": 1.8510602771748204,
+      "grad_norm": 0.42476477246752786,
+      "learning_rate": 3.8398133063532685e-06,
+      "loss": 0.2123,
+      "step": 5543
+    },
+    {
+      "epoch": 1.851394222741693,
+      "grad_norm": 0.4270712881303755,
+      "learning_rate": 3.837923100336155e-06,
+      "loss": 0.2102,
+      "step": 5544
+    },
+    {
+      "epoch": 1.8517281683085658,
+      "grad_norm": 0.40813919718254027,
+      "learning_rate": 3.836033069864441e-06,
+      "loss": 0.1959,
+      "step": 5545
+    },
+    {
+      "epoch": 1.8520621138754383,
+      "grad_norm": 0.5303707806361242,
+      "learning_rate": 3.834143215223637e-06,
+      "loss": 0.2061,
+      "step": 5546
+    },
+    {
+      "epoch": 1.8523960594423108,
+      "grad_norm": 0.5552333428331633,
+      "learning_rate": 3.832253536699227e-06,
+      "loss": 0.2078,
+      "step": 5547
+    },
+    {
+      "epoch": 1.8527300050091835,
+      "grad_norm": 0.38601722650526543,
+      "learning_rate": 3.8303640345766714e-06,
+      "loss": 0.1881,
+      "step": 5548
+    },
+    {
+      "epoch": 1.8530639505760562,
+      "grad_norm": 0.3968436369022213,
+      "learning_rate": 3.8284747091414e-06,
+      "loss": 0.1951,
+      "step": 5549
+    },
+    {
+      "epoch": 1.8533978961429287,
+      "grad_norm": 0.3590138296067034,
+      "learning_rate": 3.826585560678816e-06,
+      "loss": 0.1755,
+      "step": 5550
+    },
+    {
+      "epoch": 1.8537318417098012,
+      "grad_norm": 0.4474793230992464,
+      "learning_rate": 3.824696589474301e-06,
+      "loss": 0.1995,
+      "step": 5551
+    },
+    {
+      "epoch": 1.854065787276674,
+      "grad_norm": 0.4182641931292139,
+      "learning_rate": 3.8228077958132055e-06,
+      "loss": 0.2056,
+      "step": 5552
+    },
+    {
+      "epoch": 1.8543997328435466,
+      "grad_norm": 0.41418885944491485,
+      "learning_rate": 3.8209191799808535e-06,
+      "loss": 0.175,
+      "step": 5553
+    },
+    {
+      "epoch": 1.8547336784104191,
+      "grad_norm": 0.4098383640034617,
+      "learning_rate": 3.819030742262542e-06,
+      "loss": 0.2002,
+      "step": 5554
+    },
+    {
+      "epoch": 1.8550676239772916,
+      "grad_norm": 0.43699938557858403,
+      "learning_rate": 3.817142482943543e-06,
+      "loss": 0.1906,
+      "step": 5555
+    },
+    {
+      "epoch": 1.8554015695441644,
+      "grad_norm": 0.41360176753527556,
+      "learning_rate": 3.815254402309097e-06,
+      "loss": 0.1967,
+      "step": 5556
+    },
+    {
+      "epoch": 1.8557355151110368,
+      "grad_norm": 0.42771060635906877,
+      "learning_rate": 3.813366500644426e-06,
+      "loss": 0.2154,
+      "step": 5557
+    },
+    {
+      "epoch": 1.8560694606779093,
+      "grad_norm": 0.36870116160817834,
+      "learning_rate": 3.8114787782347172e-06,
+      "loss": 0.1812,
+      "step": 5558
+    },
+    {
+      "epoch": 1.856403406244782,
+      "grad_norm": 0.40901216532316104,
+      "learning_rate": 3.809591235365133e-06,
+      "loss": 0.1928,
+      "step": 5559
+    },
+    {
+      "epoch": 1.8567373518116548,
+      "grad_norm": 0.37696959669230656,
+      "learning_rate": 3.807703872320809e-06,
+      "loss": 0.1763,
+      "step": 5560
+    },
+    {
+      "epoch": 1.8570712973785273,
+      "grad_norm": 0.3596874200077695,
+      "learning_rate": 3.8058166893868543e-06,
+      "loss": 0.1714,
+      "step": 5561
+    },
+    {
+      "epoch": 1.8574052429453998,
+      "grad_norm": 0.35321557980968193,
+      "learning_rate": 3.8039296868483493e-06,
+      "loss": 0.1739,
+      "step": 5562
+    },
+    {
+      "epoch": 1.8577391885122725,
+      "grad_norm": 0.38771500147851023,
+      "learning_rate": 3.802042864990349e-06,
+      "loss": 0.1897,
+      "step": 5563
+    },
+    {
+      "epoch": 1.8580731340791452,
+      "grad_norm": 0.4151413292305868,
+      "learning_rate": 3.8001562240978785e-06,
+      "loss": 0.1842,
+      "step": 5564
+    },
+    {
+      "epoch": 1.8584070796460177,
+      "grad_norm": 0.40997672571652566,
+      "learning_rate": 3.7982697644559385e-06,
+      "loss": 0.1959,
+      "step": 5565
+    },
+    {
+      "epoch": 1.8587410252128902,
+      "grad_norm": 0.44628840605004716,
+      "learning_rate": 3.7963834863495013e-06,
+      "loss": 0.2154,
+      "step": 5566
+    },
+    {
+      "epoch": 1.859074970779763,
+      "grad_norm": 0.3581451638522709,
+      "learning_rate": 3.794497390063509e-06,
+      "loss": 0.1749,
+      "step": 5567
+    },
+    {
+      "epoch": 1.8594089163466356,
+      "grad_norm": 0.4141106417920125,
+      "learning_rate": 3.792611475882881e-06,
+      "loss": 0.1854,
+      "step": 5568
+    },
+    {
+      "epoch": 1.8597428619135081,
+      "grad_norm": 0.3939536353320779,
+      "learning_rate": 3.790725744092507e-06,
+      "loss": 0.1864,
+      "step": 5569
+    },
+    {
+      "epoch": 1.8600768074803806,
+      "grad_norm": 0.4176080845444923,
+      "learning_rate": 3.788840194977248e-06,
+      "loss": 0.19,
+      "step": 5570
+    },
+    {
+      "epoch": 1.8604107530472533,
+      "grad_norm": 0.3773316234488866,
+      "learning_rate": 3.7869548288219383e-06,
+      "loss": 0.1827,
+      "step": 5571
+    },
+    {
+      "epoch": 1.860744698614126,
+      "grad_norm": 0.45447820265924405,
+      "learning_rate": 3.7850696459113845e-06,
+      "loss": 0.2136,
+      "step": 5572
+    },
+    {
+      "epoch": 1.8610786441809983,
+      "grad_norm": 0.40973336466501775,
+      "learning_rate": 3.783184646530364e-06,
+      "loss": 0.1974,
+      "step": 5573
+    },
+    {
+      "epoch": 1.861412589747871,
+      "grad_norm": 0.38507853427875804,
+      "learning_rate": 3.7812998309636323e-06,
+      "loss": 0.1843,
+      "step": 5574
+    },
+    {
+      "epoch": 1.8617465353147438,
+      "grad_norm": 0.42639231490055507,
+      "learning_rate": 3.779415199495911e-06,
+      "loss": 0.19,
+      "step": 5575
+    },
+    {
+      "epoch": 1.8620804808816163,
+      "grad_norm": 0.39541454547140653,
+      "learning_rate": 3.777530752411896e-06,
+      "loss": 0.1969,
+      "step": 5576
+    },
+    {
+      "epoch": 1.8624144264484888,
+      "grad_norm": 0.41359196848838,
+      "learning_rate": 3.7756464899962546e-06,
+      "loss": 0.1888,
+      "step": 5577
+    },
+    {
+      "epoch": 1.8627483720153615,
+      "grad_norm": 0.36075624301082615,
+      "learning_rate": 3.773762412533627e-06,
+      "loss": 0.1792,
+      "step": 5578
+    },
+    {
+      "epoch": 1.8630823175822342,
+      "grad_norm": 0.4098671567571041,
+      "learning_rate": 3.771878520308624e-06,
+      "loss": 0.1912,
+      "step": 5579
+    },
+    {
+      "epoch": 1.8634162631491067,
+      "grad_norm": 0.3970704176477835,
+      "learning_rate": 3.7699948136058327e-06,
+      "loss": 0.2007,
+      "step": 5580
+    },
+    {
+      "epoch": 1.8637502087159792,
+      "grad_norm": 0.41258075692882656,
+      "learning_rate": 3.768111292709808e-06,
+      "loss": 0.1952,
+      "step": 5581
+    },
+    {
+      "epoch": 1.864084154282852,
+      "grad_norm": 0.4985649821234798,
+      "learning_rate": 3.7662279579050777e-06,
+      "loss": 0.1976,
+      "step": 5582
+    },
+    {
+      "epoch": 1.8644180998497246,
+      "grad_norm": 0.39543825043818337,
+      "learning_rate": 3.764344809476141e-06,
+      "loss": 0.1839,
+      "step": 5583
+    },
+    {
+      "epoch": 1.8647520454165971,
+      "grad_norm": 0.4169230905277188,
+      "learning_rate": 3.7624618477074705e-06,
+      "loss": 0.1792,
+      "step": 5584
+    },
+    {
+      "epoch": 1.8650859909834696,
+      "grad_norm": 0.40159431111505056,
+      "learning_rate": 3.760579072883508e-06,
+      "loss": 0.1851,
+      "step": 5585
+    },
+    {
+      "epoch": 1.8654199365503423,
+      "grad_norm": 0.42377227839386244,
+      "learning_rate": 3.758696485288672e-06,
+      "loss": 0.201,
+      "step": 5586
+    },
+    {
+      "epoch": 1.865753882117215,
+      "grad_norm": 0.4257414605932439,
+      "learning_rate": 3.7568140852073464e-06,
+      "loss": 0.1975,
+      "step": 5587
+    },
+    {
+      "epoch": 1.8660878276840875,
+      "grad_norm": 0.4169457936193867,
+      "learning_rate": 3.754931872923892e-06,
+      "loss": 0.1923,
+      "step": 5588
+    },
+    {
+      "epoch": 1.86642177325096,
+      "grad_norm": 0.4043813591817317,
+      "learning_rate": 3.7530498487226384e-06,
+      "loss": 0.1929,
+      "step": 5589
+    },
+    {
+      "epoch": 1.8667557188178328,
+      "grad_norm": 0.41635875636597164,
+      "learning_rate": 3.751168012887888e-06,
+      "loss": 0.1935,
+      "step": 5590
+    },
+    {
+      "epoch": 1.8670896643847052,
+      "grad_norm": 0.38859957319609856,
+      "learning_rate": 3.7492863657039126e-06,
+      "loss": 0.1879,
+      "step": 5591
+    },
+    {
+      "epoch": 1.8674236099515777,
+      "grad_norm": 0.4261527050568269,
+      "learning_rate": 3.7474049074549596e-06,
+      "loss": 0.1999,
+      "step": 5592
+    },
+    {
+      "epoch": 1.8677575555184505,
+      "grad_norm": 0.42976803187002927,
+      "learning_rate": 3.7455236384252435e-06,
+      "loss": 0.2078,
+      "step": 5593
+    },
+    {
+      "epoch": 1.8680915010853232,
+      "grad_norm": 0.3885539430483641,
+      "learning_rate": 3.743642558898953e-06,
+      "loss": 0.1803,
+      "step": 5594
+    },
+    {
+      "epoch": 1.8684254466521957,
+      "grad_norm": 0.475836117741637,
+      "learning_rate": 3.7417616691602477e-06,
+      "loss": 0.1962,
+      "step": 5595
+    },
+    {
+      "epoch": 1.8687593922190682,
+      "grad_norm": 0.4264306061995,
+      "learning_rate": 3.739880969493257e-06,
+      "loss": 0.2007,
+      "step": 5596
+    },
+    {
+      "epoch": 1.869093337785941,
+      "grad_norm": 0.4188722723961078,
+      "learning_rate": 3.738000460182081e-06,
+      "loss": 0.199,
+      "step": 5597
+    },
+    {
+      "epoch": 1.8694272833528136,
+      "grad_norm": 0.43449811676434813,
+      "learning_rate": 3.736120141510798e-06,
+      "loss": 0.2002,
+      "step": 5598
+    },
+    {
+      "epoch": 1.869761228919686,
+      "grad_norm": 0.4255881141855972,
+      "learning_rate": 3.734240013763448e-06,
+      "loss": 0.2031,
+      "step": 5599
+    },
+    {
+      "epoch": 1.8700951744865586,
+      "grad_norm": 0.3924986018956876,
+      "learning_rate": 3.732360077224049e-06,
+      "loss": 0.1834,
+      "step": 5600
+    },
+    {
+      "epoch": 1.8704291200534313,
+      "grad_norm": 0.41303090285019545,
+      "learning_rate": 3.730480332176586e-06,
+      "loss": 0.1947,
+      "step": 5601
+    },
+    {
+      "epoch": 1.870763065620304,
+      "grad_norm": 0.4173686398679209,
+      "learning_rate": 3.7286007789050147e-06,
+      "loss": 0.2048,
+      "step": 5602
+    },
+    {
+      "epoch": 1.8710970111871765,
+      "grad_norm": 0.42256726256761623,
+      "learning_rate": 3.726721417693268e-06,
+      "loss": 0.196,
+      "step": 5603
+    },
+    {
+      "epoch": 1.871430956754049,
+      "grad_norm": 0.37707316686734915,
+      "learning_rate": 3.7248422488252433e-06,
+      "loss": 0.184,
+      "step": 5604
+    },
+    {
+      "epoch": 1.8717649023209217,
+      "grad_norm": 0.42889234026369716,
+      "learning_rate": 3.722963272584812e-06,
+      "loss": 0.1908,
+      "step": 5605
+    },
+    {
+      "epoch": 1.8720988478877942,
+      "grad_norm": 0.3962460242481815,
+      "learning_rate": 3.721084489255815e-06,
+      "loss": 0.1839,
+      "step": 5606
+    },
+    {
+      "epoch": 1.8724327934546667,
+      "grad_norm": 0.4230266260559559,
+      "learning_rate": 3.719205899122064e-06,
+      "loss": 0.1995,
+      "step": 5607
+    },
+    {
+      "epoch": 1.8727667390215395,
+      "grad_norm": 0.42644054894229216,
+      "learning_rate": 3.7173275024673424e-06,
+      "loss": 0.197,
+      "step": 5608
+    },
+    {
+      "epoch": 1.8731006845884122,
+      "grad_norm": 0.4754937302170339,
+      "learning_rate": 3.7154492995754046e-06,
+      "loss": 0.2116,
+      "step": 5609
+    },
+    {
+      "epoch": 1.8734346301552847,
+      "grad_norm": 0.4370210334991797,
+      "learning_rate": 3.7135712907299753e-06,
+      "loss": 0.1991,
+      "step": 5610
+    },
+    {
+      "epoch": 1.8737685757221572,
+      "grad_norm": 0.3959125955396685,
+      "learning_rate": 3.7116934762147504e-06,
+      "loss": 0.1768,
+      "step": 5611
+    },
+    {
+      "epoch": 1.8741025212890299,
+      "grad_norm": 0.40337586546501275,
+      "learning_rate": 3.709815856313395e-06,
+      "loss": 0.1843,
+      "step": 5612
+    },
+    {
+      "epoch": 1.8744364668559026,
+      "grad_norm": 0.4036266931746901,
+      "learning_rate": 3.7079384313095464e-06,
+      "loss": 0.1948,
+      "step": 5613
+    },
+    {
+      "epoch": 1.874770412422775,
+      "grad_norm": 0.40647248544698983,
+      "learning_rate": 3.70606120148681e-06,
+      "loss": 0.1949,
+      "step": 5614
+    },
+    {
+      "epoch": 1.8751043579896476,
+      "grad_norm": 0.4983920786650823,
+      "learning_rate": 3.7041841671287654e-06,
+      "loss": 0.1782,
+      "step": 5615
+    },
+    {
+      "epoch": 1.8754383035565203,
+      "grad_norm": 0.42161416739832963,
+      "learning_rate": 3.70230732851896e-06,
+      "loss": 0.1899,
+      "step": 5616
+    },
+    {
+      "epoch": 1.875772249123393,
+      "grad_norm": 0.5092564041054836,
+      "learning_rate": 3.7004306859409134e-06,
+      "loss": 0.2121,
+      "step": 5617
+    },
+    {
+      "epoch": 1.8761061946902655,
+      "grad_norm": 0.42450305824700535,
+      "learning_rate": 3.6985542396781127e-06,
+      "loss": 0.1945,
+      "step": 5618
+    },
+    {
+      "epoch": 1.876440140257138,
+      "grad_norm": 0.41218596453526796,
+      "learning_rate": 3.6966779900140193e-06,
+      "loss": 0.1883,
+      "step": 5619
+    },
+    {
+      "epoch": 1.8767740858240107,
+      "grad_norm": 0.45585778407093536,
+      "learning_rate": 3.694801937232058e-06,
+      "loss": 0.1923,
+      "step": 5620
+    },
+    {
+      "epoch": 1.8771080313908834,
+      "grad_norm": 0.43555642104196896,
+      "learning_rate": 3.6929260816156353e-06,
+      "loss": 0.2081,
+      "step": 5621
+    },
+    {
+      "epoch": 1.8774419769577557,
+      "grad_norm": 0.38653206137255025,
+      "learning_rate": 3.691050423448118e-06,
+      "loss": 0.188,
+      "step": 5622
+    },
+    {
+      "epoch": 1.8777759225246284,
+      "grad_norm": 0.44802164116477605,
+      "learning_rate": 3.689174963012847e-06,
+      "loss": 0.204,
+      "step": 5623
+    },
+    {
+      "epoch": 1.8781098680915012,
+      "grad_norm": 0.3824309407516524,
+      "learning_rate": 3.6872997005931323e-06,
+      "loss": 0.1876,
+      "step": 5624
+    },
+    {
+      "epoch": 1.8784438136583737,
+      "grad_norm": 0.4373752040761639,
+      "learning_rate": 3.6854246364722534e-06,
+      "loss": 0.1988,
+      "step": 5625
+    },
+    {
+      "epoch": 1.8787777592252461,
+      "grad_norm": 0.3912895000237093,
+      "learning_rate": 3.683549770933461e-06,
+      "loss": 0.1888,
+      "step": 5626
+    },
+    {
+      "epoch": 1.8791117047921189,
+      "grad_norm": 0.46238272413161635,
+      "learning_rate": 3.6816751042599774e-06,
+      "loss": 0.2081,
+      "step": 5627
+    },
+    {
+      "epoch": 1.8794456503589916,
+      "grad_norm": 0.38589801494425297,
+      "learning_rate": 3.6798006367349926e-06,
+      "loss": 0.1829,
+      "step": 5628
+    },
+    {
+      "epoch": 1.879779595925864,
+      "grad_norm": 0.3752621047827673,
+      "learning_rate": 3.6779263686416668e-06,
+      "loss": 0.1925,
+      "step": 5629
+    },
+    {
+      "epoch": 1.8801135414927366,
+      "grad_norm": 0.3765889397099863,
+      "learning_rate": 3.676052300263129e-06,
+      "loss": 0.1846,
+      "step": 5630
+    },
+    {
+      "epoch": 1.8804474870596093,
+      "grad_norm": 0.3676402690847071,
+      "learning_rate": 3.6741784318824814e-06,
+      "loss": 0.1808,
+      "step": 5631
+    },
+    {
+      "epoch": 1.880781432626482,
+      "grad_norm": 0.3551659592860279,
+      "learning_rate": 3.6723047637827897e-06,
+      "loss": 0.1712,
+      "step": 5632
+    },
+    {
+      "epoch": 1.8811153781933545,
+      "grad_norm": 0.410879058236752,
+      "learning_rate": 3.670431296247099e-06,
+      "loss": 0.2003,
+      "step": 5633
+    },
+    {
+      "epoch": 1.881449323760227,
+      "grad_norm": 0.40283302371260815,
+      "learning_rate": 3.6685580295584162e-06,
+      "loss": 0.188,
+      "step": 5634
+    },
+    {
+      "epoch": 1.8817832693270997,
+      "grad_norm": 0.3891040548112664,
+      "learning_rate": 3.6666849639997205e-06,
+      "loss": 0.1849,
+      "step": 5635
+    },
+    {
+      "epoch": 1.8821172148939724,
+      "grad_norm": 0.41052509920344155,
+      "learning_rate": 3.6648120998539596e-06,
+      "loss": 0.1862,
+      "step": 5636
+    },
+    {
+      "epoch": 1.882451160460845,
+      "grad_norm": 0.4079072214569817,
+      "learning_rate": 3.662939437404053e-06,
+      "loss": 0.1982,
+      "step": 5637
+    },
+    {
+      "epoch": 1.8827851060277174,
+      "grad_norm": 0.42609676874734265,
+      "learning_rate": 3.6610669769328853e-06,
+      "loss": 0.1851,
+      "step": 5638
+    },
+    {
+      "epoch": 1.8831190515945901,
+      "grad_norm": 0.4182804674313846,
+      "learning_rate": 3.659194718723319e-06,
+      "loss": 0.1946,
+      "step": 5639
+    },
+    {
+      "epoch": 1.8834529971614626,
+      "grad_norm": 0.406048729629456,
+      "learning_rate": 3.657322663058177e-06,
+      "loss": 0.1954,
+      "step": 5640
+    },
+    {
+      "epoch": 1.8837869427283351,
+      "grad_norm": 0.4220538626404686,
+      "learning_rate": 3.655450810220257e-06,
+      "loss": 0.1911,
+      "step": 5641
+    },
+    {
+      "epoch": 1.8841208882952079,
+      "grad_norm": 0.45423348313302325,
+      "learning_rate": 3.6535791604923225e-06,
+      "loss": 0.1924,
+      "step": 5642
+    },
+    {
+      "epoch": 1.8844548338620806,
+      "grad_norm": 0.3981725690261841,
+      "learning_rate": 3.6517077141571076e-06,
+      "loss": 0.1844,
+      "step": 5643
+    },
+    {
+      "epoch": 1.884788779428953,
+      "grad_norm": 0.559974848713012,
+      "learning_rate": 3.649836471497321e-06,
+      "loss": 0.1981,
+      "step": 5644
+    },
+    {
+      "epoch": 1.8851227249958256,
+      "grad_norm": 0.42473441487267527,
+      "learning_rate": 3.6479654327956325e-06,
+      "loss": 0.2059,
+      "step": 5645
+    },
+    {
+      "epoch": 1.8854566705626983,
+      "grad_norm": 0.4574118801471747,
+      "learning_rate": 3.646094598334685e-06,
+      "loss": 0.1964,
+      "step": 5646
+    },
+    {
+      "epoch": 1.885790616129571,
+      "grad_norm": 0.4137780911436003,
+      "learning_rate": 3.64422396839709e-06,
+      "loss": 0.1848,
+      "step": 5647
+    },
+    {
+      "epoch": 1.8861245616964435,
+      "grad_norm": 0.40265674895445425,
+      "learning_rate": 3.642353543265429e-06,
+      "loss": 0.1887,
+      "step": 5648
+    },
+    {
+      "epoch": 1.886458507263316,
+      "grad_norm": 0.3859651283367934,
+      "learning_rate": 3.640483323222248e-06,
+      "loss": 0.1923,
+      "step": 5649
+    },
+    {
+      "epoch": 1.8867924528301887,
+      "grad_norm": 0.4131845833522721,
+      "learning_rate": 3.638613308550072e-06,
+      "loss": 0.2009,
+      "step": 5650
+    },
+    {
+      "epoch": 1.8871263983970614,
+      "grad_norm": 0.3875014614187485,
+      "learning_rate": 3.636743499531385e-06,
+      "loss": 0.1782,
+      "step": 5651
+    },
+    {
+      "epoch": 1.887460343963934,
+      "grad_norm": 0.38824358254338215,
+      "learning_rate": 3.634873896448644e-06,
+      "loss": 0.1885,
+      "step": 5652
+    },
+    {
+      "epoch": 1.8877942895308064,
+      "grad_norm": 0.35317513678915435,
+      "learning_rate": 3.633004499584275e-06,
+      "loss": 0.1774,
+      "step": 5653
+    },
+    {
+      "epoch": 1.8881282350976791,
+      "grad_norm": 0.3775118670972697,
+      "learning_rate": 3.6311353092206723e-06,
+      "loss": 0.186,
+      "step": 5654
+    },
+    {
+      "epoch": 1.8884621806645516,
+      "grad_norm": 0.4040450531349787,
+      "learning_rate": 3.6292663256401967e-06,
+      "loss": 0.1935,
+      "step": 5655
+    },
+    {
+      "epoch": 1.8887961262314241,
+      "grad_norm": 0.4223300761091552,
+      "learning_rate": 3.6273975491251844e-06,
+      "loss": 0.2026,
+      "step": 5656
+    },
+    {
+      "epoch": 1.8891300717982968,
+      "grad_norm": 0.3819512409614182,
+      "learning_rate": 3.625528979957935e-06,
+      "loss": 0.2008,
+      "step": 5657
+    },
+    {
+      "epoch": 1.8894640173651696,
+      "grad_norm": 0.40376004385819364,
+      "learning_rate": 3.6236606184207164e-06,
+      "loss": 0.184,
+      "step": 5658
+    },
+    {
+      "epoch": 1.889797962932042,
+      "grad_norm": 0.4018629191997057,
+      "learning_rate": 3.621792464795767e-06,
+      "loss": 0.1921,
+      "step": 5659
+    },
+    {
+      "epoch": 1.8901319084989145,
+      "grad_norm": 0.38134294753922704,
+      "learning_rate": 3.6199245193652944e-06,
+      "loss": 0.1823,
+      "step": 5660
+    },
+    {
+      "epoch": 1.8904658540657873,
+      "grad_norm": 0.4086245882829037,
+      "learning_rate": 3.6180567824114715e-06,
+      "loss": 0.2045,
+      "step": 5661
+    },
+    {
+      "epoch": 1.89079979963266,
+      "grad_norm": 0.4147921609611881,
+      "learning_rate": 3.6161892542164444e-06,
+      "loss": 0.2035,
+      "step": 5662
+    },
+    {
+      "epoch": 1.8911337451995325,
+      "grad_norm": 0.403633243120622,
+      "learning_rate": 3.614321935062325e-06,
+      "loss": 0.1905,
+      "step": 5663
+    },
+    {
+      "epoch": 1.891467690766405,
+      "grad_norm": 0.4062716446644573,
+      "learning_rate": 3.6124548252311918e-06,
+      "loss": 0.2017,
+      "step": 5664
+    },
+    {
+      "epoch": 1.8918016363332777,
+      "grad_norm": 0.43732722649221445,
+      "learning_rate": 3.610587925005097e-06,
+      "loss": 0.1922,
+      "step": 5665
+    },
+    {
+      "epoch": 1.8921355819001504,
+      "grad_norm": 0.4168207296619112,
+      "learning_rate": 3.608721234666054e-06,
+      "loss": 0.1872,
+      "step": 5666
+    },
+    {
+      "epoch": 1.892469527467023,
+      "grad_norm": 0.4001633405949691,
+      "learning_rate": 3.6068547544960493e-06,
+      "loss": 0.1999,
+      "step": 5667
+    },
+    {
+      "epoch": 1.8928034730338954,
+      "grad_norm": 0.359620384256815,
+      "learning_rate": 3.6049884847770396e-06,
+      "loss": 0.1879,
+      "step": 5668
+    },
+    {
+      "epoch": 1.8931374186007681,
+      "grad_norm": 0.3814699560183258,
+      "learning_rate": 3.6031224257909448e-06,
+      "loss": 0.1925,
+      "step": 5669
+    },
+    {
+      "epoch": 1.8934713641676408,
+      "grad_norm": 0.4249305582407571,
+      "learning_rate": 3.6012565778196552e-06,
+      "loss": 0.2018,
+      "step": 5670
+    },
+    {
+      "epoch": 1.893805309734513,
+      "grad_norm": 0.44929761623627773,
+      "learning_rate": 3.5993909411450297e-06,
+      "loss": 0.2176,
+      "step": 5671
+    },
+    {
+      "epoch": 1.8941392553013858,
+      "grad_norm": 0.4402284991993238,
+      "learning_rate": 3.597525516048894e-06,
+      "loss": 0.1985,
+      "step": 5672
+    },
+    {
+      "epoch": 1.8944732008682585,
+      "grad_norm": 0.41033454175195433,
+      "learning_rate": 3.5956603028130397e-06,
+      "loss": 0.179,
+      "step": 5673
+    },
+    {
+      "epoch": 1.894807146435131,
+      "grad_norm": 0.41702509272834387,
+      "learning_rate": 3.5937953017192356e-06,
+      "loss": 0.1944,
+      "step": 5674
+    },
+    {
+      "epoch": 1.8951410920020035,
+      "grad_norm": 0.417384380584999,
+      "learning_rate": 3.591930513049208e-06,
+      "loss": 0.2002,
+      "step": 5675
+    },
+    {
+      "epoch": 1.8954750375688763,
+      "grad_norm": 0.404210881524086,
+      "learning_rate": 3.5900659370846556e-06,
+      "loss": 0.1915,
+      "step": 5676
+    },
+    {
+      "epoch": 1.895808983135749,
+      "grad_norm": 0.40054054358372826,
+      "learning_rate": 3.5882015741072464e-06,
+      "loss": 0.1893,
+      "step": 5677
+    },
+    {
+      "epoch": 1.8961429287026215,
+      "grad_norm": 0.4716609545548075,
+      "learning_rate": 3.586337424398609e-06,
+      "loss": 0.2013,
+      "step": 5678
+    },
+    {
+      "epoch": 1.896476874269494,
+      "grad_norm": 0.4086034807646897,
+      "learning_rate": 3.584473488240352e-06,
+      "loss": 0.1891,
+      "step": 5679
+    },
+    {
+      "epoch": 1.8968108198363667,
+      "grad_norm": 0.3875982015231373,
+      "learning_rate": 3.5826097659140413e-06,
+      "loss": 0.189,
+      "step": 5680
+    },
+    {
+      "epoch": 1.8971447654032394,
+      "grad_norm": 0.35728633673821014,
+      "learning_rate": 3.5807462577012152e-06,
+      "loss": 0.181,
+      "step": 5681
+    },
+    {
+      "epoch": 1.897478710970112,
+      "grad_norm": 0.3834926713626789,
+      "learning_rate": 3.5788829638833777e-06,
+      "loss": 0.189,
+      "step": 5682
+    },
+    {
+      "epoch": 1.8978126565369844,
+      "grad_norm": 0.39649251448728734,
+      "learning_rate": 3.5770198847420016e-06,
+      "loss": 0.1892,
+      "step": 5683
+    },
+    {
+      "epoch": 1.898146602103857,
+      "grad_norm": 0.42377795991739803,
+      "learning_rate": 3.5751570205585264e-06,
+      "loss": 0.1914,
+      "step": 5684
+    },
+    {
+      "epoch": 1.8984805476707298,
+      "grad_norm": 0.4002763802703904,
+      "learning_rate": 3.573294371614361e-06,
+      "loss": 0.1905,
+      "step": 5685
+    },
+    {
+      "epoch": 1.8988144932376023,
+      "grad_norm": 0.4236073409479343,
+      "learning_rate": 3.571431938190879e-06,
+      "loss": 0.1981,
+      "step": 5686
+    },
+    {
+      "epoch": 1.8991484388044748,
+      "grad_norm": 0.3916550676455339,
+      "learning_rate": 3.5695697205694246e-06,
+      "loss": 0.1984,
+      "step": 5687
+    },
+    {
+      "epoch": 1.8994823843713475,
+      "grad_norm": 0.4034111232421996,
+      "learning_rate": 3.567707719031306e-06,
+      "loss": 0.195,
+      "step": 5688
+    },
+    {
+      "epoch": 1.89981632993822,
+      "grad_norm": 0.41960190854899787,
+      "learning_rate": 3.5658459338578016e-06,
+      "loss": 0.2041,
+      "step": 5689
+    },
+    {
+      "epoch": 1.9001502755050925,
+      "grad_norm": 0.40615605662054505,
+      "learning_rate": 3.563984365330153e-06,
+      "loss": 0.1915,
+      "step": 5690
+    },
+    {
+      "epoch": 1.9004842210719652,
+      "grad_norm": 0.3654123514969394,
+      "learning_rate": 3.562123013729577e-06,
+      "loss": 0.1834,
+      "step": 5691
+    },
+    {
+      "epoch": 1.900818166638838,
+      "grad_norm": 0.43052112290651406,
+      "learning_rate": 3.56026187933725e-06,
+      "loss": 0.2023,
+      "step": 5692
+    },
+    {
+      "epoch": 1.9011521122057105,
+      "grad_norm": 0.4043205360016579,
+      "learning_rate": 3.5584009624343187e-06,
+      "loss": 0.2009,
+      "step": 5693
+    },
+    {
+      "epoch": 1.901486057772583,
+      "grad_norm": 0.38893024556161726,
+      "learning_rate": 3.5565402633018963e-06,
+      "loss": 0.1861,
+      "step": 5694
+    },
+    {
+      "epoch": 1.9018200033394557,
+      "grad_norm": 0.4320452390717206,
+      "learning_rate": 3.554679782221063e-06,
+      "loss": 0.19,
+      "step": 5695
+    },
+    {
+      "epoch": 1.9021539489063284,
+      "grad_norm": 0.3916662102911691,
+      "learning_rate": 3.552819519472865e-06,
+      "loss": 0.1877,
+      "step": 5696
+    },
+    {
+      "epoch": 1.9024878944732009,
+      "grad_norm": 0.44565474629895935,
+      "learning_rate": 3.5509594753383202e-06,
+      "loss": 0.2008,
+      "step": 5697
+    },
+    {
+      "epoch": 1.9028218400400734,
+      "grad_norm": 0.41234932037291344,
+      "learning_rate": 3.5490996500984085e-06,
+      "loss": 0.1962,
+      "step": 5698
+    },
+    {
+      "epoch": 1.903155785606946,
+      "grad_norm": 0.3981277646668938,
+      "learning_rate": 3.547240044034079e-06,
+      "loss": 0.1839,
+      "step": 5699
+    },
+    {
+      "epoch": 1.9034897311738188,
+      "grad_norm": 0.3711733861198523,
+      "learning_rate": 3.545380657426247e-06,
+      "loss": 0.1824,
+      "step": 5700
+    },
+    {
+      "epoch": 1.9038236767406913,
+      "grad_norm": 0.4233679288176297,
+      "learning_rate": 3.5435214905557937e-06,
+      "loss": 0.1844,
+      "step": 5701
+    },
+    {
+      "epoch": 1.9041576223075638,
+      "grad_norm": 0.3896155487732712,
+      "learning_rate": 3.5416625437035656e-06,
+      "loss": 0.1896,
+      "step": 5702
+    },
+    {
+      "epoch": 1.9044915678744365,
+      "grad_norm": 0.37448542490424214,
+      "learning_rate": 3.539803817150385e-06,
+      "loss": 0.184,
+      "step": 5703
+    },
+    {
+      "epoch": 1.904825513441309,
+      "grad_norm": 0.4181501138090368,
+      "learning_rate": 3.5379453111770313e-06,
+      "loss": 0.2077,
+      "step": 5704
+    },
+    {
+      "epoch": 1.9051594590081815,
+      "grad_norm": 0.43828339837035857,
+      "learning_rate": 3.536087026064252e-06,
+      "loss": 0.2003,
+      "step": 5705
+    },
+    {
+      "epoch": 1.9054934045750542,
+      "grad_norm": 0.3939854984655365,
+      "learning_rate": 3.534228962092766e-06,
+      "loss": 0.1927,
+      "step": 5706
+    },
+    {
+      "epoch": 1.905827350141927,
+      "grad_norm": 0.3850822547742392,
+      "learning_rate": 3.5323711195432533e-06,
+      "loss": 0.1921,
+      "step": 5707
+    },
+    {
+      "epoch": 1.9061612957087994,
+      "grad_norm": 0.3797573178782776,
+      "learning_rate": 3.530513498696363e-06,
+      "loss": 0.1868,
+      "step": 5708
+    },
+    {
+      "epoch": 1.906495241275672,
+      "grad_norm": 0.3974351606525824,
+      "learning_rate": 3.5286560998327125e-06,
+      "loss": 0.1957,
+      "step": 5709
+    },
+    {
+      "epoch": 1.9068291868425447,
+      "grad_norm": 0.37982353045389616,
+      "learning_rate": 3.5267989232328827e-06,
+      "loss": 0.1912,
+      "step": 5710
+    },
+    {
+      "epoch": 1.9071631324094174,
+      "grad_norm": 0.4056309512887979,
+      "learning_rate": 3.5249419691774212e-06,
+      "loss": 0.1878,
+      "step": 5711
+    },
+    {
+      "epoch": 1.9074970779762899,
+      "grad_norm": 0.41189171348380066,
+      "learning_rate": 3.523085237946844e-06,
+      "loss": 0.1928,
+      "step": 5712
+    },
+    {
+      "epoch": 1.9078310235431624,
+      "grad_norm": 0.39755751196618816,
+      "learning_rate": 3.5212287298216306e-06,
+      "loss": 0.1903,
+      "step": 5713
+    },
+    {
+      "epoch": 1.908164969110035,
+      "grad_norm": 0.4965401018841489,
+      "learning_rate": 3.5193724450822296e-06,
+      "loss": 0.2001,
+      "step": 5714
+    },
+    {
+      "epoch": 1.9084989146769078,
+      "grad_norm": 0.41952782189260945,
+      "learning_rate": 3.517516384009056e-06,
+      "loss": 0.1897,
+      "step": 5715
+    },
+    {
+      "epoch": 1.9088328602437803,
+      "grad_norm": 0.36265255842951255,
+      "learning_rate": 3.515660546882488e-06,
+      "loss": 0.1814,
+      "step": 5716
+    },
+    {
+      "epoch": 1.9091668058106528,
+      "grad_norm": 0.41875774466473636,
+      "learning_rate": 3.5138049339828718e-06,
+      "loss": 0.1956,
+      "step": 5717
+    },
+    {
+      "epoch": 1.9095007513775255,
+      "grad_norm": 0.45427593055322385,
+      "learning_rate": 3.5119495455905194e-06,
+      "loss": 0.1929,
+      "step": 5718
+    },
+    {
+      "epoch": 1.9098346969443982,
+      "grad_norm": 0.4283125075634053,
+      "learning_rate": 3.5100943819857082e-06,
+      "loss": 0.1849,
+      "step": 5719
+    },
+    {
+      "epoch": 1.9101686425112705,
+      "grad_norm": 0.3912738538037045,
+      "learning_rate": 3.508239443448685e-06,
+      "loss": 0.1754,
+      "step": 5720
+    },
+    {
+      "epoch": 1.9105025880781432,
+      "grad_norm": 0.3783746014875604,
+      "learning_rate": 3.5063847302596587e-06,
+      "loss": 0.1858,
+      "step": 5721
+    },
+    {
+      "epoch": 1.910836533645016,
+      "grad_norm": 0.40411330753733493,
+      "learning_rate": 3.504530242698806e-06,
+      "loss": 0.185,
+      "step": 5722
+    },
+    {
+      "epoch": 1.9111704792118884,
+      "grad_norm": 0.367902521163063,
+      "learning_rate": 3.5026759810462687e-06,
+      "loss": 0.1823,
+      "step": 5723
+    },
+    {
+      "epoch": 1.911504424778761,
+      "grad_norm": 0.36691214379783854,
+      "learning_rate": 3.5008219455821546e-06,
+      "loss": 0.1837,
+      "step": 5724
+    },
+    {
+      "epoch": 1.9118383703456336,
+      "grad_norm": 0.4113856631417277,
+      "learning_rate": 3.4989681365865363e-06,
+      "loss": 0.1943,
+      "step": 5725
+    },
+    {
+      "epoch": 1.9121723159125064,
+      "grad_norm": 0.38908322636188303,
+      "learning_rate": 3.497114554339457e-06,
+      "loss": 0.175,
+      "step": 5726
+    },
+    {
+      "epoch": 1.9125062614793789,
+      "grad_norm": 0.4302171916585288,
+      "learning_rate": 3.4952611991209197e-06,
+      "loss": 0.1912,
+      "step": 5727
+    },
+    {
+      "epoch": 1.9128402070462514,
+      "grad_norm": 0.42508386899522627,
+      "learning_rate": 3.4934080712108964e-06,
+      "loss": 0.1986,
+      "step": 5728
+    },
+    {
+      "epoch": 1.913174152613124,
+      "grad_norm": 0.40828670712191045,
+      "learning_rate": 3.4915551708893236e-06,
+      "loss": 0.1983,
+      "step": 5729
+    },
+    {
+      "epoch": 1.9135080981799968,
+      "grad_norm": 0.3826098539335381,
+      "learning_rate": 3.489702498436103e-06,
+      "loss": 0.1892,
+      "step": 5730
+    },
+    {
+      "epoch": 1.9138420437468693,
+      "grad_norm": 0.37640018946704573,
+      "learning_rate": 3.487850054131103e-06,
+      "loss": 0.1825,
+      "step": 5731
+    },
+    {
+      "epoch": 1.9141759893137418,
+      "grad_norm": 0.423540829528593,
+      "learning_rate": 3.4859978382541575e-06,
+      "loss": 0.2008,
+      "step": 5732
+    },
+    {
+      "epoch": 1.9145099348806145,
+      "grad_norm": 0.40110211348855723,
+      "learning_rate": 3.4841458510850656e-06,
+      "loss": 0.1928,
+      "step": 5733
+    },
+    {
+      "epoch": 1.9148438804474872,
+      "grad_norm": 0.4202765037424674,
+      "learning_rate": 3.482294092903592e-06,
+      "loss": 0.1913,
+      "step": 5734
+    },
+    {
+      "epoch": 1.9151778260143597,
+      "grad_norm": 0.3992739294133403,
+      "learning_rate": 3.480442563989466e-06,
+      "loss": 0.1855,
+      "step": 5735
+    },
+    {
+      "epoch": 1.9155117715812322,
+      "grad_norm": 0.39103883707441306,
+      "learning_rate": 3.4785912646223813e-06,
+      "loss": 0.198,
+      "step": 5736
+    },
+    {
+      "epoch": 1.915845717148105,
+      "grad_norm": 0.42159492620557487,
+      "learning_rate": 3.4767401950820003e-06,
+      "loss": 0.2061,
+      "step": 5737
+    },
+    {
+      "epoch": 1.9161796627149774,
+      "grad_norm": 0.38979029077665317,
+      "learning_rate": 3.4748893556479497e-06,
+      "loss": 0.185,
+      "step": 5738
+    },
+    {
+      "epoch": 1.91651360828185,
+      "grad_norm": 0.3767039745449479,
+      "learning_rate": 3.4730387465998194e-06,
+      "loss": 0.1911,
+      "step": 5739
+    },
+    {
+      "epoch": 1.9168475538487226,
+      "grad_norm": 0.4421262922770817,
+      "learning_rate": 3.4711883682171666e-06,
+      "loss": 0.2017,
+      "step": 5740
+    },
+    {
+      "epoch": 1.9171814994155953,
+      "grad_norm": 0.4073953632957591,
+      "learning_rate": 3.4693382207795114e-06,
+      "loss": 0.1913,
+      "step": 5741
+    },
+    {
+      "epoch": 1.9175154449824678,
+      "grad_norm": 0.3976848565725884,
+      "learning_rate": 3.4674883045663404e-06,
+      "loss": 0.1888,
+      "step": 5742
+    },
+    {
+      "epoch": 1.9178493905493403,
+      "grad_norm": 0.3853218853012135,
+      "learning_rate": 3.465638619857104e-06,
+      "loss": 0.1828,
+      "step": 5743
+    },
+    {
+      "epoch": 1.918183336116213,
+      "grad_norm": 0.40753207467467006,
+      "learning_rate": 3.463789166931223e-06,
+      "loss": 0.1907,
+      "step": 5744
+    },
+    {
+      "epoch": 1.9185172816830858,
+      "grad_norm": 0.3765078672210008,
+      "learning_rate": 3.4619399460680757e-06,
+      "loss": 0.1825,
+      "step": 5745
+    },
+    {
+      "epoch": 1.9188512272499583,
+      "grad_norm": 0.434986647967096,
+      "learning_rate": 3.460090957547011e-06,
+      "loss": 0.1806,
+      "step": 5746
+    },
+    {
+      "epoch": 1.9191851728168308,
+      "grad_norm": 0.40851362414259823,
+      "learning_rate": 3.4582422016473384e-06,
+      "loss": 0.1935,
+      "step": 5747
+    },
+    {
+      "epoch": 1.9195191183837035,
+      "grad_norm": 0.41083715107382207,
+      "learning_rate": 3.4563936786483345e-06,
+      "loss": 0.1988,
+      "step": 5748
+    },
+    {
+      "epoch": 1.9198530639505762,
+      "grad_norm": 0.4074208018702095,
+      "learning_rate": 3.454545388829239e-06,
+      "loss": 0.1944,
+      "step": 5749
+    },
+    {
+      "epoch": 1.9201870095174487,
+      "grad_norm": 0.39520620444751453,
+      "learning_rate": 3.4526973324692614e-06,
+      "loss": 0.1908,
+      "step": 5750
+    },
+    {
+      "epoch": 1.9205209550843212,
+      "grad_norm": 0.43906524174111994,
+      "learning_rate": 3.4508495098475712e-06,
+      "loss": 0.1992,
+      "step": 5751
+    },
+    {
+      "epoch": 1.920854900651194,
+      "grad_norm": 0.4502858723264202,
+      "learning_rate": 3.4490019212433035e-06,
+      "loss": 0.2044,
+      "step": 5752
+    },
+    {
+      "epoch": 1.9211888462180664,
+      "grad_norm": 0.40705908631119775,
+      "learning_rate": 3.447154566935557e-06,
+      "loss": 0.1943,
+      "step": 5753
+    },
+    {
+      "epoch": 1.921522791784939,
+      "grad_norm": 0.39945366422919054,
+      "learning_rate": 3.4453074472033975e-06,
+      "loss": 0.1863,
+      "step": 5754
+    },
+    {
+      "epoch": 1.9218567373518116,
+      "grad_norm": 0.39100600573805244,
+      "learning_rate": 3.443460562325853e-06,
+      "loss": 0.1823,
+      "step": 5755
+    },
+    {
+      "epoch": 1.9221906829186843,
+      "grad_norm": 0.3867960086824861,
+      "learning_rate": 3.4416139125819204e-06,
+      "loss": 0.1833,
+      "step": 5756
+    },
+    {
+      "epoch": 1.9225246284855568,
+      "grad_norm": 0.41121060111634883,
+      "learning_rate": 3.4397674982505546e-06,
+      "loss": 0.1887,
+      "step": 5757
+    },
+    {
+      "epoch": 1.9228585740524293,
+      "grad_norm": 0.42275941705337255,
+      "learning_rate": 3.43792131961068e-06,
+      "loss": 0.2092,
+      "step": 5758
+    },
+    {
+      "epoch": 1.923192519619302,
+      "grad_norm": 0.3843148443080461,
+      "learning_rate": 3.4360753769411816e-06,
+      "loss": 0.1749,
+      "step": 5759
+    },
+    {
+      "epoch": 1.9235264651861748,
+      "grad_norm": 0.39501568437932283,
+      "learning_rate": 3.4342296705209112e-06,
+      "loss": 0.1731,
+      "step": 5760
+    },
+    {
+      "epoch": 1.9238604107530473,
+      "grad_norm": 0.3831115342716545,
+      "learning_rate": 3.432384200628688e-06,
+      "loss": 0.1848,
+      "step": 5761
+    },
+    {
+      "epoch": 1.9241943563199198,
+      "grad_norm": 0.3925566970493553,
+      "learning_rate": 3.4305389675432882e-06,
+      "loss": 0.1923,
+      "step": 5762
+    },
+    {
+      "epoch": 1.9245283018867925,
+      "grad_norm": 0.4353970565860827,
+      "learning_rate": 3.4286939715434573e-06,
+      "loss": 0.1941,
+      "step": 5763
+    },
+    {
+      "epoch": 1.9248622474536652,
+      "grad_norm": 0.3853234833941961,
+      "learning_rate": 3.4268492129079047e-06,
+      "loss": 0.1862,
+      "step": 5764
+    },
+    {
+      "epoch": 1.9251961930205377,
+      "grad_norm": 0.43244855006490596,
+      "learning_rate": 3.4250046919153e-06,
+      "loss": 0.199,
+      "step": 5765
+    },
+    {
+      "epoch": 1.9255301385874102,
+      "grad_norm": 0.44410365215339354,
+      "learning_rate": 3.4231604088442806e-06,
+      "loss": 0.196,
+      "step": 5766
+    },
+    {
+      "epoch": 1.925864084154283,
+      "grad_norm": 0.4221818596900024,
+      "learning_rate": 3.4213163639734504e-06,
+      "loss": 0.1965,
+      "step": 5767
+    },
+    {
+      "epoch": 1.9261980297211556,
+      "grad_norm": 0.40430757506097365,
+      "learning_rate": 3.4194725575813707e-06,
+      "loss": 0.1871,
+      "step": 5768
+    },
+    {
+      "epoch": 1.9265319752880279,
+      "grad_norm": 0.39690720279215563,
+      "learning_rate": 3.417628989946572e-06,
+      "loss": 0.1866,
+      "step": 5769
+    },
+    {
+      "epoch": 1.9268659208549006,
+      "grad_norm": 0.41882654166179956,
+      "learning_rate": 3.415785661347546e-06,
+      "loss": 0.1976,
+      "step": 5770
+    },
+    {
+      "epoch": 1.9271998664217733,
+      "grad_norm": 0.42589760681235683,
+      "learning_rate": 3.4139425720627494e-06,
+      "loss": 0.1931,
+      "step": 5771
+    },
+    {
+      "epoch": 1.9275338119886458,
+      "grad_norm": 0.42698531652640753,
+      "learning_rate": 3.412099722370601e-06,
+      "loss": 0.1954,
+      "step": 5772
+    },
+    {
+      "epoch": 1.9278677575555183,
+      "grad_norm": 0.4458391967702739,
+      "learning_rate": 3.4102571125494877e-06,
+      "loss": 0.2061,
+      "step": 5773
+    },
+    {
+      "epoch": 1.928201703122391,
+      "grad_norm": 0.42935606022669165,
+      "learning_rate": 3.408414742877757e-06,
+      "loss": 0.2027,
+      "step": 5774
+    },
+    {
+      "epoch": 1.9285356486892637,
+      "grad_norm": 0.4383502567902757,
+      "learning_rate": 3.406572613633719e-06,
+      "loss": 0.1894,
+      "step": 5775
+    },
+    {
+      "epoch": 1.9288695942561362,
+      "grad_norm": 0.40057496070461995,
+      "learning_rate": 3.40473072509565e-06,
+      "loss": 0.199,
+      "step": 5776
+    },
+    {
+      "epoch": 1.9292035398230087,
+      "grad_norm": 0.39138634692508933,
+      "learning_rate": 3.4028890775417887e-06,
+      "loss": 0.1884,
+      "step": 5777
+    },
+    {
+      "epoch": 1.9295374853898815,
+      "grad_norm": 0.4266213221820064,
+      "learning_rate": 3.4010476712503367e-06,
+      "loss": 0.1957,
+      "step": 5778
+    },
+    {
+      "epoch": 1.9298714309567542,
+      "grad_norm": 0.41962434822821776,
+      "learning_rate": 3.3992065064994615e-06,
+      "loss": 0.1986,
+      "step": 5779
+    },
+    {
+      "epoch": 1.9302053765236267,
+      "grad_norm": 0.40653053851480586,
+      "learning_rate": 3.3973655835672923e-06,
+      "loss": 0.1872,
+      "step": 5780
+    },
+    {
+      "epoch": 1.9305393220904992,
+      "grad_norm": 0.3992762796239201,
+      "learning_rate": 3.3955249027319214e-06,
+      "loss": 0.1926,
+      "step": 5781
+    },
+    {
+      "epoch": 1.9308732676573719,
+      "grad_norm": 0.39309596029124866,
+      "learning_rate": 3.3936844642714073e-06,
+      "loss": 0.1885,
+      "step": 5782
+    },
+    {
+      "epoch": 1.9312072132242446,
+      "grad_norm": 0.3967296969011724,
+      "learning_rate": 3.3918442684637687e-06,
+      "loss": 0.1945,
+      "step": 5783
+    },
+    {
+      "epoch": 1.931541158791117,
+      "grad_norm": 0.37389292859170953,
+      "learning_rate": 3.3900043155869865e-06,
+      "loss": 0.1857,
+      "step": 5784
+    },
+    {
+      "epoch": 1.9318751043579896,
+      "grad_norm": 0.406835201145131,
+      "learning_rate": 3.388164605919012e-06,
+      "loss": 0.1912,
+      "step": 5785
+    },
+    {
+      "epoch": 1.9322090499248623,
+      "grad_norm": 0.40644414055673206,
+      "learning_rate": 3.3863251397377516e-06,
+      "loss": 0.1882,
+      "step": 5786
+    },
+    {
+      "epoch": 1.9325429954917348,
+      "grad_norm": 0.4534582493017801,
+      "learning_rate": 3.3844859173210797e-06,
+      "loss": 0.2098,
+      "step": 5787
+    },
+    {
+      "epoch": 1.9328769410586073,
+      "grad_norm": 0.40570770824932884,
+      "learning_rate": 3.382646938946832e-06,
+      "loss": 0.182,
+      "step": 5788
+    },
+    {
+      "epoch": 1.93321088662548,
+      "grad_norm": 0.4125254859784587,
+      "learning_rate": 3.3808082048928083e-06,
+      "loss": 0.1827,
+      "step": 5789
+    },
+    {
+      "epoch": 1.9335448321923527,
+      "grad_norm": 0.42050886372039403,
+      "learning_rate": 3.378969715436767e-06,
+      "loss": 0.1788,
+      "step": 5790
+    },
+    {
+      "epoch": 1.9338787777592252,
+      "grad_norm": 0.4029343552361385,
+      "learning_rate": 3.3771314708564408e-06,
+      "loss": 0.1932,
+      "step": 5791
+    },
+    {
+      "epoch": 1.9342127233260977,
+      "grad_norm": 0.4431266567668056,
+      "learning_rate": 3.3752934714295146e-06,
+      "loss": 0.2059,
+      "step": 5792
+    },
+    {
+      "epoch": 1.9345466688929704,
+      "grad_norm": 0.393345686460719,
+      "learning_rate": 3.373455717433639e-06,
+      "loss": 0.2035,
+      "step": 5793
+    },
+    {
+      "epoch": 1.9348806144598432,
+      "grad_norm": 0.4071248248099155,
+      "learning_rate": 3.3716182091464295e-06,
+      "loss": 0.1797,
+      "step": 5794
+    },
+    {
+      "epoch": 1.9352145600267157,
+      "grad_norm": 0.41122336422676053,
+      "learning_rate": 3.3697809468454634e-06,
+      "loss": 0.1922,
+      "step": 5795
+    },
+    {
+      "epoch": 1.9355485055935882,
+      "grad_norm": 0.4212957583988024,
+      "learning_rate": 3.3679439308082777e-06,
+      "loss": 0.1996,
+      "step": 5796
+    },
+    {
+      "epoch": 1.9358824511604609,
+      "grad_norm": 0.3994840928223961,
+      "learning_rate": 3.366107161312381e-06,
+      "loss": 0.1845,
+      "step": 5797
+    },
+    {
+      "epoch": 1.9362163967273336,
+      "grad_norm": 0.42573497457455245,
+      "learning_rate": 3.3642706386352355e-06,
+      "loss": 0.1994,
+      "step": 5798
+    },
+    {
+      "epoch": 1.936550342294206,
+      "grad_norm": 0.3631236184721829,
+      "learning_rate": 3.3624343630542707e-06,
+      "loss": 0.176,
+      "step": 5799
+    },
+    {
+      "epoch": 1.9368842878610786,
+      "grad_norm": 0.38488554412626735,
+      "learning_rate": 3.3605983348468764e-06,
+      "loss": 0.1882,
+      "step": 5800
+    },
+    {
+      "epoch": 1.9372182334279513,
+      "grad_norm": 0.47199555184017095,
+      "learning_rate": 3.3587625542904063e-06,
+      "loss": 0.2026,
+      "step": 5801
+    },
+    {
+      "epoch": 1.9375521789948238,
+      "grad_norm": 0.4601046109909035,
+      "learning_rate": 3.356927021662178e-06,
+      "loss": 0.1916,
+      "step": 5802
+    },
+    {
+      "epoch": 1.9378861245616963,
+      "grad_norm": 0.3994198582194862,
+      "learning_rate": 3.3550917372394696e-06,
+      "loss": 0.184,
+      "step": 5803
+    },
+    {
+      "epoch": 1.938220070128569,
+      "grad_norm": 0.3734853608354193,
+      "learning_rate": 3.353256701299522e-06,
+      "loss": 0.1833,
+      "step": 5804
+    },
+    {
+      "epoch": 1.9385540156954417,
+      "grad_norm": 0.38879660452435616,
+      "learning_rate": 3.3514219141195404e-06,
+      "loss": 0.1845,
+      "step": 5805
+    },
+    {
+      "epoch": 1.9388879612623142,
+      "grad_norm": 0.38560538917320175,
+      "learning_rate": 3.3495873759766897e-06,
+      "loss": 0.1926,
+      "step": 5806
+    },
+    {
+      "epoch": 1.9392219068291867,
+      "grad_norm": 0.38353142779479343,
+      "learning_rate": 3.347753087148098e-06,
+      "loss": 0.1838,
+      "step": 5807
+    },
+    {
+      "epoch": 1.9395558523960594,
+      "grad_norm": 0.44448297494571126,
+      "learning_rate": 3.3459190479108583e-06,
+      "loss": 0.199,
+      "step": 5808
+    },
+    {
+      "epoch": 1.9398897979629321,
+      "grad_norm": 0.4529776498547492,
+      "learning_rate": 3.344085258542022e-06,
+      "loss": 0.1979,
+      "step": 5809
+    },
+    {
+      "epoch": 1.9402237435298046,
+      "grad_norm": 0.4124811126918051,
+      "learning_rate": 3.3422517193186056e-06,
+      "loss": 0.2068,
+      "step": 5810
+    },
+    {
+      "epoch": 1.9405576890966771,
+      "grad_norm": 0.39920971830243007,
+      "learning_rate": 3.340418430517586e-06,
+      "loss": 0.1938,
+      "step": 5811
+    },
+    {
+      "epoch": 1.9408916346635499,
+      "grad_norm": 0.4222813253850214,
+      "learning_rate": 3.338585392415904e-06,
+      "loss": 0.2053,
+      "step": 5812
+    },
+    {
+      "epoch": 1.9412255802304226,
+      "grad_norm": 0.4161721442405356,
+      "learning_rate": 3.3367526052904585e-06,
+      "loss": 0.1787,
+      "step": 5813
+    },
+    {
+      "epoch": 1.941559525797295,
+      "grad_norm": 0.4879776692990766,
+      "learning_rate": 3.3349200694181182e-06,
+      "loss": 0.2053,
+      "step": 5814
+    },
+    {
+      "epoch": 1.9418934713641676,
+      "grad_norm": 0.4021223636861502,
+      "learning_rate": 3.333087785075707e-06,
+      "loss": 0.1964,
+      "step": 5815
+    },
+    {
+      "epoch": 1.9422274169310403,
+      "grad_norm": 0.37651593810978384,
+      "learning_rate": 3.3312557525400133e-06,
+      "loss": 0.1715,
+      "step": 5816
+    },
+    {
+      "epoch": 1.942561362497913,
+      "grad_norm": 0.4319318531805005,
+      "learning_rate": 3.329423972087787e-06,
+      "loss": 0.1955,
+      "step": 5817
+    },
+    {
+      "epoch": 1.9428953080647853,
+      "grad_norm": 0.4141613206128405,
+      "learning_rate": 3.3275924439957397e-06,
+      "loss": 0.2074,
+      "step": 5818
+    },
+    {
+      "epoch": 1.943229253631658,
+      "grad_norm": 0.38424218929059795,
+      "learning_rate": 3.3257611685405444e-06,
+      "loss": 0.1827,
+      "step": 5819
+    },
+    {
+      "epoch": 1.9435631991985307,
+      "grad_norm": 0.45704517395065586,
+      "learning_rate": 3.3239301459988395e-06,
+      "loss": 0.204,
+      "step": 5820
+    },
+    {
+      "epoch": 1.9438971447654032,
+      "grad_norm": 0.4519944942975536,
+      "learning_rate": 3.322099376647221e-06,
+      "loss": 0.1961,
+      "step": 5821
+    },
+    {
+      "epoch": 1.9442310903322757,
+      "grad_norm": 0.3865117153162869,
+      "learning_rate": 3.320268860762249e-06,
+      "loss": 0.1842,
+      "step": 5822
+    },
+    {
+      "epoch": 1.9445650358991484,
+      "grad_norm": 0.39394807708092394,
+      "learning_rate": 3.318438598620444e-06,
+      "loss": 0.1934,
+      "step": 5823
+    },
+    {
+      "epoch": 1.9448989814660211,
+      "grad_norm": 0.4040471014908394,
+      "learning_rate": 3.316608590498287e-06,
+      "loss": 0.1833,
+      "step": 5824
+    },
+    {
+      "epoch": 1.9452329270328936,
+      "grad_norm": 0.42368161376371266,
+      "learning_rate": 3.314778836672224e-06,
+      "loss": 0.1989,
+      "step": 5825
+    },
+    {
+      "epoch": 1.9455668725997661,
+      "grad_norm": 0.3799786063158331,
+      "learning_rate": 3.312949337418661e-06,
+      "loss": 0.1903,
+      "step": 5826
+    },
+    {
+      "epoch": 1.9459008181666388,
+      "grad_norm": 0.42001314369222936,
+      "learning_rate": 3.311120093013964e-06,
+      "loss": 0.2029,
+      "step": 5827
+    },
+    {
+      "epoch": 1.9462347637335116,
+      "grad_norm": 0.4491441693231125,
+      "learning_rate": 3.3092911037344642e-06,
+      "loss": 0.1899,
+      "step": 5828
+    },
+    {
+      "epoch": 1.946568709300384,
+      "grad_norm": 0.40616158884667664,
+      "learning_rate": 3.30746236985645e-06,
+      "loss": 0.1902,
+      "step": 5829
+    },
+    {
+      "epoch": 1.9469026548672566,
+      "grad_norm": 0.41435867845292346,
+      "learning_rate": 3.305633891656175e-06,
+      "loss": 0.1944,
+      "step": 5830
+    },
+    {
+      "epoch": 1.9472366004341293,
+      "grad_norm": 0.3835569025294799,
+      "learning_rate": 3.3038056694098485e-06,
+      "loss": 0.1866,
+      "step": 5831
+    },
+    {
+      "epoch": 1.947570546001002,
+      "grad_norm": 0.41557182152841543,
+      "learning_rate": 3.3019777033936497e-06,
+      "loss": 0.1941,
+      "step": 5832
+    },
+    {
+      "epoch": 1.9479044915678745,
+      "grad_norm": 0.4136102028672037,
+      "learning_rate": 3.3001499938837124e-06,
+      "loss": 0.199,
+      "step": 5833
+    },
+    {
+      "epoch": 1.948238437134747,
+      "grad_norm": 0.42808162442284475,
+      "learning_rate": 3.2983225411561338e-06,
+      "loss": 0.1888,
+      "step": 5834
+    },
+    {
+      "epoch": 1.9485723827016197,
+      "grad_norm": 0.4239596741209335,
+      "learning_rate": 3.296495345486971e-06,
+      "loss": 0.1947,
+      "step": 5835
+    },
+    {
+      "epoch": 1.9489063282684922,
+      "grad_norm": 0.38609956620089547,
+      "learning_rate": 3.294668407152245e-06,
+      "loss": 0.1887,
+      "step": 5836
+    },
+    {
+      "epoch": 1.9492402738353647,
+      "grad_norm": 0.39282968924115996,
+      "learning_rate": 3.2928417264279338e-06,
+      "loss": 0.1866,
+      "step": 5837
+    },
+    {
+      "epoch": 1.9495742194022374,
+      "grad_norm": 0.37824694049603946,
+      "learning_rate": 3.2910153035899826e-06,
+      "loss": 0.1867,
+      "step": 5838
+    },
+    {
+      "epoch": 1.9499081649691101,
+      "grad_norm": 0.43800186138716113,
+      "learning_rate": 3.2891891389142933e-06,
+      "loss": 0.2164,
+      "step": 5839
+    },
+    {
+      "epoch": 1.9502421105359826,
+      "grad_norm": 0.38749810000277757,
+      "learning_rate": 3.2873632326767278e-06,
+      "loss": 0.1806,
+      "step": 5840
+    },
+    {
+      "epoch": 1.9505760561028551,
+      "grad_norm": 0.393525628067674,
+      "learning_rate": 3.2855375851531122e-06,
+      "loss": 0.1861,
+      "step": 5841
+    },
+    {
+      "epoch": 1.9509100016697278,
+      "grad_norm": 0.4351462404079153,
+      "learning_rate": 3.283712196619229e-06,
+      "loss": 0.2038,
+      "step": 5842
+    },
+    {
+      "epoch": 1.9512439472366006,
+      "grad_norm": 0.42090443021448815,
+      "learning_rate": 3.2818870673508297e-06,
+      "loss": 0.2048,
+      "step": 5843
+    },
+    {
+      "epoch": 1.951577892803473,
+      "grad_norm": 0.3790280867900267,
+      "learning_rate": 3.2800621976236184e-06,
+      "loss": 0.1874,
+      "step": 5844
+    },
+    {
+      "epoch": 1.9519118383703455,
+      "grad_norm": 0.40688808218095657,
+      "learning_rate": 3.2782375877132643e-06,
+      "loss": 0.1808,
+      "step": 5845
+    },
+    {
+      "epoch": 1.9522457839372183,
+      "grad_norm": 0.4256516285672986,
+      "learning_rate": 3.276413237895395e-06,
+      "loss": 0.1954,
+      "step": 5846
+    },
+    {
+      "epoch": 1.952579729504091,
+      "grad_norm": 0.4171161147626943,
+      "learning_rate": 3.2745891484456016e-06,
+      "loss": 0.2009,
+      "step": 5847
+    },
+    {
+      "epoch": 1.9529136750709635,
+      "grad_norm": 0.5044605426457384,
+      "learning_rate": 3.2727653196394314e-06,
+      "loss": 0.194,
+      "step": 5848
+    },
+    {
+      "epoch": 1.953247620637836,
+      "grad_norm": 0.40522989200496995,
+      "learning_rate": 3.270941751752398e-06,
+      "loss": 0.191,
+      "step": 5849
+    },
+    {
+      "epoch": 1.9535815662047087,
+      "grad_norm": 0.39708955867580614,
+      "learning_rate": 3.269118445059973e-06,
+      "loss": 0.1799,
+      "step": 5850
+    },
+    {
+      "epoch": 1.9539155117715812,
+      "grad_norm": 0.40876661685989146,
+      "learning_rate": 3.267295399837587e-06,
+      "loss": 0.1914,
+      "step": 5851
+    },
+    {
+      "epoch": 1.9542494573384537,
+      "grad_norm": 0.39832148352306906,
+      "learning_rate": 3.2654726163606333e-06,
+      "loss": 0.1843,
+      "step": 5852
+    },
+    {
+      "epoch": 1.9545834029053264,
+      "grad_norm": 0.4339968232593733,
+      "learning_rate": 3.2636500949044637e-06,
+      "loss": 0.204,
+      "step": 5853
+    },
+    {
+      "epoch": 1.9549173484721991,
+      "grad_norm": 0.40521624053581823,
+      "learning_rate": 3.2618278357443913e-06,
+      "loss": 0.1923,
+      "step": 5854
+    },
+    {
+      "epoch": 1.9552512940390716,
+      "grad_norm": 0.40925796162009104,
+      "learning_rate": 3.260005839155691e-06,
+      "loss": 0.1942,
+      "step": 5855
+    },
+    {
+      "epoch": 1.955585239605944,
+      "grad_norm": 0.4065519345018825,
+      "learning_rate": 3.258184105413597e-06,
+      "loss": 0.1929,
+      "step": 5856
+    },
+    {
+      "epoch": 1.9559191851728168,
+      "grad_norm": 0.40373783185327916,
+      "learning_rate": 3.256362634793303e-06,
+      "loss": 0.1836,
+      "step": 5857
+    },
+    {
+      "epoch": 1.9562531307396895,
+      "grad_norm": 0.466912546541022,
+      "learning_rate": 3.2545414275699638e-06,
+      "loss": 0.2051,
+      "step": 5858
+    },
+    {
+      "epoch": 1.956587076306562,
+      "grad_norm": 0.4119496217398985,
+      "learning_rate": 3.2527204840186944e-06,
+      "loss": 0.2009,
+      "step": 5859
+    },
+    {
+      "epoch": 1.9569210218734345,
+      "grad_norm": 0.427568161677971,
+      "learning_rate": 3.2508998044145674e-06,
+      "loss": 0.1853,
+      "step": 5860
+    },
+    {
+      "epoch": 1.9572549674403072,
+      "grad_norm": 0.3964003737577245,
+      "learning_rate": 3.249079389032621e-06,
+      "loss": 0.1931,
+      "step": 5861
+    },
+    {
+      "epoch": 1.95758891300718,
+      "grad_norm": 0.3733715502028118,
+      "learning_rate": 3.247259238147851e-06,
+      "loss": 0.1891,
+      "step": 5862
+    },
+    {
+      "epoch": 1.9579228585740525,
+      "grad_norm": 0.387516171710147,
+      "learning_rate": 3.245439352035209e-06,
+      "loss": 0.1937,
+      "step": 5863
+    },
+    {
+      "epoch": 1.958256804140925,
+      "grad_norm": 0.41829601573720177,
+      "learning_rate": 3.243619730969614e-06,
+      "loss": 0.2014,
+      "step": 5864
+    },
+    {
+      "epoch": 1.9585907497077977,
+      "grad_norm": 0.37178507032150165,
+      "learning_rate": 3.2418003752259374e-06,
+      "loss": 0.1829,
+      "step": 5865
+    },
+    {
+      "epoch": 1.9589246952746704,
+      "grad_norm": 0.3755322252943328,
+      "learning_rate": 3.239981285079016e-06,
+      "loss": 0.1816,
+      "step": 5866
+    },
+    {
+      "epoch": 1.9592586408415427,
+      "grad_norm": 0.5064830664211579,
+      "learning_rate": 3.238162460803646e-06,
+      "loss": 0.1801,
+      "step": 5867
+    },
+    {
+      "epoch": 1.9595925864084154,
+      "grad_norm": 0.40514585437042433,
+      "learning_rate": 3.2363439026745813e-06,
+      "loss": 0.1937,
+      "step": 5868
+    },
+    {
+      "epoch": 1.959926531975288,
+      "grad_norm": 0.3949337542067921,
+      "learning_rate": 3.2345256109665366e-06,
+      "loss": 0.1769,
+      "step": 5869
+    },
+    {
+      "epoch": 1.9602604775421606,
+      "grad_norm": 0.39658019897229246,
+      "learning_rate": 3.2327075859541867e-06,
+      "loss": 0.1755,
+      "step": 5870
+    },
+    {
+      "epoch": 1.960594423109033,
+      "grad_norm": 0.4007129170784194,
+      "learning_rate": 3.2308898279121646e-06,
+      "loss": 0.1887,
+      "step": 5871
+    },
+    {
+      "epoch": 1.9609283686759058,
+      "grad_norm": 0.407882174298797,
+      "learning_rate": 3.2290723371150627e-06,
+      "loss": 0.1933,
+      "step": 5872
+    },
+    {
+      "epoch": 1.9612623142427785,
+      "grad_norm": 0.38390199585380574,
+      "learning_rate": 3.2272551138374387e-06,
+      "loss": 0.1857,
+      "step": 5873
+    },
+    {
+      "epoch": 1.961596259809651,
+      "grad_norm": 0.4144353996140741,
+      "learning_rate": 3.2254381583538025e-06,
+      "loss": 0.2052,
+      "step": 5874
+    },
+    {
+      "epoch": 1.9619302053765235,
+      "grad_norm": 0.42520106945104263,
+      "learning_rate": 3.223621470938628e-06,
+      "loss": 0.1989,
+      "step": 5875
+    },
+    {
+      "epoch": 1.9622641509433962,
+      "grad_norm": 0.39528450044868774,
+      "learning_rate": 3.2218050518663457e-06,
+      "loss": 0.1952,
+      "step": 5876
+    },
+    {
+      "epoch": 1.962598096510269,
+      "grad_norm": 0.38997068609060653,
+      "learning_rate": 3.219988901411347e-06,
+      "loss": 0.1905,
+      "step": 5877
+    },
+    {
+      "epoch": 1.9629320420771414,
+      "grad_norm": 0.43396687286925384,
+      "learning_rate": 3.218173019847985e-06,
+      "loss": 0.195,
+      "step": 5878
+    },
+    {
+      "epoch": 1.963265987644014,
+      "grad_norm": 0.41610373323796274,
+      "learning_rate": 3.2163574074505686e-06,
+      "loss": 0.2019,
+      "step": 5879
+    },
+    {
+      "epoch": 1.9635999332108867,
+      "grad_norm": 0.41341146951828495,
+      "learning_rate": 3.214542064493367e-06,
+      "loss": 0.2027,
+      "step": 5880
+    },
+    {
+      "epoch": 1.9639338787777594,
+      "grad_norm": 0.41138072181872515,
+      "learning_rate": 3.2127269912506103e-06,
+      "loss": 0.1868,
+      "step": 5881
+    },
+    {
+      "epoch": 1.9642678243446319,
+      "grad_norm": 0.3866031619654102,
+      "learning_rate": 3.210912187996486e-06,
+      "loss": 0.1932,
+      "step": 5882
+    },
+    {
+      "epoch": 1.9646017699115044,
+      "grad_norm": 0.4016530046534839,
+      "learning_rate": 3.2090976550051393e-06,
+      "loss": 0.2022,
+      "step": 5883
+    },
+    {
+      "epoch": 1.964935715478377,
+      "grad_norm": 0.4546880535883009,
+      "learning_rate": 3.207283392550681e-06,
+      "loss": 0.1937,
+      "step": 5884
+    },
+    {
+      "epoch": 1.9652696610452496,
+      "grad_norm": 0.3751451368240068,
+      "learning_rate": 3.2054694009071753e-06,
+      "loss": 0.1766,
+      "step": 5885
+    },
+    {
+      "epoch": 1.965603606612122,
+      "grad_norm": 0.4249952733519631,
+      "learning_rate": 3.2036556803486465e-06,
+      "loss": 0.1897,
+      "step": 5886
+    },
+    {
+      "epoch": 1.9659375521789948,
+      "grad_norm": 0.4289533302323695,
+      "learning_rate": 3.2018422311490778e-06,
+      "loss": 0.1872,
+      "step": 5887
+    },
+    {
+      "epoch": 1.9662714977458675,
+      "grad_norm": 0.4156318288794827,
+      "learning_rate": 3.200029053582413e-06,
+      "loss": 0.2013,
+      "step": 5888
+    },
+    {
+      "epoch": 1.96660544331274,
+      "grad_norm": 0.4234857984820954,
+      "learning_rate": 3.1982161479225514e-06,
+      "loss": 0.2035,
+      "step": 5889
+    },
+    {
+      "epoch": 1.9669393888796125,
+      "grad_norm": 0.4631603579410054,
+      "learning_rate": 3.196403514443358e-06,
+      "loss": 0.212,
+      "step": 5890
+    },
+    {
+      "epoch": 1.9672733344464852,
+      "grad_norm": 0.39403994004750187,
+      "learning_rate": 3.19459115341865e-06,
+      "loss": 0.1785,
+      "step": 5891
+    },
+    {
+      "epoch": 1.967607280013358,
+      "grad_norm": 0.3926567448420319,
+      "learning_rate": 3.1927790651222073e-06,
+      "loss": 0.189,
+      "step": 5892
+    },
+    {
+      "epoch": 1.9679412255802304,
+      "grad_norm": 0.3989322852990989,
+      "learning_rate": 3.1909672498277656e-06,
+      "loss": 0.2059,
+      "step": 5893
+    },
+    {
+      "epoch": 1.968275171147103,
+      "grad_norm": 0.42083685608122595,
+      "learning_rate": 3.1891557078090218e-06,
+      "loss": 0.1998,
+      "step": 5894
+    },
+    {
+      "epoch": 1.9686091167139756,
+      "grad_norm": 0.4383329776305396,
+      "learning_rate": 3.187344439339628e-06,
+      "loss": 0.1941,
+      "step": 5895
+    },
+    {
+      "epoch": 1.9689430622808484,
+      "grad_norm": 0.41449458340242895,
+      "learning_rate": 3.1855334446932025e-06,
+      "loss": 0.1944,
+      "step": 5896
+    },
+    {
+      "epoch": 1.9692770078477209,
+      "grad_norm": 0.3717497165199124,
+      "learning_rate": 3.1837227241433145e-06,
+      "loss": 0.1877,
+      "step": 5897
+    },
+    {
+      "epoch": 1.9696109534145934,
+      "grad_norm": 0.4011174641338493,
+      "learning_rate": 3.181912277963495e-06,
+      "loss": 0.1987,
+      "step": 5898
+    },
+    {
+      "epoch": 1.969944898981466,
+      "grad_norm": 0.38902286671665814,
+      "learning_rate": 3.180102106427233e-06,
+      "loss": 0.1867,
+      "step": 5899
+    },
+    {
+      "epoch": 1.9702788445483386,
+      "grad_norm": 0.42521419169467267,
+      "learning_rate": 3.178292209807976e-06,
+      "loss": 0.2018,
+      "step": 5900
+    },
+    {
+      "epoch": 1.970612790115211,
+      "grad_norm": 0.4065707042117532,
+      "learning_rate": 3.1764825883791306e-06,
+      "loss": 0.1885,
+      "step": 5901
+    },
+    {
+      "epoch": 1.9709467356820838,
+      "grad_norm": 0.412725809040078,
+      "learning_rate": 3.174673242414062e-06,
+      "loss": 0.1897,
+      "step": 5902
+    },
+    {
+      "epoch": 1.9712806812489565,
+      "grad_norm": 0.375413582686983,
+      "learning_rate": 3.1728641721860925e-06,
+      "loss": 0.185,
+      "step": 5903
+    },
+    {
+      "epoch": 1.971614626815829,
+      "grad_norm": 0.41467689840017014,
+      "learning_rate": 3.1710553779685036e-06,
+      "loss": 0.1973,
+      "step": 5904
+    },
+    {
+      "epoch": 1.9719485723827015,
+      "grad_norm": 0.41671275410394143,
+      "learning_rate": 3.169246860034535e-06,
+      "loss": 0.2082,
+      "step": 5905
+    },
+    {
+      "epoch": 1.9722825179495742,
+      "grad_norm": 0.4057263289652559,
+      "learning_rate": 3.1674386186573853e-06,
+      "loss": 0.1897,
+      "step": 5906
+    },
+    {
+      "epoch": 1.972616463516447,
+      "grad_norm": 0.38913647089351844,
+      "learning_rate": 3.1656306541102073e-06,
+      "loss": 0.1856,
+      "step": 5907
+    },
+    {
+      "epoch": 1.9729504090833194,
+      "grad_norm": 0.3486855638934541,
+      "learning_rate": 3.16382296666612e-06,
+      "loss": 0.1679,
+      "step": 5908
+    },
+    {
+      "epoch": 1.973284354650192,
+      "grad_norm": 0.4014595623992815,
+      "learning_rate": 3.1620155565981942e-06,
+      "loss": 0.191,
+      "step": 5909
+    },
+    {
+      "epoch": 1.9736183002170646,
+      "grad_norm": 0.40381926251995,
+      "learning_rate": 3.1602084241794595e-06,
+      "loss": 0.1839,
+      "step": 5910
+    },
+    {
+      "epoch": 1.9739522457839374,
+      "grad_norm": 0.37605441183464833,
+      "learning_rate": 3.158401569682906e-06,
+      "loss": 0.1875,
+      "step": 5911
+    },
+    {
+      "epoch": 1.9742861913508098,
+      "grad_norm": 0.3777469896395668,
+      "learning_rate": 3.156594993381479e-06,
+      "loss": 0.1886,
+      "step": 5912
+    },
+    {
+      "epoch": 1.9746201369176823,
+      "grad_norm": 0.3736417903821561,
+      "learning_rate": 3.154788695548082e-06,
+      "loss": 0.1822,
+      "step": 5913
+    },
+    {
+      "epoch": 1.974954082484555,
+      "grad_norm": 0.3663210624038179,
+      "learning_rate": 3.152982676455581e-06,
+      "loss": 0.1726,
+      "step": 5914
+    },
+    {
+      "epoch": 1.9752880280514278,
+      "grad_norm": 0.39921583039669706,
+      "learning_rate": 3.151176936376794e-06,
+      "loss": 0.1952,
+      "step": 5915
+    },
+    {
+      "epoch": 1.9756219736183,
+      "grad_norm": 0.41004780294376225,
+      "learning_rate": 3.1493714755845013e-06,
+      "loss": 0.1892,
+      "step": 5916
+    },
+    {
+      "epoch": 1.9759559191851728,
+      "grad_norm": 0.3953073598725052,
+      "learning_rate": 3.1475662943514366e-06,
+      "loss": 0.1846,
+      "step": 5917
+    },
+    {
+      "epoch": 1.9762898647520455,
+      "grad_norm": 0.41934031809748096,
+      "learning_rate": 3.145761392950293e-06,
+      "loss": 0.1938,
+      "step": 5918
+    },
+    {
+      "epoch": 1.976623810318918,
+      "grad_norm": 0.3991536627450681,
+      "learning_rate": 3.1439567716537268e-06,
+      "loss": 0.1842,
+      "step": 5919
+    },
+    {
+      "epoch": 1.9769577558857905,
+      "grad_norm": 0.42211108035671385,
+      "learning_rate": 3.142152430734343e-06,
+      "loss": 0.1926,
+      "step": 5920
+    },
+    {
+      "epoch": 1.9772917014526632,
+      "grad_norm": 0.3874004858043167,
+      "learning_rate": 3.140348370464711e-06,
+      "loss": 0.1825,
+      "step": 5921
+    },
+    {
+      "epoch": 1.977625647019536,
+      "grad_norm": 0.4959700492993034,
+      "learning_rate": 3.138544591117354e-06,
+      "loss": 0.1881,
+      "step": 5922
+    },
+    {
+      "epoch": 1.9779595925864084,
+      "grad_norm": 0.4058894672658346,
+      "learning_rate": 3.1367410929647544e-06,
+      "loss": 0.2038,
+      "step": 5923
+    },
+    {
+      "epoch": 1.978293538153281,
+      "grad_norm": 0.4178550089316045,
+      "learning_rate": 3.1349378762793515e-06,
+      "loss": 0.2012,
+      "step": 5924
+    },
+    {
+      "epoch": 1.9786274837201536,
+      "grad_norm": 0.4004663381408806,
+      "learning_rate": 3.133134941333543e-06,
+      "loss": 0.1837,
+      "step": 5925
+    },
+    {
+      "epoch": 1.9789614292870263,
+      "grad_norm": 0.41889122105286186,
+      "learning_rate": 3.1313322883996833e-06,
+      "loss": 0.2016,
+      "step": 5926
+    },
+    {
+      "epoch": 1.9792953748538988,
+      "grad_norm": 0.39423179870453967,
+      "learning_rate": 3.129529917750085e-06,
+      "loss": 0.1859,
+      "step": 5927
+    },
+    {
+      "epoch": 1.9796293204207713,
+      "grad_norm": 0.4176399622736979,
+      "learning_rate": 3.1277278296570157e-06,
+      "loss": 0.1969,
+      "step": 5928
+    },
+    {
+      "epoch": 1.979963265987644,
+      "grad_norm": 0.720364647524329,
+      "learning_rate": 3.1259260243927035e-06,
+      "loss": 0.1926,
+      "step": 5929
+    },
+    {
+      "epoch": 1.9802972115545168,
+      "grad_norm": 0.41290547958373885,
+      "learning_rate": 3.12412450222933e-06,
+      "loss": 0.1898,
+      "step": 5930
+    },
+    {
+      "epoch": 1.9806311571213893,
+      "grad_norm": 0.3671561676737529,
+      "learning_rate": 3.12232326343904e-06,
+      "loss": 0.1741,
+      "step": 5931
+    },
+    {
+      "epoch": 1.9809651026882618,
+      "grad_norm": 0.4006350592225924,
+      "learning_rate": 3.1205223082939302e-06,
+      "loss": 0.1908,
+      "step": 5932
+    },
+    {
+      "epoch": 1.9812990482551345,
+      "grad_norm": 0.398681215978086,
+      "learning_rate": 3.1187216370660558e-06,
+      "loss": 0.1936,
+      "step": 5933
+    },
+    {
+      "epoch": 1.981632993822007,
+      "grad_norm": 0.43415498393757535,
+      "learning_rate": 3.1169212500274294e-06,
+      "loss": 0.1943,
+      "step": 5934
+    },
+    {
+      "epoch": 1.9819669393888795,
+      "grad_norm": 0.4135058440013624,
+      "learning_rate": 3.11512114745002e-06,
+      "loss": 0.2106,
+      "step": 5935
+    },
+    {
+      "epoch": 1.9823008849557522,
+      "grad_norm": 0.3846174813571026,
+      "learning_rate": 3.113321329605754e-06,
+      "loss": 0.1796,
+      "step": 5936
+    },
+    {
+      "epoch": 1.982634830522625,
+      "grad_norm": 0.383507587269331,
+      "learning_rate": 3.1115217967665174e-06,
+      "loss": 0.1855,
+      "step": 5937
+    },
+    {
+      "epoch": 1.9829687760894974,
+      "grad_norm": 0.37537114616087025,
+      "learning_rate": 3.1097225492041494e-06,
+      "loss": 0.1747,
+      "step": 5938
+    },
+    {
+      "epoch": 1.98330272165637,
+      "grad_norm": 0.39957620124069226,
+      "learning_rate": 3.107923587190448e-06,
+      "loss": 0.1931,
+      "step": 5939
+    },
+    {
+      "epoch": 1.9836366672232426,
+      "grad_norm": 0.403505274568005,
+      "learning_rate": 3.106124910997168e-06,
+      "loss": 0.1886,
+      "step": 5940
+    },
+    {
+      "epoch": 1.9839706127901153,
+      "grad_norm": 0.38362226676493627,
+      "learning_rate": 3.1043265208960187e-06,
+      "loss": 0.1823,
+      "step": 5941
+    },
+    {
+      "epoch": 1.9843045583569878,
+      "grad_norm": 0.39122961355599384,
+      "learning_rate": 3.102528417158668e-06,
+      "loss": 0.1859,
+      "step": 5942
+    },
+    {
+      "epoch": 1.9846385039238603,
+      "grad_norm": 0.41049641938952836,
+      "learning_rate": 3.1007306000567434e-06,
+      "loss": 0.1918,
+      "step": 5943
+    },
+    {
+      "epoch": 1.984972449490733,
+      "grad_norm": 0.36272757102046843,
+      "learning_rate": 3.0989330698618248e-06,
+      "loss": 0.175,
+      "step": 5944
+    },
+    {
+      "epoch": 1.9853063950576058,
+      "grad_norm": 0.4101938673103447,
+      "learning_rate": 3.097135826845451e-06,
+      "loss": 0.193,
+      "step": 5945
+    },
+    {
+      "epoch": 1.9856403406244783,
+      "grad_norm": 0.3690806719292599,
+      "learning_rate": 3.0953388712791155e-06,
+      "loss": 0.1597,
+      "step": 5946
+    },
+    {
+      "epoch": 1.9859742861913507,
+      "grad_norm": 0.4225456794662782,
+      "learning_rate": 3.09354220343427e-06,
+      "loss": 0.1867,
+      "step": 5947
+    },
+    {
+      "epoch": 1.9863082317582235,
+      "grad_norm": 0.4282830064350807,
+      "learning_rate": 3.0917458235823215e-06,
+      "loss": 0.2025,
+      "step": 5948
+    },
+    {
+      "epoch": 1.986642177325096,
+      "grad_norm": 0.4349543625503407,
+      "learning_rate": 3.089949731994637e-06,
+      "loss": 0.2014,
+      "step": 5949
+    },
+    {
+      "epoch": 1.9869761228919685,
+      "grad_norm": 0.37381968406717025,
+      "learning_rate": 3.088153928942535e-06,
+      "loss": 0.1842,
+      "step": 5950
+    },
+    {
+      "epoch": 1.9873100684588412,
+      "grad_norm": 0.39590274000895553,
+      "learning_rate": 3.0863584146972935e-06,
+      "loss": 0.1806,
+      "step": 5951
+    },
+    {
+      "epoch": 1.987644014025714,
+      "grad_norm": 0.3863725717307747,
+      "learning_rate": 3.084563189530146e-06,
+      "loss": 0.1792,
+      "step": 5952
+    },
+    {
+      "epoch": 1.9879779595925864,
+      "grad_norm": 0.40619770607208006,
+      "learning_rate": 3.0827682537122817e-06,
+      "loss": 0.1942,
+      "step": 5953
+    },
+    {
+      "epoch": 1.9883119051594589,
+      "grad_norm": 0.4483264558049833,
+      "learning_rate": 3.0809736075148456e-06,
+      "loss": 0.2014,
+      "step": 5954
+    },
+    {
+      "epoch": 1.9886458507263316,
+      "grad_norm": 0.4199360913214994,
+      "learning_rate": 3.0791792512089443e-06,
+      "loss": 0.182,
+      "step": 5955
+    },
+    {
+      "epoch": 1.9889797962932043,
+      "grad_norm": 0.4058012474543562,
+      "learning_rate": 3.0773851850656335e-06,
+      "loss": 0.1855,
+      "step": 5956
+    },
+    {
+      "epoch": 1.9893137418600768,
+      "grad_norm": 0.3894507983132781,
+      "learning_rate": 3.075591409355929e-06,
+      "loss": 0.1914,
+      "step": 5957
+    },
+    {
+      "epoch": 1.9896476874269493,
+      "grad_norm": 0.39437724621404885,
+      "learning_rate": 3.073797924350801e-06,
+      "loss": 0.195,
+      "step": 5958
+    },
+    {
+      "epoch": 1.989981632993822,
+      "grad_norm": 0.39982310703857066,
+      "learning_rate": 3.0720047303211746e-06,
+      "loss": 0.1943,
+      "step": 5959
+    },
+    {
+      "epoch": 1.9903155785606947,
+      "grad_norm": 0.43063169248758626,
+      "learning_rate": 3.0702118275379376e-06,
+      "loss": 0.1986,
+      "step": 5960
+    },
+    {
+      "epoch": 1.9906495241275672,
+      "grad_norm": 0.39695348018292154,
+      "learning_rate": 3.0684192162719263e-06,
+      "loss": 0.1922,
+      "step": 5961
+    },
+    {
+      "epoch": 1.9909834696944397,
+      "grad_norm": 0.398425444924352,
+      "learning_rate": 3.066626896793936e-06,
+      "loss": 0.1887,
+      "step": 5962
+    },
+    {
+      "epoch": 1.9913174152613125,
+      "grad_norm": 0.3756162081582355,
+      "learning_rate": 3.0648348693747177e-06,
+      "loss": 0.1855,
+      "step": 5963
+    },
+    {
+      "epoch": 1.9916513608281852,
+      "grad_norm": 0.4199243577842171,
+      "learning_rate": 3.063043134284979e-06,
+      "loss": 0.1933,
+      "step": 5964
+    },
+    {
+      "epoch": 1.9919853063950574,
+      "grad_norm": 0.38131482579075354,
+      "learning_rate": 3.0612516917953783e-06,
+      "loss": 0.1843,
+      "step": 5965
+    },
+    {
+      "epoch": 1.9923192519619302,
+      "grad_norm": 0.369216400903443,
+      "learning_rate": 3.0594605421765406e-06,
+      "loss": 0.1797,
+      "step": 5966
+    },
+    {
+      "epoch": 1.9926531975288029,
+      "grad_norm": 0.4121903275320649,
+      "learning_rate": 3.057669685699037e-06,
+      "loss": 0.1863,
+      "step": 5967
+    },
+    {
+      "epoch": 1.9929871430956754,
+      "grad_norm": 0.4355489715595582,
+      "learning_rate": 3.0558791226333974e-06,
+      "loss": 0.1911,
+      "step": 5968
+    },
+    {
+      "epoch": 1.9933210886625479,
+      "grad_norm": 0.3789799913064823,
+      "learning_rate": 3.0540888532501075e-06,
+      "loss": 0.1926,
+      "step": 5969
+    },
+    {
+      "epoch": 1.9936550342294206,
+      "grad_norm": 0.42365316429605293,
+      "learning_rate": 3.052298877819608e-06,
+      "loss": 0.1904,
+      "step": 5970
+    },
+    {
+      "epoch": 1.9939889797962933,
+      "grad_norm": 0.41698808702160367,
+      "learning_rate": 3.050509196612297e-06,
+      "loss": 0.1963,
+      "step": 5971
+    },
+    {
+      "epoch": 1.9943229253631658,
+      "grad_norm": 0.37543262198026217,
+      "learning_rate": 3.0487198098985265e-06,
+      "loss": 0.1944,
+      "step": 5972
+    },
+    {
+      "epoch": 1.9946568709300383,
+      "grad_norm": 0.43641995145584184,
+      "learning_rate": 3.046930717948604e-06,
+      "loss": 0.1892,
+      "step": 5973
+    },
+    {
+      "epoch": 1.994990816496911,
+      "grad_norm": 0.38182184018203114,
+      "learning_rate": 3.0451419210327935e-06,
+      "loss": 0.1806,
+      "step": 5974
+    },
+    {
+      "epoch": 1.9953247620637837,
+      "grad_norm": 0.3818787684355753,
+      "learning_rate": 3.0433534194213143e-06,
+      "loss": 0.1799,
+      "step": 5975
+    },
+    {
+      "epoch": 1.9956587076306562,
+      "grad_norm": 0.3744181477267095,
+      "learning_rate": 3.0415652133843375e-06,
+      "loss": 0.1816,
+      "step": 5976
+    },
+    {
+      "epoch": 1.9959926531975287,
+      "grad_norm": 0.40734828799825423,
+      "learning_rate": 3.0397773031919966e-06,
+      "loss": 0.2016,
+      "step": 5977
+    },
+    {
+      "epoch": 1.9963265987644014,
+      "grad_norm": 0.3871428637806214,
+      "learning_rate": 3.0379896891143746e-06,
+      "loss": 0.1806,
+      "step": 5978
+    },
+    {
+      "epoch": 1.9966605443312742,
+      "grad_norm": 0.39395345656310504,
+      "learning_rate": 3.036202371421513e-06,
+      "loss": 0.1812,
+      "step": 5979
+    },
+    {
+      "epoch": 1.9969944898981467,
+      "grad_norm": 0.41099460235481833,
+      "learning_rate": 3.034415350383405e-06,
+      "loss": 0.1939,
+      "step": 5980
+    },
+    {
+      "epoch": 1.9973284354650191,
+      "grad_norm": 0.3679936586206283,
+      "learning_rate": 3.0326286262700035e-06,
+      "loss": 0.1879,
+      "step": 5981
+    },
+    {
+      "epoch": 1.9976623810318919,
+      "grad_norm": 0.3814246949048082,
+      "learning_rate": 3.030842199351212e-06,
+      "loss": 0.1879,
+      "step": 5982
+    },
+    {
+      "epoch": 1.9979963265987644,
+      "grad_norm": 0.42714571464627366,
+      "learning_rate": 3.0290560698968907e-06,
+      "loss": 0.2029,
+      "step": 5983
+    },
+    {
+      "epoch": 1.9983302721656369,
+      "grad_norm": 0.4021653615871575,
+      "learning_rate": 3.0272702381768593e-06,
+      "loss": 0.1936,
+      "step": 5984
+    },
+    {
+      "epoch": 1.9986642177325096,
+      "grad_norm": 0.39922967076189847,
+      "learning_rate": 3.0254847044608872e-06,
+      "loss": 0.2034,
+      "step": 5985
+    },
+    {
+      "epoch": 1.9989981632993823,
+      "grad_norm": 0.4103626909173217,
+      "learning_rate": 3.0236994690186983e-06,
+      "loss": 0.1874,
+      "step": 5986
+    },
+    {
+      "epoch": 1.9993321088662548,
+      "grad_norm": 0.40222096398131774,
+      "learning_rate": 3.0219145321199763e-06,
+      "loss": 0.1868,
+      "step": 5987
+    },
+    {
+      "epoch": 1.9996660544331273,
+      "grad_norm": 0.37798334008408047,
+      "learning_rate": 3.0201298940343543e-06,
+      "loss": 0.1801,
+      "step": 5988
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.41315481474910304,
+      "learning_rate": 3.018345555031422e-06,
+      "loss": 0.1861,
+      "step": 5989
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.20493414998054504,
+      "eval_runtime": 184.3833,
+      "eval_samples_per_second": 109.408,
+      "eval_steps_per_second": 1.714,
+      "step": 5989
+    },
+    {
+      "epoch": 2.0003339455668727,
+      "grad_norm": 0.3534058618254128,
+      "learning_rate": 3.0165615153807293e-06,
+      "loss": 0.155,
+      "step": 5990
+    },
+    {
+      "epoch": 2.000667891133745,
+      "grad_norm": 0.4330126571772758,
+      "learning_rate": 3.014777775351774e-06,
+      "loss": 0.1767,
+      "step": 5991
+    },
+    {
+      "epoch": 2.0010018367006177,
+      "grad_norm": 0.3993798294281268,
+      "learning_rate": 3.012994335214011e-06,
+      "loss": 0.1656,
+      "step": 5992
+    },
+    {
+      "epoch": 2.0013357822674904,
+      "grad_norm": 0.35823939570937335,
+      "learning_rate": 3.0112111952368496e-06,
+      "loss": 0.152,
+      "step": 5993
+    },
+    {
+      "epoch": 2.001669727834363,
+      "grad_norm": 0.4274610122348606,
+      "learning_rate": 3.009428355689654e-06,
+      "loss": 0.1683,
+      "step": 5994
+    },
+    {
+      "epoch": 2.0020036734012354,
+      "grad_norm": 0.39209988341414004,
+      "learning_rate": 3.007645816841743e-06,
+      "loss": 0.1553,
+      "step": 5995
+    },
+    {
+      "epoch": 2.002337618968108,
+      "grad_norm": 0.41360416403789046,
+      "learning_rate": 3.0058635789623926e-06,
+      "loss": 0.1693,
+      "step": 5996
+    },
+    {
+      "epoch": 2.002671564534981,
+      "grad_norm": 0.3964326471268367,
+      "learning_rate": 3.0040816423208276e-06,
+      "loss": 0.157,
+      "step": 5997
+    },
+    {
+      "epoch": 2.0030055101018536,
+      "grad_norm": 0.38668198118986086,
+      "learning_rate": 3.002300007186232e-06,
+      "loss": 0.1524,
+      "step": 5998
+    },
+    {
+      "epoch": 2.003339455668726,
+      "grad_norm": 0.405645756289619,
+      "learning_rate": 3.0005186738277407e-06,
+      "loss": 0.1656,
+      "step": 5999
+    },
+    {
+      "epoch": 2.0036734012355986,
+      "grad_norm": 0.4397302560149037,
+      "learning_rate": 2.9987376425144477e-06,
+      "loss": 0.1577,
+      "step": 6000
+    },
+    {
+      "epoch": 2.0040073468024713,
+      "grad_norm": 0.4394388666470143,
+      "learning_rate": 2.9969569135153985e-06,
+      "loss": 0.1639,
+      "step": 6001
+    },
+    {
+      "epoch": 2.004341292369344,
+      "grad_norm": 0.4849570632205489,
+      "learning_rate": 2.9951764870995925e-06,
+      "loss": 0.1621,
+      "step": 6002
+    },
+    {
+      "epoch": 2.0046752379362163,
+      "grad_norm": 0.48705098971685307,
+      "learning_rate": 2.9933963635359847e-06,
+      "loss": 0.1646,
+      "step": 6003
+    },
+    {
+      "epoch": 2.005009183503089,
+      "grad_norm": 0.4641189129482545,
+      "learning_rate": 2.991616543093483e-06,
+      "loss": 0.1749,
+      "step": 6004
+    },
+    {
+      "epoch": 2.0053431290699617,
+      "grad_norm": 0.46677908584428857,
+      "learning_rate": 2.9898370260409502e-06,
+      "loss": 0.1623,
+      "step": 6005
+    },
+    {
+      "epoch": 2.005677074636834,
+      "grad_norm": 0.4711820568033053,
+      "learning_rate": 2.9880578126472015e-06,
+      "loss": 0.16,
+      "step": 6006
+    },
+    {
+      "epoch": 2.0060110202037067,
+      "grad_norm": 0.47617074156351524,
+      "learning_rate": 2.9862789031810126e-06,
+      "loss": 0.1721,
+      "step": 6007
+    },
+    {
+      "epoch": 2.0063449657705794,
+      "grad_norm": 0.4487052591604857,
+      "learning_rate": 2.984500297911106e-06,
+      "loss": 0.1584,
+      "step": 6008
+    },
+    {
+      "epoch": 2.006678911337452,
+      "grad_norm": 0.447671304592757,
+      "learning_rate": 2.9827219971061607e-06,
+      "loss": 0.1654,
+      "step": 6009
+    },
+    {
+      "epoch": 2.0070128569043244,
+      "grad_norm": 0.47125672755220205,
+      "learning_rate": 2.98094400103481e-06,
+      "loss": 0.1681,
+      "step": 6010
+    },
+    {
+      "epoch": 2.007346802471197,
+      "grad_norm": 0.45728456238338155,
+      "learning_rate": 2.9791663099656424e-06,
+      "loss": 0.1618,
+      "step": 6011
+    },
+    {
+      "epoch": 2.00768074803807,
+      "grad_norm": 0.47447060864994905,
+      "learning_rate": 2.977388924167196e-06,
+      "loss": 0.1492,
+      "step": 6012
+    },
+    {
+      "epoch": 2.0080146936049426,
+      "grad_norm": 0.446422979462353,
+      "learning_rate": 2.975611843907971e-06,
+      "loss": 0.1576,
+      "step": 6013
+    },
+    {
+      "epoch": 2.008348639171815,
+      "grad_norm": 0.48529971727386545,
+      "learning_rate": 2.9738350694564117e-06,
+      "loss": 0.1674,
+      "step": 6014
+    },
+    {
+      "epoch": 2.0086825847386875,
+      "grad_norm": 0.43999105968259355,
+      "learning_rate": 2.9720586010809234e-06,
+      "loss": 0.1584,
+      "step": 6015
+    },
+    {
+      "epoch": 2.0090165303055603,
+      "grad_norm": 0.42356091106399757,
+      "learning_rate": 2.9702824390498615e-06,
+      "loss": 0.155,
+      "step": 6016
+    },
+    {
+      "epoch": 2.009350475872433,
+      "grad_norm": 0.4529826679339405,
+      "learning_rate": 2.9685065836315362e-06,
+      "loss": 0.1594,
+      "step": 6017
+    },
+    {
+      "epoch": 2.0096844214393053,
+      "grad_norm": 0.44293372978393375,
+      "learning_rate": 2.9667310350942103e-06,
+      "loss": 0.1601,
+      "step": 6018
+    },
+    {
+      "epoch": 2.010018367006178,
+      "grad_norm": 0.4662546772405464,
+      "learning_rate": 2.964955793706104e-06,
+      "loss": 0.1651,
+      "step": 6019
+    },
+    {
+      "epoch": 2.0103523125730507,
+      "grad_norm": 0.4662905596333431,
+      "learning_rate": 2.963180859735387e-06,
+      "loss": 0.1564,
+      "step": 6020
+    },
+    {
+      "epoch": 2.0106862581399234,
+      "grad_norm": 0.5287577603093535,
+      "learning_rate": 2.961406233450184e-06,
+      "loss": 0.1658,
+      "step": 6021
+    },
+    {
+      "epoch": 2.0110202037067957,
+      "grad_norm": 0.5051506571548454,
+      "learning_rate": 2.9596319151185713e-06,
+      "loss": 0.178,
+      "step": 6022
+    },
+    {
+      "epoch": 2.0113541492736684,
+      "grad_norm": 0.4822552977434536,
+      "learning_rate": 2.9578579050085836e-06,
+      "loss": 0.1756,
+      "step": 6023
+    },
+    {
+      "epoch": 2.011688094840541,
+      "grad_norm": 0.5070476190943671,
+      "learning_rate": 2.956084203388204e-06,
+      "loss": 0.1565,
+      "step": 6024
+    },
+    {
+      "epoch": 2.0120220404074134,
+      "grad_norm": 0.4697832208358172,
+      "learning_rate": 2.9543108105253733e-06,
+      "loss": 0.1665,
+      "step": 6025
+    },
+    {
+      "epoch": 2.012355985974286,
+      "grad_norm": 0.47328140753003795,
+      "learning_rate": 2.9525377266879813e-06,
+      "loss": 0.1646,
+      "step": 6026
+    },
+    {
+      "epoch": 2.012689931541159,
+      "grad_norm": 0.45988128423824537,
+      "learning_rate": 2.950764952143874e-06,
+      "loss": 0.1568,
+      "step": 6027
+    },
+    {
+      "epoch": 2.0130238771080315,
+      "grad_norm": 0.4671425277097255,
+      "learning_rate": 2.9489924871608495e-06,
+      "loss": 0.1643,
+      "step": 6028
+    },
+    {
+      "epoch": 2.013357822674904,
+      "grad_norm": 0.4991890618376096,
+      "learning_rate": 2.9472203320066594e-06,
+      "loss": 0.1721,
+      "step": 6029
+    },
+    {
+      "epoch": 2.0136917682417765,
+      "grad_norm": 0.4946814103252608,
+      "learning_rate": 2.9454484869490074e-06,
+      "loss": 0.1696,
+      "step": 6030
+    },
+    {
+      "epoch": 2.0140257138086493,
+      "grad_norm": 0.5118310832695881,
+      "learning_rate": 2.943676952255554e-06,
+      "loss": 0.1657,
+      "step": 6031
+    },
+    {
+      "epoch": 2.014359659375522,
+      "grad_norm": 0.49387059069204864,
+      "learning_rate": 2.9419057281939106e-06,
+      "loss": 0.168,
+      "step": 6032
+    },
+    {
+      "epoch": 2.0146936049423942,
+      "grad_norm": 0.4596647029669583,
+      "learning_rate": 2.94013481503164e-06,
+      "loss": 0.1597,
+      "step": 6033
+    },
+    {
+      "epoch": 2.015027550509267,
+      "grad_norm": 0.49980994832903475,
+      "learning_rate": 2.9383642130362596e-06,
+      "loss": 0.1647,
+      "step": 6034
+    },
+    {
+      "epoch": 2.0153614960761397,
+      "grad_norm": 0.4739084813661146,
+      "learning_rate": 2.9365939224752394e-06,
+      "loss": 0.1564,
+      "step": 6035
+    },
+    {
+      "epoch": 2.0156954416430124,
+      "grad_norm": 0.48889393560437605,
+      "learning_rate": 2.934823943616001e-06,
+      "loss": 0.1683,
+      "step": 6036
+    },
+    {
+      "epoch": 2.0160293872098847,
+      "grad_norm": 0.46643889067317856,
+      "learning_rate": 2.933054276725925e-06,
+      "loss": 0.1613,
+      "step": 6037
+    },
+    {
+      "epoch": 2.0163633327767574,
+      "grad_norm": 0.45399739753399115,
+      "learning_rate": 2.9312849220723382e-06,
+      "loss": 0.1559,
+      "step": 6038
+    },
+    {
+      "epoch": 2.01669727834363,
+      "grad_norm": 0.45507324551278094,
+      "learning_rate": 2.929515879922522e-06,
+      "loss": 0.159,
+      "step": 6039
+    },
+    {
+      "epoch": 2.0170312239105024,
+      "grad_norm": 0.4612841631939222,
+      "learning_rate": 2.9277471505437105e-06,
+      "loss": 0.1602,
+      "step": 6040
+    },
+    {
+      "epoch": 2.017365169477375,
+      "grad_norm": 0.4776368978691176,
+      "learning_rate": 2.925978734203092e-06,
+      "loss": 0.1723,
+      "step": 6041
+    },
+    {
+      "epoch": 2.017699115044248,
+      "grad_norm": 0.47341321323873353,
+      "learning_rate": 2.924210631167807e-06,
+      "loss": 0.1518,
+      "step": 6042
+    },
+    {
+      "epoch": 2.0180330606111205,
+      "grad_norm": 0.4839747955511077,
+      "learning_rate": 2.922442841704948e-06,
+      "loss": 0.1621,
+      "step": 6043
+    },
+    {
+      "epoch": 2.018367006177993,
+      "grad_norm": 0.4883511258492533,
+      "learning_rate": 2.920675366081559e-06,
+      "loss": 0.1636,
+      "step": 6044
+    },
+    {
+      "epoch": 2.0187009517448655,
+      "grad_norm": 0.5464115849984775,
+      "learning_rate": 2.9189082045646404e-06,
+      "loss": 0.1777,
+      "step": 6045
+    },
+    {
+      "epoch": 2.0190348973117382,
+      "grad_norm": 0.46191595447785594,
+      "learning_rate": 2.9171413574211426e-06,
+      "loss": 0.1623,
+      "step": 6046
+    },
+    {
+      "epoch": 2.019368842878611,
+      "grad_norm": 0.49795505123926764,
+      "learning_rate": 2.9153748249179637e-06,
+      "loss": 0.1667,
+      "step": 6047
+    },
+    {
+      "epoch": 2.0197027884454832,
+      "grad_norm": 0.47251837346553677,
+      "learning_rate": 2.9136086073219665e-06,
+      "loss": 0.1539,
+      "step": 6048
+    },
+    {
+      "epoch": 2.020036734012356,
+      "grad_norm": 0.4581855742704899,
+      "learning_rate": 2.9118427048999544e-06,
+      "loss": 0.1555,
+      "step": 6049
+    },
+    {
+      "epoch": 2.0203706795792287,
+      "grad_norm": 0.5045197369778212,
+      "learning_rate": 2.9100771179186904e-06,
+      "loss": 0.17,
+      "step": 6050
+    },
+    {
+      "epoch": 2.0207046251461014,
+      "grad_norm": 0.4421223167410785,
+      "learning_rate": 2.9083118466448845e-06,
+      "loss": 0.1502,
+      "step": 6051
+    },
+    {
+      "epoch": 2.0210385707129737,
+      "grad_norm": 0.4474440028569155,
+      "learning_rate": 2.9065468913452045e-06,
+      "loss": 0.1462,
+      "step": 6052
+    },
+    {
+      "epoch": 2.0213725162798464,
+      "grad_norm": 0.4925003177758385,
+      "learning_rate": 2.904782252286264e-06,
+      "loss": 0.1618,
+      "step": 6053
+    },
+    {
+      "epoch": 2.021706461846719,
+      "grad_norm": 0.4821571393316654,
+      "learning_rate": 2.903017929734635e-06,
+      "loss": 0.1662,
+      "step": 6054
+    },
+    {
+      "epoch": 2.0220404074135914,
+      "grad_norm": 0.49103538206612174,
+      "learning_rate": 2.9012539239568405e-06,
+      "loss": 0.1633,
+      "step": 6055
+    },
+    {
+      "epoch": 2.022374352980464,
+      "grad_norm": 0.4891641946599009,
+      "learning_rate": 2.899490235219351e-06,
+      "loss": 0.1706,
+      "step": 6056
+    },
+    {
+      "epoch": 2.022708298547337,
+      "grad_norm": 0.4685788784764766,
+      "learning_rate": 2.897726863788595e-06,
+      "loss": 0.1602,
+      "step": 6057
+    },
+    {
+      "epoch": 2.0230422441142095,
+      "grad_norm": 0.4753894335902614,
+      "learning_rate": 2.8959638099309504e-06,
+      "loss": 0.1542,
+      "step": 6058
+    },
+    {
+      "epoch": 2.023376189681082,
+      "grad_norm": 0.4919778734814164,
+      "learning_rate": 2.8942010739127446e-06,
+      "loss": 0.1572,
+      "step": 6059
+    },
+    {
+      "epoch": 2.0237101352479545,
+      "grad_norm": 0.46606003547768277,
+      "learning_rate": 2.8924386560002627e-06,
+      "loss": 0.1574,
+      "step": 6060
+    },
+    {
+      "epoch": 2.0240440808148272,
+      "grad_norm": 0.48964811026319693,
+      "learning_rate": 2.8906765564597384e-06,
+      "loss": 0.1502,
+      "step": 6061
+    },
+    {
+      "epoch": 2.0243780263817,
+      "grad_norm": 0.48273759736055805,
+      "learning_rate": 2.8889147755573556e-06,
+      "loss": 0.1546,
+      "step": 6062
+    },
+    {
+      "epoch": 2.024711971948572,
+      "grad_norm": 0.5370134887436172,
+      "learning_rate": 2.8871533135592544e-06,
+      "loss": 0.1724,
+      "step": 6063
+    },
+    {
+      "epoch": 2.025045917515445,
+      "grad_norm": 0.45586288119980345,
+      "learning_rate": 2.8853921707315215e-06,
+      "loss": 0.1602,
+      "step": 6064
+    },
+    {
+      "epoch": 2.0253798630823177,
+      "grad_norm": 0.49510158520645964,
+      "learning_rate": 2.8836313473402e-06,
+      "loss": 0.1659,
+      "step": 6065
+    },
+    {
+      "epoch": 2.0257138086491904,
+      "grad_norm": 0.4668970296054312,
+      "learning_rate": 2.881870843651282e-06,
+      "loss": 0.1579,
+      "step": 6066
+    },
+    {
+      "epoch": 2.0260477542160626,
+      "grad_norm": 0.4694378333177316,
+      "learning_rate": 2.8801106599307164e-06,
+      "loss": 0.1492,
+      "step": 6067
+    },
+    {
+      "epoch": 2.0263816997829354,
+      "grad_norm": 0.4914932697805478,
+      "learning_rate": 2.8783507964443942e-06,
+      "loss": 0.1627,
+      "step": 6068
+    },
+    {
+      "epoch": 2.026715645349808,
+      "grad_norm": 0.4815900522040312,
+      "learning_rate": 2.8765912534581674e-06,
+      "loss": 0.1626,
+      "step": 6069
+    },
+    {
+      "epoch": 2.027049590916681,
+      "grad_norm": 0.48576180596656354,
+      "learning_rate": 2.874832031237833e-06,
+      "loss": 0.1565,
+      "step": 6070
+    },
+    {
+      "epoch": 2.027383536483553,
+      "grad_norm": 0.5102513368484094,
+      "learning_rate": 2.873073130049142e-06,
+      "loss": 0.1637,
+      "step": 6071
+    },
+    {
+      "epoch": 2.027717482050426,
+      "grad_norm": 0.4971694308223245,
+      "learning_rate": 2.8713145501578e-06,
+      "loss": 0.1609,
+      "step": 6072
+    },
+    {
+      "epoch": 2.0280514276172985,
+      "grad_norm": 0.5074347111244822,
+      "learning_rate": 2.869556291829461e-06,
+      "loss": 0.1503,
+      "step": 6073
+    },
+    {
+      "epoch": 2.028385373184171,
+      "grad_norm": 0.4216789477572703,
+      "learning_rate": 2.8677983553297266e-06,
+      "loss": 0.1452,
+      "step": 6074
+    },
+    {
+      "epoch": 2.0287193187510435,
+      "grad_norm": 0.49704788792406124,
+      "learning_rate": 2.8660407409241593e-06,
+      "loss": 0.1625,
+      "step": 6075
+    },
+    {
+      "epoch": 2.029053264317916,
+      "grad_norm": 0.49175915805051923,
+      "learning_rate": 2.864283448878262e-06,
+      "loss": 0.1694,
+      "step": 6076
+    },
+    {
+      "epoch": 2.029387209884789,
+      "grad_norm": 0.46571201952476754,
+      "learning_rate": 2.8625264794574975e-06,
+      "loss": 0.1572,
+      "step": 6077
+    },
+    {
+      "epoch": 2.029721155451661,
+      "grad_norm": 0.500277066125853,
+      "learning_rate": 2.860769832927276e-06,
+      "loss": 0.1662,
+      "step": 6078
+    },
+    {
+      "epoch": 2.030055101018534,
+      "grad_norm": 0.49439200808987693,
+      "learning_rate": 2.8590135095529624e-06,
+      "loss": 0.1676,
+      "step": 6079
+    },
+    {
+      "epoch": 2.0303890465854066,
+      "grad_norm": 0.44751887545978086,
+      "learning_rate": 2.8572575095998646e-06,
+      "loss": 0.1479,
+      "step": 6080
+    },
+    {
+      "epoch": 2.0307229921522794,
+      "grad_norm": 0.4366187834049901,
+      "learning_rate": 2.855501833333253e-06,
+      "loss": 0.1499,
+      "step": 6081
+    },
+    {
+      "epoch": 2.0310569377191516,
+      "grad_norm": 0.4463655726940175,
+      "learning_rate": 2.853746481018337e-06,
+      "loss": 0.1478,
+      "step": 6082
+    },
+    {
+      "epoch": 2.0313908832860244,
+      "grad_norm": 0.525002855673986,
+      "learning_rate": 2.8519914529202868e-06,
+      "loss": 0.1768,
+      "step": 6083
+    },
+    {
+      "epoch": 2.031724828852897,
+      "grad_norm": 0.46830169747916234,
+      "learning_rate": 2.8502367493042217e-06,
+      "loss": 0.1652,
+      "step": 6084
+    },
+    {
+      "epoch": 2.03205877441977,
+      "grad_norm": 0.474430022851497,
+      "learning_rate": 2.848482370435206e-06,
+      "loss": 0.1601,
+      "step": 6085
+    },
+    {
+      "epoch": 2.032392719986642,
+      "grad_norm": 0.4858834048312866,
+      "learning_rate": 2.8467283165782643e-06,
+      "loss": 0.1637,
+      "step": 6086
+    },
+    {
+      "epoch": 2.0327266655535148,
+      "grad_norm": 0.5000627831706353,
+      "learning_rate": 2.8449745879983614e-06,
+      "loss": 0.1562,
+      "step": 6087
+    },
+    {
+      "epoch": 2.0330606111203875,
+      "grad_norm": 0.48498567163926287,
+      "learning_rate": 2.8432211849604218e-06,
+      "loss": 0.1594,
+      "step": 6088
+    },
+    {
+      "epoch": 2.0333945566872598,
+      "grad_norm": 0.4706989935254033,
+      "learning_rate": 2.841468107729318e-06,
+      "loss": 0.1627,
+      "step": 6089
+    },
+    {
+      "epoch": 2.0337285022541325,
+      "grad_norm": 0.4897839906916831,
+      "learning_rate": 2.8397153565698744e-06,
+      "loss": 0.1567,
+      "step": 6090
+    },
+    {
+      "epoch": 2.034062447821005,
+      "grad_norm": 0.46017908487227666,
+      "learning_rate": 2.8379629317468604e-06,
+      "loss": 0.1588,
+      "step": 6091
+    },
+    {
+      "epoch": 2.034396393387878,
+      "grad_norm": 0.4692319755849913,
+      "learning_rate": 2.8362108335250044e-06,
+      "loss": 0.1522,
+      "step": 6092
+    },
+    {
+      "epoch": 2.03473033895475,
+      "grad_norm": 0.520679375837806,
+      "learning_rate": 2.834459062168978e-06,
+      "loss": 0.1718,
+      "step": 6093
+    },
+    {
+      "epoch": 2.035064284521623,
+      "grad_norm": 0.5246005906068195,
+      "learning_rate": 2.8327076179434088e-06,
+      "loss": 0.1669,
+      "step": 6094
+    },
+    {
+      "epoch": 2.0353982300884956,
+      "grad_norm": 0.45873793348961106,
+      "learning_rate": 2.8309565011128732e-06,
+      "loss": 0.1488,
+      "step": 6095
+    },
+    {
+      "epoch": 2.0357321756553683,
+      "grad_norm": 0.47134127061918596,
+      "learning_rate": 2.8292057119418994e-06,
+      "loss": 0.1631,
+      "step": 6096
+    },
+    {
+      "epoch": 2.0360661212222406,
+      "grad_norm": 0.4729789859407233,
+      "learning_rate": 2.827455250694961e-06,
+      "loss": 0.16,
+      "step": 6097
+    },
+    {
+      "epoch": 2.0364000667891133,
+      "grad_norm": 0.5146567715009788,
+      "learning_rate": 2.8257051176364903e-06,
+      "loss": 0.1588,
+      "step": 6098
+    },
+    {
+      "epoch": 2.036734012355986,
+      "grad_norm": 0.4937190219735824,
+      "learning_rate": 2.8239553130308604e-06,
+      "loss": 0.1733,
+      "step": 6099
+    },
+    {
+      "epoch": 2.0370679579228588,
+      "grad_norm": 0.49973854857161226,
+      "learning_rate": 2.8222058371424033e-06,
+      "loss": 0.1609,
+      "step": 6100
+    },
+    {
+      "epoch": 2.037401903489731,
+      "grad_norm": 0.5022894018079591,
+      "learning_rate": 2.820456690235397e-06,
+      "loss": 0.1618,
+      "step": 6101
+    },
+    {
+      "epoch": 2.0377358490566038,
+      "grad_norm": 0.5307860067204769,
+      "learning_rate": 2.8187078725740723e-06,
+      "loss": 0.1717,
+      "step": 6102
+    },
+    {
+      "epoch": 2.0380697946234765,
+      "grad_norm": 0.5005034713170023,
+      "learning_rate": 2.8169593844226063e-06,
+      "loss": 0.1656,
+      "step": 6103
+    },
+    {
+      "epoch": 2.0384037401903488,
+      "grad_norm": 0.44826699148693205,
+      "learning_rate": 2.815211226045131e-06,
+      "loss": 0.1444,
+      "step": 6104
+    },
+    {
+      "epoch": 2.0387376857572215,
+      "grad_norm": 0.4616708746298766,
+      "learning_rate": 2.8134633977057236e-06,
+      "loss": 0.1543,
+      "step": 6105
+    },
+    {
+      "epoch": 2.039071631324094,
+      "grad_norm": 0.456410361967302,
+      "learning_rate": 2.811715899668415e-06,
+      "loss": 0.1539,
+      "step": 6106
+    },
+    {
+      "epoch": 2.039405576890967,
+      "grad_norm": 0.47621252955779786,
+      "learning_rate": 2.8099687321971887e-06,
+      "loss": 0.1556,
+      "step": 6107
+    },
+    {
+      "epoch": 2.039739522457839,
+      "grad_norm": 0.4543629974389159,
+      "learning_rate": 2.80822189555597e-06,
+      "loss": 0.1537,
+      "step": 6108
+    },
+    {
+      "epoch": 2.040073468024712,
+      "grad_norm": 0.49525294210493037,
+      "learning_rate": 2.8064753900086427e-06,
+      "loss": 0.1654,
+      "step": 6109
+    },
+    {
+      "epoch": 2.0404074135915846,
+      "grad_norm": 0.512395040030274,
+      "learning_rate": 2.804729215819034e-06,
+      "loss": 0.1712,
+      "step": 6110
+    },
+    {
+      "epoch": 2.0407413591584573,
+      "grad_norm": 0.47930074814671675,
+      "learning_rate": 2.8029833732509282e-06,
+      "loss": 0.1545,
+      "step": 6111
+    },
+    {
+      "epoch": 2.0410753047253296,
+      "grad_norm": 0.496750207975215,
+      "learning_rate": 2.801237862568048e-06,
+      "loss": 0.1538,
+      "step": 6112
+    },
+    {
+      "epoch": 2.0414092502922023,
+      "grad_norm": 0.5533662779888088,
+      "learning_rate": 2.799492684034083e-06,
+      "loss": 0.159,
+      "step": 6113
+    },
+    {
+      "epoch": 2.041743195859075,
+      "grad_norm": 0.5055832988689022,
+      "learning_rate": 2.797747837912656e-06,
+      "loss": 0.1607,
+      "step": 6114
+    },
+    {
+      "epoch": 2.0420771414259478,
+      "grad_norm": 0.48537757955911165,
+      "learning_rate": 2.796003324467351e-06,
+      "loss": 0.1572,
+      "step": 6115
+    },
+    {
+      "epoch": 2.04241108699282,
+      "grad_norm": 0.5055379856535059,
+      "learning_rate": 2.794259143961693e-06,
+      "loss": 0.1621,
+      "step": 6116
+    },
+    {
+      "epoch": 2.0427450325596928,
+      "grad_norm": 0.5309785977167635,
+      "learning_rate": 2.7925152966591627e-06,
+      "loss": 0.168,
+      "step": 6117
+    },
+    {
+      "epoch": 2.0430789781265655,
+      "grad_norm": 0.48438645896285526,
+      "learning_rate": 2.7907717828231893e-06,
+      "loss": 0.158,
+      "step": 6118
+    },
+    {
+      "epoch": 2.043412923693438,
+      "grad_norm": 0.4991961469549531,
+      "learning_rate": 2.7890286027171532e-06,
+      "loss": 0.1591,
+      "step": 6119
+    },
+    {
+      "epoch": 2.0437468692603105,
+      "grad_norm": 0.45423321499364905,
+      "learning_rate": 2.7872857566043775e-06,
+      "loss": 0.158,
+      "step": 6120
+    },
+    {
+      "epoch": 2.044080814827183,
+      "grad_norm": 0.4865693553830601,
+      "learning_rate": 2.7855432447481444e-06,
+      "loss": 0.1649,
+      "step": 6121
+    },
+    {
+      "epoch": 2.044414760394056,
+      "grad_norm": 0.5337114471576638,
+      "learning_rate": 2.7838010674116767e-06,
+      "loss": 0.1718,
+      "step": 6122
+    },
+    {
+      "epoch": 2.044748705960928,
+      "grad_norm": 0.44948290522077555,
+      "learning_rate": 2.7820592248581523e-06,
+      "loss": 0.1487,
+      "step": 6123
+    },
+    {
+      "epoch": 2.045082651527801,
+      "grad_norm": 0.4938115357940406,
+      "learning_rate": 2.780317717350697e-06,
+      "loss": 0.163,
+      "step": 6124
+    },
+    {
+      "epoch": 2.0454165970946736,
+      "grad_norm": 0.49731409157250733,
+      "learning_rate": 2.7785765451523896e-06,
+      "loss": 0.165,
+      "step": 6125
+    },
+    {
+      "epoch": 2.0457505426615463,
+      "grad_norm": 0.45450185170041324,
+      "learning_rate": 2.7768357085262486e-06,
+      "loss": 0.1488,
+      "step": 6126
+    },
+    {
+      "epoch": 2.0460844882284186,
+      "grad_norm": 0.45184670089086093,
+      "learning_rate": 2.7750952077352534e-06,
+      "loss": 0.1578,
+      "step": 6127
+    },
+    {
+      "epoch": 2.0464184337952913,
+      "grad_norm": 0.4948991525337002,
+      "learning_rate": 2.7733550430423216e-06,
+      "loss": 0.1658,
+      "step": 6128
+    },
+    {
+      "epoch": 2.046752379362164,
+      "grad_norm": 0.5040836204485063,
+      "learning_rate": 2.7716152147103292e-06,
+      "loss": 0.1577,
+      "step": 6129
+    },
+    {
+      "epoch": 2.0470863249290367,
+      "grad_norm": 0.5335238908571901,
+      "learning_rate": 2.7698757230020986e-06,
+      "loss": 0.1572,
+      "step": 6130
+    },
+    {
+      "epoch": 2.047420270495909,
+      "grad_norm": 0.49828751734413135,
+      "learning_rate": 2.7681365681803967e-06,
+      "loss": 0.1657,
+      "step": 6131
+    },
+    {
+      "epoch": 2.0477542160627817,
+      "grad_norm": 0.4999406939225242,
+      "learning_rate": 2.7663977505079483e-06,
+      "loss": 0.1687,
+      "step": 6132
+    },
+    {
+      "epoch": 2.0480881616296545,
+      "grad_norm": 0.5085730563774729,
+      "learning_rate": 2.764659270247417e-06,
+      "loss": 0.1711,
+      "step": 6133
+    },
+    {
+      "epoch": 2.048422107196527,
+      "grad_norm": 0.47688020503244877,
+      "learning_rate": 2.7629211276614255e-06,
+      "loss": 0.1589,
+      "step": 6134
+    },
+    {
+      "epoch": 2.0487560527633994,
+      "grad_norm": 0.47192663378129107,
+      "learning_rate": 2.761183323012534e-06,
+      "loss": 0.1613,
+      "step": 6135
+    },
+    {
+      "epoch": 2.049089998330272,
+      "grad_norm": 0.4848663154654449,
+      "learning_rate": 2.7594458565632664e-06,
+      "loss": 0.1578,
+      "step": 6136
+    },
+    {
+      "epoch": 2.049423943897145,
+      "grad_norm": 0.5000980624774899,
+      "learning_rate": 2.757708728576083e-06,
+      "loss": 0.1556,
+      "step": 6137
+    },
+    {
+      "epoch": 2.049757889464017,
+      "grad_norm": 0.47224990165416336,
+      "learning_rate": 2.7559719393133987e-06,
+      "loss": 0.1538,
+      "step": 6138
+    },
+    {
+      "epoch": 2.05009183503089,
+      "grad_norm": 0.5215834737284637,
+      "learning_rate": 2.754235489037575e-06,
+      "loss": 0.1563,
+      "step": 6139
+    },
+    {
+      "epoch": 2.0504257805977626,
+      "grad_norm": 0.4972320969795835,
+      "learning_rate": 2.7524993780109254e-06,
+      "loss": 0.1491,
+      "step": 6140
+    },
+    {
+      "epoch": 2.0507597261646353,
+      "grad_norm": 0.49919640281659483,
+      "learning_rate": 2.750763606495704e-06,
+      "loss": 0.1581,
+      "step": 6141
+    },
+    {
+      "epoch": 2.0510936717315076,
+      "grad_norm": 0.5143857980365772,
+      "learning_rate": 2.7490281747541276e-06,
+      "loss": 0.1712,
+      "step": 6142
+    },
+    {
+      "epoch": 2.0514276172983803,
+      "grad_norm": 0.5088252050706469,
+      "learning_rate": 2.747293083048348e-06,
+      "loss": 0.1591,
+      "step": 6143
+    },
+    {
+      "epoch": 2.051761562865253,
+      "grad_norm": 0.5077613718070135,
+      "learning_rate": 2.7455583316404744e-06,
+      "loss": 0.1673,
+      "step": 6144
+    },
+    {
+      "epoch": 2.0520955084321257,
+      "grad_norm": 0.5097681301324266,
+      "learning_rate": 2.743823920792559e-06,
+      "loss": 0.1652,
+      "step": 6145
+    },
+    {
+      "epoch": 2.052429453998998,
+      "grad_norm": 0.5389169316301308,
+      "learning_rate": 2.742089850766607e-06,
+      "loss": 0.1742,
+      "step": 6146
+    },
+    {
+      "epoch": 2.0527633995658707,
+      "grad_norm": 0.48373136141051654,
+      "learning_rate": 2.7403561218245654e-06,
+      "loss": 0.1589,
+      "step": 6147
+    },
+    {
+      "epoch": 2.0530973451327434,
+      "grad_norm": 0.4750482606605971,
+      "learning_rate": 2.7386227342283423e-06,
+      "loss": 0.1573,
+      "step": 6148
+    },
+    {
+      "epoch": 2.053431290699616,
+      "grad_norm": 0.5005599752487294,
+      "learning_rate": 2.73688968823978e-06,
+      "loss": 0.1723,
+      "step": 6149
+    },
+    {
+      "epoch": 2.0537652362664884,
+      "grad_norm": 0.48098693092080774,
+      "learning_rate": 2.7351569841206792e-06,
+      "loss": 0.1565,
+      "step": 6150
+    },
+    {
+      "epoch": 2.054099181833361,
+      "grad_norm": 0.5329904578572746,
+      "learning_rate": 2.733424622132782e-06,
+      "loss": 0.1717,
+      "step": 6151
+    },
+    {
+      "epoch": 2.054433127400234,
+      "grad_norm": 0.4760378622271746,
+      "learning_rate": 2.7316926025377855e-06,
+      "loss": 0.1588,
+      "step": 6152
+    },
+    {
+      "epoch": 2.054767072967106,
+      "grad_norm": 0.4462805437410892,
+      "learning_rate": 2.729960925597328e-06,
+      "loss": 0.1476,
+      "step": 6153
+    },
+    {
+      "epoch": 2.055101018533979,
+      "grad_norm": 0.49345870296933825,
+      "learning_rate": 2.7282295915730016e-06,
+      "loss": 0.1642,
+      "step": 6154
+    },
+    {
+      "epoch": 2.0554349641008516,
+      "grad_norm": 0.48456295420893086,
+      "learning_rate": 2.726498600726346e-06,
+      "loss": 0.1632,
+      "step": 6155
+    },
+    {
+      "epoch": 2.0557689096677243,
+      "grad_norm": 0.45486223885521615,
+      "learning_rate": 2.7247679533188446e-06,
+      "loss": 0.1519,
+      "step": 6156
+    },
+    {
+      "epoch": 2.0561028552345966,
+      "grad_norm": 0.5200442659549483,
+      "learning_rate": 2.723037649611936e-06,
+      "loss": 0.1635,
+      "step": 6157
+    },
+    {
+      "epoch": 2.0564368008014693,
+      "grad_norm": 0.4909804430537013,
+      "learning_rate": 2.721307689866997e-06,
+      "loss": 0.1635,
+      "step": 6158
+    },
+    {
+      "epoch": 2.056770746368342,
+      "grad_norm": 0.4594697562211968,
+      "learning_rate": 2.719578074345366e-06,
+      "loss": 0.151,
+      "step": 6159
+    },
+    {
+      "epoch": 2.0571046919352147,
+      "grad_norm": 0.5065496098738638,
+      "learning_rate": 2.7178488033083163e-06,
+      "loss": 0.1652,
+      "step": 6160
+    },
+    {
+      "epoch": 2.057438637502087,
+      "grad_norm": 0.5436549815894377,
+      "learning_rate": 2.7161198770170784e-06,
+      "loss": 0.1702,
+      "step": 6161
+    },
+    {
+      "epoch": 2.0577725830689597,
+      "grad_norm": 0.49925937165886347,
+      "learning_rate": 2.714391295732822e-06,
+      "loss": 0.1647,
+      "step": 6162
+    },
+    {
+      "epoch": 2.0581065286358324,
+      "grad_norm": 0.4852644890591728,
+      "learning_rate": 2.712663059716675e-06,
+      "loss": 0.1458,
+      "step": 6163
+    },
+    {
+      "epoch": 2.058440474202705,
+      "grad_norm": 0.48613041699076365,
+      "learning_rate": 2.7109351692297015e-06,
+      "loss": 0.1657,
+      "step": 6164
+    },
+    {
+      "epoch": 2.0587744197695774,
+      "grad_norm": 0.4648594550935304,
+      "learning_rate": 2.7092076245329273e-06,
+      "loss": 0.1598,
+      "step": 6165
+    },
+    {
+      "epoch": 2.05910836533645,
+      "grad_norm": 0.4584401434296162,
+      "learning_rate": 2.7074804258873127e-06,
+      "loss": 0.1527,
+      "step": 6166
+    },
+    {
+      "epoch": 2.059442310903323,
+      "grad_norm": 0.48690854592041694,
+      "learning_rate": 2.7057535735537754e-06,
+      "loss": 0.1588,
+      "step": 6167
+    },
+    {
+      "epoch": 2.0597762564701956,
+      "grad_norm": 0.4726518812957776,
+      "learning_rate": 2.704027067793173e-06,
+      "loss": 0.1596,
+      "step": 6168
+    },
+    {
+      "epoch": 2.060110202037068,
+      "grad_norm": 0.4848631026115318,
+      "learning_rate": 2.7023009088663176e-06,
+      "loss": 0.1622,
+      "step": 6169
+    },
+    {
+      "epoch": 2.0604441476039406,
+      "grad_norm": 0.5010288195268352,
+      "learning_rate": 2.7005750970339607e-06,
+      "loss": 0.1493,
+      "step": 6170
+    },
+    {
+      "epoch": 2.0607780931708133,
+      "grad_norm": 0.5589263121127757,
+      "learning_rate": 2.698849632556815e-06,
+      "loss": 0.1801,
+      "step": 6171
+    },
+    {
+      "epoch": 2.0611120387376856,
+      "grad_norm": 0.4738720262047763,
+      "learning_rate": 2.697124515695524e-06,
+      "loss": 0.1539,
+      "step": 6172
+    },
+    {
+      "epoch": 2.0614459843045583,
+      "grad_norm": 0.4860672112480917,
+      "learning_rate": 2.695399746710693e-06,
+      "loss": 0.1598,
+      "step": 6173
+    },
+    {
+      "epoch": 2.061779929871431,
+      "grad_norm": 0.5056145335563673,
+      "learning_rate": 2.6936753258628643e-06,
+      "loss": 0.1633,
+      "step": 6174
+    },
+    {
+      "epoch": 2.0621138754383037,
+      "grad_norm": 0.526585125988662,
+      "learning_rate": 2.691951253412536e-06,
+      "loss": 0.1569,
+      "step": 6175
+    },
+    {
+      "epoch": 2.062447821005176,
+      "grad_norm": 0.5141605422386901,
+      "learning_rate": 2.6902275296201445e-06,
+      "loss": 0.1628,
+      "step": 6176
+    },
+    {
+      "epoch": 2.0627817665720487,
+      "grad_norm": 0.4813014518093507,
+      "learning_rate": 2.688504154746082e-06,
+      "loss": 0.1613,
+      "step": 6177
+    },
+    {
+      "epoch": 2.0631157121389214,
+      "grad_norm": 0.4515762636761851,
+      "learning_rate": 2.686781129050685e-06,
+      "loss": 0.1475,
+      "step": 6178
+    },
+    {
+      "epoch": 2.063449657705794,
+      "grad_norm": 0.4709359444645056,
+      "learning_rate": 2.685058452794235e-06,
+      "loss": 0.1603,
+      "step": 6179
+    },
+    {
+      "epoch": 2.0637836032726664,
+      "grad_norm": 0.4820755120947043,
+      "learning_rate": 2.6833361262369644e-06,
+      "loss": 0.1532,
+      "step": 6180
+    },
+    {
+      "epoch": 2.064117548839539,
+      "grad_norm": 0.45748798633211313,
+      "learning_rate": 2.681614149639048e-06,
+      "loss": 0.1603,
+      "step": 6181
+    },
+    {
+      "epoch": 2.064451494406412,
+      "grad_norm": 0.47928383265863794,
+      "learning_rate": 2.679892523260612e-06,
+      "loss": 0.1565,
+      "step": 6182
+    },
+    {
+      "epoch": 2.0647854399732846,
+      "grad_norm": 0.4912860713391019,
+      "learning_rate": 2.6781712473617293e-06,
+      "loss": 0.1631,
+      "step": 6183
+    },
+    {
+      "epoch": 2.065119385540157,
+      "grad_norm": 0.4937495874322415,
+      "learning_rate": 2.6764503222024202e-06,
+      "loss": 0.1631,
+      "step": 6184
+    },
+    {
+      "epoch": 2.0654533311070296,
+      "grad_norm": 0.5250388336695911,
+      "learning_rate": 2.674729748042647e-06,
+      "loss": 0.1712,
+      "step": 6185
+    },
+    {
+      "epoch": 2.0657872766739023,
+      "grad_norm": 0.47623079436105503,
+      "learning_rate": 2.673009525142326e-06,
+      "loss": 0.1564,
+      "step": 6186
+    },
+    {
+      "epoch": 2.0661212222407745,
+      "grad_norm": 0.4874432578932814,
+      "learning_rate": 2.6712896537613143e-06,
+      "loss": 0.1579,
+      "step": 6187
+    },
+    {
+      "epoch": 2.0664551678076473,
+      "grad_norm": 0.45987158697491903,
+      "learning_rate": 2.6695701341594193e-06,
+      "loss": 0.1505,
+      "step": 6188
+    },
+    {
+      "epoch": 2.06678911337452,
+      "grad_norm": 0.44396258874408284,
+      "learning_rate": 2.667850966596396e-06,
+      "loss": 0.1597,
+      "step": 6189
+    },
+    {
+      "epoch": 2.0671230589413927,
+      "grad_norm": 0.47997705349381775,
+      "learning_rate": 2.6661321513319467e-06,
+      "loss": 0.1575,
+      "step": 6190
+    },
+    {
+      "epoch": 2.067457004508265,
+      "grad_norm": 0.4639530958941431,
+      "learning_rate": 2.6644136886257138e-06,
+      "loss": 0.1554,
+      "step": 6191
+    },
+    {
+      "epoch": 2.0677909500751377,
+      "grad_norm": 0.47671499247486326,
+      "learning_rate": 2.6626955787372962e-06,
+      "loss": 0.1528,
+      "step": 6192
+    },
+    {
+      "epoch": 2.0681248956420104,
+      "grad_norm": 0.5172377476601987,
+      "learning_rate": 2.6609778219262296e-06,
+      "loss": 0.1572,
+      "step": 6193
+    },
+    {
+      "epoch": 2.068458841208883,
+      "grad_norm": 0.46150608992405706,
+      "learning_rate": 2.659260418452005e-06,
+      "loss": 0.1508,
+      "step": 6194
+    },
+    {
+      "epoch": 2.0687927867757554,
+      "grad_norm": 0.4864260006069014,
+      "learning_rate": 2.6575433685740547e-06,
+      "loss": 0.1503,
+      "step": 6195
+    },
+    {
+      "epoch": 2.069126732342628,
+      "grad_norm": 0.47462807507812255,
+      "learning_rate": 2.655826672551762e-06,
+      "loss": 0.1544,
+      "step": 6196
+    },
+    {
+      "epoch": 2.069460677909501,
+      "grad_norm": 0.5047881595787677,
+      "learning_rate": 2.6541103306444516e-06,
+      "loss": 0.1627,
+      "step": 6197
+    },
+    {
+      "epoch": 2.0697946234763736,
+      "grad_norm": 0.4847110255829372,
+      "learning_rate": 2.6523943431113985e-06,
+      "loss": 0.1575,
+      "step": 6198
+    },
+    {
+      "epoch": 2.070128569043246,
+      "grad_norm": 0.5684634058179412,
+      "learning_rate": 2.6506787102118204e-06,
+      "loss": 0.1793,
+      "step": 6199
+    },
+    {
+      "epoch": 2.0704625146101185,
+      "grad_norm": 0.4668023927238211,
+      "learning_rate": 2.6489634322048853e-06,
+      "loss": 0.1475,
+      "step": 6200
+    },
+    {
+      "epoch": 2.0707964601769913,
+      "grad_norm": 0.4823615249413857,
+      "learning_rate": 2.647248509349708e-06,
+      "loss": 0.1591,
+      "step": 6201
+    },
+    {
+      "epoch": 2.0711304057438635,
+      "grad_norm": 0.5284323022808656,
+      "learning_rate": 2.645533941905345e-06,
+      "loss": 0.1711,
+      "step": 6202
+    },
+    {
+      "epoch": 2.0714643513107363,
+      "grad_norm": 0.4629106304307049,
+      "learning_rate": 2.6438197301308045e-06,
+      "loss": 0.153,
+      "step": 6203
+    },
+    {
+      "epoch": 2.071798296877609,
+      "grad_norm": 0.4756702290381628,
+      "learning_rate": 2.6421058742850346e-06,
+      "loss": 0.1595,
+      "step": 6204
+    },
+    {
+      "epoch": 2.0721322424444817,
+      "grad_norm": 0.4651028891547631,
+      "learning_rate": 2.6403923746269368e-06,
+      "loss": 0.1439,
+      "step": 6205
+    },
+    {
+      "epoch": 2.072466188011354,
+      "grad_norm": 0.5038789264372514,
+      "learning_rate": 2.638679231415353e-06,
+      "loss": 0.1612,
+      "step": 6206
+    },
+    {
+      "epoch": 2.0728001335782267,
+      "grad_norm": 0.465142161854533,
+      "learning_rate": 2.636966444909077e-06,
+      "loss": 0.1581,
+      "step": 6207
+    },
+    {
+      "epoch": 2.0731340791450994,
+      "grad_norm": 0.471800296558469,
+      "learning_rate": 2.635254015366842e-06,
+      "loss": 0.1617,
+      "step": 6208
+    },
+    {
+      "epoch": 2.073468024711972,
+      "grad_norm": 0.4929302779325763,
+      "learning_rate": 2.633541943047334e-06,
+      "loss": 0.1617,
+      "step": 6209
+    },
+    {
+      "epoch": 2.0738019702788444,
+      "grad_norm": 0.5009578484887693,
+      "learning_rate": 2.6318302282091772e-06,
+      "loss": 0.1582,
+      "step": 6210
+    },
+    {
+      "epoch": 2.074135915845717,
+      "grad_norm": 0.4885347046833819,
+      "learning_rate": 2.6301188711109494e-06,
+      "loss": 0.1608,
+      "step": 6211
+    },
+    {
+      "epoch": 2.07446986141259,
+      "grad_norm": 0.46827738776381345,
+      "learning_rate": 2.6284078720111693e-06,
+      "loss": 0.1463,
+      "step": 6212
+    },
+    {
+      "epoch": 2.0748038069794625,
+      "grad_norm": 0.548235859732826,
+      "learning_rate": 2.626697231168308e-06,
+      "loss": 0.1672,
+      "step": 6213
+    },
+    {
+      "epoch": 2.075137752546335,
+      "grad_norm": 0.4807495227938576,
+      "learning_rate": 2.624986948840772e-06,
+      "loss": 0.1597,
+      "step": 6214
+    },
+    {
+      "epoch": 2.0754716981132075,
+      "grad_norm": 0.5208785862675631,
+      "learning_rate": 2.6232770252869243e-06,
+      "loss": 0.1652,
+      "step": 6215
+    },
+    {
+      "epoch": 2.0758056436800802,
+      "grad_norm": 0.5392394769474849,
+      "learning_rate": 2.6215674607650653e-06,
+      "loss": 0.1674,
+      "step": 6216
+    },
+    {
+      "epoch": 2.076139589246953,
+      "grad_norm": 0.4812349060192288,
+      "learning_rate": 2.619858255533446e-06,
+      "loss": 0.1555,
+      "step": 6217
+    },
+    {
+      "epoch": 2.0764735348138252,
+      "grad_norm": 0.4772963896311151,
+      "learning_rate": 2.6181494098502626e-06,
+      "loss": 0.1508,
+      "step": 6218
+    },
+    {
+      "epoch": 2.076807480380698,
+      "grad_norm": 0.4820130088588004,
+      "learning_rate": 2.616440923973659e-06,
+      "loss": 0.1579,
+      "step": 6219
+    },
+    {
+      "epoch": 2.0771414259475707,
+      "grad_norm": 0.47326464893761655,
+      "learning_rate": 2.6147327981617167e-06,
+      "loss": 0.156,
+      "step": 6220
+    },
+    {
+      "epoch": 2.077475371514443,
+      "grad_norm": 0.4814515104395897,
+      "learning_rate": 2.613025032672472e-06,
+      "loss": 0.1565,
+      "step": 6221
+    },
+    {
+      "epoch": 2.0778093170813157,
+      "grad_norm": 0.484941581282741,
+      "learning_rate": 2.611317627763901e-06,
+      "loss": 0.1559,
+      "step": 6222
+    },
+    {
+      "epoch": 2.0781432626481884,
+      "grad_norm": 0.46578841936370874,
+      "learning_rate": 2.609610583693928e-06,
+      "loss": 0.1565,
+      "step": 6223
+    },
+    {
+      "epoch": 2.078477208215061,
+      "grad_norm": 0.5122244564079962,
+      "learning_rate": 2.6079039007204238e-06,
+      "loss": 0.1628,
+      "step": 6224
+    },
+    {
+      "epoch": 2.0788111537819334,
+      "grad_norm": 0.4916632187074603,
+      "learning_rate": 2.6061975791011996e-06,
+      "loss": 0.1529,
+      "step": 6225
+    },
+    {
+      "epoch": 2.079145099348806,
+      "grad_norm": 0.4816831224754778,
+      "learning_rate": 2.6044916190940194e-06,
+      "loss": 0.1592,
+      "step": 6226
+    },
+    {
+      "epoch": 2.079479044915679,
+      "grad_norm": 0.4506436759636466,
+      "learning_rate": 2.6027860209565835e-06,
+      "loss": 0.145,
+      "step": 6227
+    },
+    {
+      "epoch": 2.0798129904825515,
+      "grad_norm": 0.5060128940585642,
+      "learning_rate": 2.6010807849465468e-06,
+      "loss": 0.1579,
+      "step": 6228
+    },
+    {
+      "epoch": 2.080146936049424,
+      "grad_norm": 0.5398542683218175,
+      "learning_rate": 2.5993759113215032e-06,
+      "loss": 0.178,
+      "step": 6229
+    },
+    {
+      "epoch": 2.0804808816162965,
+      "grad_norm": 0.46194809764874106,
+      "learning_rate": 2.5976714003389963e-06,
+      "loss": 0.1558,
+      "step": 6230
+    },
+    {
+      "epoch": 2.0808148271831692,
+      "grad_norm": 0.5504935947232371,
+      "learning_rate": 2.5959672522565095e-06,
+      "loss": 0.1697,
+      "step": 6231
+    },
+    {
+      "epoch": 2.081148772750042,
+      "grad_norm": 0.5461299363549844,
+      "learning_rate": 2.594263467331477e-06,
+      "loss": 0.1653,
+      "step": 6232
+    },
+    {
+      "epoch": 2.0814827183169142,
+      "grad_norm": 0.5349505003346677,
+      "learning_rate": 2.592560045821273e-06,
+      "loss": 0.1631,
+      "step": 6233
+    },
+    {
+      "epoch": 2.081816663883787,
+      "grad_norm": 0.5028722732399603,
+      "learning_rate": 2.5908569879832223e-06,
+      "loss": 0.1587,
+      "step": 6234
+    },
+    {
+      "epoch": 2.0821506094506597,
+      "grad_norm": 0.4847488245991641,
+      "learning_rate": 2.5891542940745873e-06,
+      "loss": 0.1645,
+      "step": 6235
+    },
+    {
+      "epoch": 2.082484555017532,
+      "grad_norm": 0.5179191384302583,
+      "learning_rate": 2.5874519643525864e-06,
+      "loss": 0.1652,
+      "step": 6236
+    },
+    {
+      "epoch": 2.0828185005844047,
+      "grad_norm": 0.5235684818125339,
+      "learning_rate": 2.5857499990743706e-06,
+      "loss": 0.1662,
+      "step": 6237
+    },
+    {
+      "epoch": 2.0831524461512774,
+      "grad_norm": 0.4954099960770102,
+      "learning_rate": 2.584048398497047e-06,
+      "loss": 0.1576,
+      "step": 6238
+    },
+    {
+      "epoch": 2.08348639171815,
+      "grad_norm": 0.5232717340566326,
+      "learning_rate": 2.5823471628776574e-06,
+      "loss": 0.1611,
+      "step": 6239
+    },
+    {
+      "epoch": 2.0838203372850224,
+      "grad_norm": 0.518556939706427,
+      "learning_rate": 2.5806462924731955e-06,
+      "loss": 0.1651,
+      "step": 6240
+    },
+    {
+      "epoch": 2.084154282851895,
+      "grad_norm": 0.4616700212286318,
+      "learning_rate": 2.5789457875405986e-06,
+      "loss": 0.1516,
+      "step": 6241
+    },
+    {
+      "epoch": 2.084488228418768,
+      "grad_norm": 0.5116149450430864,
+      "learning_rate": 2.57724564833675e-06,
+      "loss": 0.1602,
+      "step": 6242
+    },
+    {
+      "epoch": 2.0848221739856405,
+      "grad_norm": 0.4893670944675162,
+      "learning_rate": 2.5755458751184705e-06,
+      "loss": 0.1565,
+      "step": 6243
+    },
+    {
+      "epoch": 2.085156119552513,
+      "grad_norm": 0.4520354757290471,
+      "learning_rate": 2.5738464681425356e-06,
+      "loss": 0.1531,
+      "step": 6244
+    },
+    {
+      "epoch": 2.0854900651193855,
+      "grad_norm": 0.5233378545227989,
+      "learning_rate": 2.5721474276656566e-06,
+      "loss": 0.1656,
+      "step": 6245
+    },
+    {
+      "epoch": 2.0858240106862582,
+      "grad_norm": 0.5282369500802311,
+      "learning_rate": 2.5704487539444956e-06,
+      "loss": 0.17,
+      "step": 6246
+    },
+    {
+      "epoch": 2.086157956253131,
+      "grad_norm": 0.572079378431354,
+      "learning_rate": 2.5687504472356596e-06,
+      "loss": 0.1673,
+      "step": 6247
+    },
+    {
+      "epoch": 2.086491901820003,
+      "grad_norm": 0.5136764358982354,
+      "learning_rate": 2.5670525077956944e-06,
+      "loss": 0.1672,
+      "step": 6248
+    },
+    {
+      "epoch": 2.086825847386876,
+      "grad_norm": 0.4780306930943609,
+      "learning_rate": 2.5653549358810957e-06,
+      "loss": 0.1512,
+      "step": 6249
+    },
+    {
+      "epoch": 2.0871597929537486,
+      "grad_norm": 0.5351708052915466,
+      "learning_rate": 2.563657731748299e-06,
+      "loss": 0.1616,
+      "step": 6250
+    },
+    {
+      "epoch": 2.087493738520621,
+      "grad_norm": 0.48212931576855894,
+      "learning_rate": 2.5619608956536895e-06,
+      "loss": 0.1541,
+      "step": 6251
+    },
+    {
+      "epoch": 2.0878276840874936,
+      "grad_norm": 0.4882011252089522,
+      "learning_rate": 2.5602644278535937e-06,
+      "loss": 0.1689,
+      "step": 6252
+    },
+    {
+      "epoch": 2.0881616296543664,
+      "grad_norm": 0.47356586066482304,
+      "learning_rate": 2.558568328604285e-06,
+      "loss": 0.1535,
+      "step": 6253
+    },
+    {
+      "epoch": 2.088495575221239,
+      "grad_norm": 0.5437862879169038,
+      "learning_rate": 2.5568725981619747e-06,
+      "loss": 0.1729,
+      "step": 6254
+    },
+    {
+      "epoch": 2.0888295207881113,
+      "grad_norm": 0.5035531617051279,
+      "learning_rate": 2.5551772367828276e-06,
+      "loss": 0.1774,
+      "step": 6255
+    },
+    {
+      "epoch": 2.089163466354984,
+      "grad_norm": 0.4478316519506567,
+      "learning_rate": 2.5534822447229436e-06,
+      "loss": 0.146,
+      "step": 6256
+    },
+    {
+      "epoch": 2.089497411921857,
+      "grad_norm": 0.5394101004610315,
+      "learning_rate": 2.551787622238376e-06,
+      "loss": 0.1699,
+      "step": 6257
+    },
+    {
+      "epoch": 2.0898313574887295,
+      "grad_norm": 0.49097878747683316,
+      "learning_rate": 2.5500933695851104e-06,
+      "loss": 0.155,
+      "step": 6258
+    },
+    {
+      "epoch": 2.0901653030556018,
+      "grad_norm": 0.5146220478613275,
+      "learning_rate": 2.548399487019092e-06,
+      "loss": 0.1591,
+      "step": 6259
+    },
+    {
+      "epoch": 2.0904992486224745,
+      "grad_norm": 0.5347843554604156,
+      "learning_rate": 2.5467059747961953e-06,
+      "loss": 0.1591,
+      "step": 6260
+    },
+    {
+      "epoch": 2.090833194189347,
+      "grad_norm": 0.5233614282885114,
+      "learning_rate": 2.54501283317225e-06,
+      "loss": 0.1649,
+      "step": 6261
+    },
+    {
+      "epoch": 2.09116713975622,
+      "grad_norm": 0.5146286619898135,
+      "learning_rate": 2.5433200624030212e-06,
+      "loss": 0.161,
+      "step": 6262
+    },
+    {
+      "epoch": 2.091501085323092,
+      "grad_norm": 0.5020319234607654,
+      "learning_rate": 2.541627662744225e-06,
+      "loss": 0.1574,
+      "step": 6263
+    },
+    {
+      "epoch": 2.091835030889965,
+      "grad_norm": 0.5004055836984991,
+      "learning_rate": 2.5399356344515138e-06,
+      "loss": 0.1572,
+      "step": 6264
+    },
+    {
+      "epoch": 2.0921689764568376,
+      "grad_norm": 0.4979420565995763,
+      "learning_rate": 2.538243977780494e-06,
+      "loss": 0.1564,
+      "step": 6265
+    },
+    {
+      "epoch": 2.0925029220237104,
+      "grad_norm": 0.51833341034248,
+      "learning_rate": 2.5365526929867056e-06,
+      "loss": 0.1628,
+      "step": 6266
+    },
+    {
+      "epoch": 2.0928368675905826,
+      "grad_norm": 0.5459444259176912,
+      "learning_rate": 2.534861780325642e-06,
+      "loss": 0.1555,
+      "step": 6267
+    },
+    {
+      "epoch": 2.0931708131574553,
+      "grad_norm": 0.4788442607046698,
+      "learning_rate": 2.53317124005273e-06,
+      "loss": 0.1499,
+      "step": 6268
+    },
+    {
+      "epoch": 2.093504758724328,
+      "grad_norm": 0.5008550202971368,
+      "learning_rate": 2.5314810724233502e-06,
+      "loss": 0.1683,
+      "step": 6269
+    },
+    {
+      "epoch": 2.0938387042912003,
+      "grad_norm": 0.522704888977689,
+      "learning_rate": 2.529791277692818e-06,
+      "loss": 0.1655,
+      "step": 6270
+    },
+    {
+      "epoch": 2.094172649858073,
+      "grad_norm": 0.507328669366408,
+      "learning_rate": 2.5281018561163996e-06,
+      "loss": 0.1635,
+      "step": 6271
+    },
+    {
+      "epoch": 2.0945065954249458,
+      "grad_norm": 0.5081138020513082,
+      "learning_rate": 2.5264128079493033e-06,
+      "loss": 0.1694,
+      "step": 6272
+    },
+    {
+      "epoch": 2.0948405409918185,
+      "grad_norm": 0.4772258614297116,
+      "learning_rate": 2.524724133446676e-06,
+      "loss": 0.1616,
+      "step": 6273
+    },
+    {
+      "epoch": 2.0951744865586908,
+      "grad_norm": 0.495524661859938,
+      "learning_rate": 2.523035832863614e-06,
+      "loss": 0.1641,
+      "step": 6274
+    },
+    {
+      "epoch": 2.0955084321255635,
+      "grad_norm": 0.4529427960497026,
+      "learning_rate": 2.521347906455154e-06,
+      "loss": 0.1517,
+      "step": 6275
+    },
+    {
+      "epoch": 2.095842377692436,
+      "grad_norm": 0.511282048473958,
+      "learning_rate": 2.5196603544762804e-06,
+      "loss": 0.1628,
+      "step": 6276
+    },
+    {
+      "epoch": 2.096176323259309,
+      "grad_norm": 0.4720720037080335,
+      "learning_rate": 2.5179731771819133e-06,
+      "loss": 0.1573,
+      "step": 6277
+    },
+    {
+      "epoch": 2.096510268826181,
+      "grad_norm": 0.46615925544747905,
+      "learning_rate": 2.5162863748269247e-06,
+      "loss": 0.1518,
+      "step": 6278
+    },
+    {
+      "epoch": 2.096844214393054,
+      "grad_norm": 0.5095981778636587,
+      "learning_rate": 2.514599947666122e-06,
+      "loss": 0.1593,
+      "step": 6279
+    },
+    {
+      "epoch": 2.0971781599599266,
+      "grad_norm": 0.5176486525595663,
+      "learning_rate": 2.5129138959542633e-06,
+      "loss": 0.1702,
+      "step": 6280
+    },
+    {
+      "epoch": 2.0975121055267993,
+      "grad_norm": 0.476234499249456,
+      "learning_rate": 2.5112282199460415e-06,
+      "loss": 0.1543,
+      "step": 6281
+    },
+    {
+      "epoch": 2.0978460510936716,
+      "grad_norm": 0.522206312703102,
+      "learning_rate": 2.5095429198961056e-06,
+      "loss": 0.173,
+      "step": 6282
+    },
+    {
+      "epoch": 2.0981799966605443,
+      "grad_norm": 0.5078914776921112,
+      "learning_rate": 2.507857996059034e-06,
+      "loss": 0.1628,
+      "step": 6283
+    },
+    {
+      "epoch": 2.098513942227417,
+      "grad_norm": 0.48432480768589903,
+      "learning_rate": 2.5061734486893574e-06,
+      "loss": 0.1541,
+      "step": 6284
+    },
+    {
+      "epoch": 2.0988478877942893,
+      "grad_norm": 0.49121279758470837,
+      "learning_rate": 2.504489278041544e-06,
+      "loss": 0.1597,
+      "step": 6285
+    },
+    {
+      "epoch": 2.099181833361162,
+      "grad_norm": 0.47385292608458507,
+      "learning_rate": 2.5028054843700102e-06,
+      "loss": 0.1531,
+      "step": 6286
+    },
+    {
+      "epoch": 2.0995157789280348,
+      "grad_norm": 0.4841879216861,
+      "learning_rate": 2.501122067929108e-06,
+      "loss": 0.1517,
+      "step": 6287
+    },
+    {
+      "epoch": 2.0998497244949075,
+      "grad_norm": 0.479421292773347,
+      "learning_rate": 2.4994390289731446e-06,
+      "loss": 0.1551,
+      "step": 6288
+    },
+    {
+      "epoch": 2.1001836700617798,
+      "grad_norm": 0.5527499779534556,
+      "learning_rate": 2.497756367756357e-06,
+      "loss": 0.1618,
+      "step": 6289
+    },
+    {
+      "epoch": 2.1005176156286525,
+      "grad_norm": 0.4913185942010325,
+      "learning_rate": 2.496074084532935e-06,
+      "loss": 0.154,
+      "step": 6290
+    },
+    {
+      "epoch": 2.100851561195525,
+      "grad_norm": 0.5031384055965269,
+      "learning_rate": 2.4943921795570033e-06,
+      "loss": 0.1479,
+      "step": 6291
+    },
+    {
+      "epoch": 2.101185506762398,
+      "grad_norm": 0.49927324992800404,
+      "learning_rate": 2.4927106530826372e-06,
+      "loss": 0.1658,
+      "step": 6292
+    },
+    {
+      "epoch": 2.10151945232927,
+      "grad_norm": 0.5309889957366104,
+      "learning_rate": 2.491029505363848e-06,
+      "loss": 0.1422,
+      "step": 6293
+    },
+    {
+      "epoch": 2.101853397896143,
+      "grad_norm": 0.49560503238802656,
+      "learning_rate": 2.489348736654593e-06,
+      "loss": 0.1625,
+      "step": 6294
+    },
+    {
+      "epoch": 2.1021873434630156,
+      "grad_norm": 0.4782744370282557,
+      "learning_rate": 2.4876683472087767e-06,
+      "loss": 0.1508,
+      "step": 6295
+    },
+    {
+      "epoch": 2.1025212890298883,
+      "grad_norm": 0.4648251202295055,
+      "learning_rate": 2.4859883372802357e-06,
+      "loss": 0.1514,
+      "step": 6296
+    },
+    {
+      "epoch": 2.1028552345967606,
+      "grad_norm": 0.47761287277219516,
+      "learning_rate": 2.484308707122758e-06,
+      "loss": 0.1469,
+      "step": 6297
+    },
+    {
+      "epoch": 2.1031891801636333,
+      "grad_norm": 0.5120606241455523,
+      "learning_rate": 2.4826294569900725e-06,
+      "loss": 0.1599,
+      "step": 6298
+    },
+    {
+      "epoch": 2.103523125730506,
+      "grad_norm": 0.5192893295764826,
+      "learning_rate": 2.4809505871358476e-06,
+      "loss": 0.1647,
+      "step": 6299
+    },
+    {
+      "epoch": 2.1038570712973783,
+      "grad_norm": 0.49615949952418487,
+      "learning_rate": 2.4792720978136967e-06,
+      "loss": 0.1615,
+      "step": 6300
+    },
+    {
+      "epoch": 2.104191016864251,
+      "grad_norm": 0.5171662212170216,
+      "learning_rate": 2.4775939892771787e-06,
+      "loss": 0.1544,
+      "step": 6301
+    },
+    {
+      "epoch": 2.1045249624311237,
+      "grad_norm": 0.5549893785210516,
+      "learning_rate": 2.4759162617797873e-06,
+      "loss": 0.1687,
+      "step": 6302
+    },
+    {
+      "epoch": 2.1048589079979965,
+      "grad_norm": 0.5230583873016571,
+      "learning_rate": 2.4742389155749657e-06,
+      "loss": 0.1651,
+      "step": 6303
+    },
+    {
+      "epoch": 2.1051928535648687,
+      "grad_norm": 0.4862306544241593,
+      "learning_rate": 2.472561950916094e-06,
+      "loss": 0.1553,
+      "step": 6304
+    },
+    {
+      "epoch": 2.1055267991317415,
+      "grad_norm": 0.5309377784725847,
+      "learning_rate": 2.4708853680565e-06,
+      "loss": 0.1652,
+      "step": 6305
+    },
+    {
+      "epoch": 2.105860744698614,
+      "grad_norm": 0.5055232116428447,
+      "learning_rate": 2.4692091672494494e-06,
+      "loss": 0.163,
+      "step": 6306
+    },
+    {
+      "epoch": 2.106194690265487,
+      "grad_norm": 0.5642390943911669,
+      "learning_rate": 2.4675333487481558e-06,
+      "loss": 0.1711,
+      "step": 6307
+    },
+    {
+      "epoch": 2.106528635832359,
+      "grad_norm": 0.461679532517826,
+      "learning_rate": 2.4658579128057665e-06,
+      "loss": 0.1534,
+      "step": 6308
+    },
+    {
+      "epoch": 2.106862581399232,
+      "grad_norm": 0.5100394097368949,
+      "learning_rate": 2.4641828596753803e-06,
+      "loss": 0.1637,
+      "step": 6309
+    },
+    {
+      "epoch": 2.1071965269661046,
+      "grad_norm": 0.46839005082728324,
+      "learning_rate": 2.4625081896100294e-06,
+      "loss": 0.1535,
+      "step": 6310
+    },
+    {
+      "epoch": 2.1075304725329773,
+      "grad_norm": 0.4964670435707832,
+      "learning_rate": 2.4608339028626943e-06,
+      "loss": 0.1647,
+      "step": 6311
+    },
+    {
+      "epoch": 2.1078644180998496,
+      "grad_norm": 0.48855274829211875,
+      "learning_rate": 2.4591599996862957e-06,
+      "loss": 0.1639,
+      "step": 6312
+    },
+    {
+      "epoch": 2.1081983636667223,
+      "grad_norm": 0.5450080590558118,
+      "learning_rate": 2.457486480333699e-06,
+      "loss": 0.174,
+      "step": 6313
+    },
+    {
+      "epoch": 2.108532309233595,
+      "grad_norm": 0.5224423801079273,
+      "learning_rate": 2.4558133450577044e-06,
+      "loss": 0.1577,
+      "step": 6314
+    },
+    {
+      "epoch": 2.1088662548004677,
+      "grad_norm": 0.5174969356229879,
+      "learning_rate": 2.4541405941110626e-06,
+      "loss": 0.162,
+      "step": 6315
+    },
+    {
+      "epoch": 2.10920020036734,
+      "grad_norm": 0.4964356122182152,
+      "learning_rate": 2.452468227746459e-06,
+      "loss": 0.1575,
+      "step": 6316
+    },
+    {
+      "epoch": 2.1095341459342127,
+      "grad_norm": 0.5613011623009709,
+      "learning_rate": 2.4507962462165254e-06,
+      "loss": 0.1669,
+      "step": 6317
+    },
+    {
+      "epoch": 2.1098680915010855,
+      "grad_norm": 0.47625742049767633,
+      "learning_rate": 2.449124649773835e-06,
+      "loss": 0.1559,
+      "step": 6318
+    },
+    {
+      "epoch": 2.1102020370679577,
+      "grad_norm": 0.6372927108132521,
+      "learning_rate": 2.4474534386709036e-06,
+      "loss": 0.1766,
+      "step": 6319
+    },
+    {
+      "epoch": 2.1105359826348304,
+      "grad_norm": 0.5062800748969212,
+      "learning_rate": 2.4457826131601835e-06,
+      "loss": 0.1498,
+      "step": 6320
+    },
+    {
+      "epoch": 2.110869928201703,
+      "grad_norm": 0.5307208312611601,
+      "learning_rate": 2.444112173494077e-06,
+      "loss": 0.1615,
+      "step": 6321
+    },
+    {
+      "epoch": 2.111203873768576,
+      "grad_norm": 0.5407648531518704,
+      "learning_rate": 2.4424421199249194e-06,
+      "loss": 0.1691,
+      "step": 6322
+    },
+    {
+      "epoch": 2.111537819335448,
+      "grad_norm": 0.5343008151179298,
+      "learning_rate": 2.440772452704993e-06,
+      "loss": 0.1642,
+      "step": 6323
+    },
+    {
+      "epoch": 2.111871764902321,
+      "grad_norm": 0.49431699281947467,
+      "learning_rate": 2.4391031720865246e-06,
+      "loss": 0.1634,
+      "step": 6324
+    },
+    {
+      "epoch": 2.1122057104691936,
+      "grad_norm": 0.46526916827882536,
+      "learning_rate": 2.4374342783216732e-06,
+      "loss": 0.1483,
+      "step": 6325
+    },
+    {
+      "epoch": 2.1125396560360663,
+      "grad_norm": 0.5278048019270066,
+      "learning_rate": 2.435765771662549e-06,
+      "loss": 0.1569,
+      "step": 6326
+    },
+    {
+      "epoch": 2.1128736016029386,
+      "grad_norm": 0.48140891039036754,
+      "learning_rate": 2.4340976523611957e-06,
+      "loss": 0.1496,
+      "step": 6327
+    },
+    {
+      "epoch": 2.1132075471698113,
+      "grad_norm": 0.49088279957734005,
+      "learning_rate": 2.4324299206696057e-06,
+      "loss": 0.1609,
+      "step": 6328
+    },
+    {
+      "epoch": 2.113541492736684,
+      "grad_norm": 0.49893234907819617,
+      "learning_rate": 2.4307625768397077e-06,
+      "loss": 0.164,
+      "step": 6329
+    },
+    {
+      "epoch": 2.1138754383035567,
+      "grad_norm": 0.5409361980416141,
+      "learning_rate": 2.4290956211233757e-06,
+      "loss": 0.1665,
+      "step": 6330
+    },
+    {
+      "epoch": 2.114209383870429,
+      "grad_norm": 0.5101008229874319,
+      "learning_rate": 2.42742905377242e-06,
+      "loss": 0.1529,
+      "step": 6331
+    },
+    {
+      "epoch": 2.1145433294373017,
+      "grad_norm": 0.5238806985611678,
+      "learning_rate": 2.4257628750385987e-06,
+      "loss": 0.1633,
+      "step": 6332
+    },
+    {
+      "epoch": 2.1148772750041744,
+      "grad_norm": 0.4728995120386821,
+      "learning_rate": 2.424097085173604e-06,
+      "loss": 0.1476,
+      "step": 6333
+    },
+    {
+      "epoch": 2.1152112205710467,
+      "grad_norm": 0.5183928863262791,
+      "learning_rate": 2.4224316844290747e-06,
+      "loss": 0.1603,
+      "step": 6334
+    },
+    {
+      "epoch": 2.1155451661379194,
+      "grad_norm": 0.5244171017941279,
+      "learning_rate": 2.4207666730565893e-06,
+      "loss": 0.1601,
+      "step": 6335
+    },
+    {
+      "epoch": 2.115879111704792,
+      "grad_norm": 0.483669792524147,
+      "learning_rate": 2.4191020513076697e-06,
+      "loss": 0.1643,
+      "step": 6336
+    },
+    {
+      "epoch": 2.116213057271665,
+      "grad_norm": 0.5036617136208132,
+      "learning_rate": 2.4174378194337715e-06,
+      "loss": 0.1617,
+      "step": 6337
+    },
+    {
+      "epoch": 2.116547002838537,
+      "grad_norm": 0.5000061906999408,
+      "learning_rate": 2.4157739776863023e-06,
+      "loss": 0.16,
+      "step": 6338
+    },
+    {
+      "epoch": 2.11688094840541,
+      "grad_norm": 0.48562974176186696,
+      "learning_rate": 2.4141105263166e-06,
+      "loss": 0.1594,
+      "step": 6339
+    },
+    {
+      "epoch": 2.1172148939722826,
+      "grad_norm": 0.4796849682215605,
+      "learning_rate": 2.41244746557595e-06,
+      "loss": 0.152,
+      "step": 6340
+    },
+    {
+      "epoch": 2.1175488395391553,
+      "grad_norm": 0.45539955587257097,
+      "learning_rate": 2.4107847957155784e-06,
+      "loss": 0.157,
+      "step": 6341
+    },
+    {
+      "epoch": 2.1178827851060276,
+      "grad_norm": 0.46438375956752614,
+      "learning_rate": 2.409122516986652e-06,
+      "loss": 0.1549,
+      "step": 6342
+    },
+    {
+      "epoch": 2.1182167306729003,
+      "grad_norm": 0.4514297641602212,
+      "learning_rate": 2.4074606296402735e-06,
+      "loss": 0.1519,
+      "step": 6343
+    },
+    {
+      "epoch": 2.118550676239773,
+      "grad_norm": 0.4859800183560971,
+      "learning_rate": 2.405799133927496e-06,
+      "loss": 0.1593,
+      "step": 6344
+    },
+    {
+      "epoch": 2.1188846218066457,
+      "grad_norm": 0.5005344670210375,
+      "learning_rate": 2.404138030099303e-06,
+      "loss": 0.1597,
+      "step": 6345
+    },
+    {
+      "epoch": 2.119218567373518,
+      "grad_norm": 0.48375070616490734,
+      "learning_rate": 2.4024773184066253e-06,
+      "loss": 0.1665,
+      "step": 6346
+    },
+    {
+      "epoch": 2.1195525129403907,
+      "grad_norm": 0.472840448624495,
+      "learning_rate": 2.4008169991003356e-06,
+      "loss": 0.1452,
+      "step": 6347
+    },
+    {
+      "epoch": 2.1198864585072634,
+      "grad_norm": 0.49084422815911033,
+      "learning_rate": 2.3991570724312405e-06,
+      "loss": 0.1528,
+      "step": 6348
+    },
+    {
+      "epoch": 2.1202204040741357,
+      "grad_norm": 0.4692583452343208,
+      "learning_rate": 2.3974975386500958e-06,
+      "loss": 0.155,
+      "step": 6349
+    },
+    {
+      "epoch": 2.1205543496410084,
+      "grad_norm": 0.5062663536376966,
+      "learning_rate": 2.3958383980075896e-06,
+      "loss": 0.1633,
+      "step": 6350
+    },
+    {
+      "epoch": 2.120888295207881,
+      "grad_norm": 0.5403727659874564,
+      "learning_rate": 2.394179650754358e-06,
+      "loss": 0.1671,
+      "step": 6351
+    },
+    {
+      "epoch": 2.121222240774754,
+      "grad_norm": 0.5235630447457905,
+      "learning_rate": 2.3925212971409688e-06,
+      "loss": 0.1622,
+      "step": 6352
+    },
+    {
+      "epoch": 2.121556186341626,
+      "grad_norm": 0.4699026763388292,
+      "learning_rate": 2.3908633374179436e-06,
+      "loss": 0.1518,
+      "step": 6353
+    },
+    {
+      "epoch": 2.121890131908499,
+      "grad_norm": 0.5167884671729314,
+      "learning_rate": 2.3892057718357308e-06,
+      "loss": 0.1604,
+      "step": 6354
+    },
+    {
+      "epoch": 2.1222240774753716,
+      "grad_norm": 0.4623642417869892,
+      "learning_rate": 2.3875486006447294e-06,
+      "loss": 0.1491,
+      "step": 6355
+    },
+    {
+      "epoch": 2.1225580230422443,
+      "grad_norm": 0.5158989070059895,
+      "learning_rate": 2.3858918240952703e-06,
+      "loss": 0.1576,
+      "step": 6356
+    },
+    {
+      "epoch": 2.1228919686091166,
+      "grad_norm": 0.5110491439871296,
+      "learning_rate": 2.384235442437632e-06,
+      "loss": 0.1651,
+      "step": 6357
+    },
+    {
+      "epoch": 2.1232259141759893,
+      "grad_norm": 0.5119404415532335,
+      "learning_rate": 2.3825794559220296e-06,
+      "loss": 0.1729,
+      "step": 6358
+    },
+    {
+      "epoch": 2.123559859742862,
+      "grad_norm": 0.5501830648498629,
+      "learning_rate": 2.380923864798621e-06,
+      "loss": 0.1692,
+      "step": 6359
+    },
+    {
+      "epoch": 2.1238938053097347,
+      "grad_norm": 0.5038516827919598,
+      "learning_rate": 2.3792686693174993e-06,
+      "loss": 0.1627,
+      "step": 6360
+    },
+    {
+      "epoch": 2.124227750876607,
+      "grad_norm": 0.5332702528419736,
+      "learning_rate": 2.3776138697287055e-06,
+      "loss": 0.1651,
+      "step": 6361
+    },
+    {
+      "epoch": 2.1245616964434797,
+      "grad_norm": 0.4763237636455076,
+      "learning_rate": 2.3759594662822122e-06,
+      "loss": 0.1436,
+      "step": 6362
+    },
+    {
+      "epoch": 2.1248956420103524,
+      "grad_norm": 0.5286176751132988,
+      "learning_rate": 2.3743054592279386e-06,
+      "loss": 0.1605,
+      "step": 6363
+    },
+    {
+      "epoch": 2.125229587577225,
+      "grad_norm": 0.47547415889114697,
+      "learning_rate": 2.372651848815742e-06,
+      "loss": 0.1473,
+      "step": 6364
+    },
+    {
+      "epoch": 2.1255635331440974,
+      "grad_norm": 0.4533019654138708,
+      "learning_rate": 2.370998635295421e-06,
+      "loss": 0.1531,
+      "step": 6365
+    },
+    {
+      "epoch": 2.12589747871097,
+      "grad_norm": 0.45228899024811703,
+      "learning_rate": 2.3693458189167106e-06,
+      "loss": 0.1517,
+      "step": 6366
+    },
+    {
+      "epoch": 2.126231424277843,
+      "grad_norm": 0.5400654982070013,
+      "learning_rate": 2.3676933999292905e-06,
+      "loss": 0.1697,
+      "step": 6367
+    },
+    {
+      "epoch": 2.126565369844715,
+      "grad_norm": 0.5439371706519016,
+      "learning_rate": 2.366041378582775e-06,
+      "loss": 0.171,
+      "step": 6368
+    },
+    {
+      "epoch": 2.126899315411588,
+      "grad_norm": 0.5018152963433691,
+      "learning_rate": 2.364389755126723e-06,
+      "loss": 0.1528,
+      "step": 6369
+    },
+    {
+      "epoch": 2.1272332609784605,
+      "grad_norm": 0.487011525737873,
+      "learning_rate": 2.3627385298106344e-06,
+      "loss": 0.1598,
+      "step": 6370
+    },
+    {
+      "epoch": 2.1275672065453333,
+      "grad_norm": 0.5186675987624135,
+      "learning_rate": 2.361087702883941e-06,
+      "loss": 0.1655,
+      "step": 6371
+    },
+    {
+      "epoch": 2.1279011521122055,
+      "grad_norm": 0.4740219296383837,
+      "learning_rate": 2.359437274596024e-06,
+      "loss": 0.1535,
+      "step": 6372
+    },
+    {
+      "epoch": 2.1282350976790783,
+      "grad_norm": 0.4723892674328739,
+      "learning_rate": 2.357787245196197e-06,
+      "loss": 0.1461,
+      "step": 6373
+    },
+    {
+      "epoch": 2.128569043245951,
+      "grad_norm": 0.4653901182497193,
+      "learning_rate": 2.3561376149337188e-06,
+      "loss": 0.1542,
+      "step": 6374
+    },
+    {
+      "epoch": 2.1289029888128237,
+      "grad_norm": 0.5055198140337246,
+      "learning_rate": 2.3544883840577815e-06,
+      "loss": 0.1614,
+      "step": 6375
+    },
+    {
+      "epoch": 2.129236934379696,
+      "grad_norm": 0.49423346871659607,
+      "learning_rate": 2.352839552817527e-06,
+      "loss": 0.1606,
+      "step": 6376
+    },
+    {
+      "epoch": 2.1295708799465687,
+      "grad_norm": 0.5028984209451132,
+      "learning_rate": 2.3511911214620255e-06,
+      "loss": 0.1524,
+      "step": 6377
+    },
+    {
+      "epoch": 2.1299048255134414,
+      "grad_norm": 0.534742967358469,
+      "learning_rate": 2.3495430902402956e-06,
+      "loss": 0.1685,
+      "step": 6378
+    },
+    {
+      "epoch": 2.1302387710803137,
+      "grad_norm": 0.5180111072516133,
+      "learning_rate": 2.3478954594012884e-06,
+      "loss": 0.1613,
+      "step": 6379
+    },
+    {
+      "epoch": 2.1305727166471864,
+      "grad_norm": 0.4735330750800575,
+      "learning_rate": 2.346248229193901e-06,
+      "loss": 0.1488,
+      "step": 6380
+    },
+    {
+      "epoch": 2.130906662214059,
+      "grad_norm": 0.5283937925064244,
+      "learning_rate": 2.344601399866962e-06,
+      "loss": 0.1635,
+      "step": 6381
+    },
+    {
+      "epoch": 2.131240607780932,
+      "grad_norm": 0.5098981708034045,
+      "learning_rate": 2.342954971669252e-06,
+      "loss": 0.1621,
+      "step": 6382
+    },
+    {
+      "epoch": 2.131574553347804,
+      "grad_norm": 0.5297193277625377,
+      "learning_rate": 2.341308944849477e-06,
+      "loss": 0.1697,
+      "step": 6383
+    },
+    {
+      "epoch": 2.131908498914677,
+      "grad_norm": 0.4851259261560039,
+      "learning_rate": 2.3396633196562924e-06,
+      "loss": 0.154,
+      "step": 6384
+    },
+    {
+      "epoch": 2.1322424444815495,
+      "grad_norm": 0.5309045479819359,
+      "learning_rate": 2.3380180963382866e-06,
+      "loss": 0.1625,
+      "step": 6385
+    },
+    {
+      "epoch": 2.1325763900484223,
+      "grad_norm": 0.529874878718374,
+      "learning_rate": 2.3363732751439926e-06,
+      "loss": 0.1707,
+      "step": 6386
+    },
+    {
+      "epoch": 2.1329103356152945,
+      "grad_norm": 0.5019851431122059,
+      "learning_rate": 2.334728856321875e-06,
+      "loss": 0.1647,
+      "step": 6387
+    },
+    {
+      "epoch": 2.1332442811821672,
+      "grad_norm": 0.5565660036653327,
+      "learning_rate": 2.33308484012035e-06,
+      "loss": 0.1724,
+      "step": 6388
+    },
+    {
+      "epoch": 2.13357822674904,
+      "grad_norm": 0.4636960099887346,
+      "learning_rate": 2.33144122678776e-06,
+      "loss": 0.1483,
+      "step": 6389
+    },
+    {
+      "epoch": 2.1339121723159127,
+      "grad_norm": 0.4886503505718227,
+      "learning_rate": 2.3297980165723953e-06,
+      "loss": 0.1556,
+      "step": 6390
+    },
+    {
+      "epoch": 2.134246117882785,
+      "grad_norm": 0.5243203293316878,
+      "learning_rate": 2.3281552097224798e-06,
+      "loss": 0.1692,
+      "step": 6391
+    },
+    {
+      "epoch": 2.1345800634496577,
+      "grad_norm": 0.5512527072406636,
+      "learning_rate": 2.326512806486181e-06,
+      "loss": 0.1702,
+      "step": 6392
+    },
+    {
+      "epoch": 2.1349140090165304,
+      "grad_norm": 0.4869770099606462,
+      "learning_rate": 2.3248708071116005e-06,
+      "loss": 0.1604,
+      "step": 6393
+    },
+    {
+      "epoch": 2.135247954583403,
+      "grad_norm": 0.4892570672975904,
+      "learning_rate": 2.323229211846783e-06,
+      "loss": 0.1547,
+      "step": 6394
+    },
+    {
+      "epoch": 2.1355819001502754,
+      "grad_norm": 0.49250010065010924,
+      "learning_rate": 2.3215880209397133e-06,
+      "loss": 0.1537,
+      "step": 6395
+    },
+    {
+      "epoch": 2.135915845717148,
+      "grad_norm": 0.47927021514041757,
+      "learning_rate": 2.319947234638308e-06,
+      "loss": 0.1532,
+      "step": 6396
+    },
+    {
+      "epoch": 2.136249791284021,
+      "grad_norm": 0.5035140794080505,
+      "learning_rate": 2.3183068531904317e-06,
+      "loss": 0.1624,
+      "step": 6397
+    },
+    {
+      "epoch": 2.136583736850893,
+      "grad_norm": 0.5127475929816925,
+      "learning_rate": 2.3166668768438772e-06,
+      "loss": 0.1566,
+      "step": 6398
+    },
+    {
+      "epoch": 2.136917682417766,
+      "grad_norm": 0.5021756770030403,
+      "learning_rate": 2.31502730584639e-06,
+      "loss": 0.16,
+      "step": 6399
+    },
+    {
+      "epoch": 2.1372516279846385,
+      "grad_norm": 0.49666732574803907,
+      "learning_rate": 2.313388140445641e-06,
+      "loss": 0.1579,
+      "step": 6400
+    },
+    {
+      "epoch": 2.1375855735515112,
+      "grad_norm": 0.48350078859534124,
+      "learning_rate": 2.311749380889249e-06,
+      "loss": 0.1654,
+      "step": 6401
+    },
+    {
+      "epoch": 2.1379195191183835,
+      "grad_norm": 0.5439150444830929,
+      "learning_rate": 2.310111027424764e-06,
+      "loss": 0.1725,
+      "step": 6402
+    },
+    {
+      "epoch": 2.1382534646852562,
+      "grad_norm": 0.48456783877805504,
+      "learning_rate": 2.308473080299683e-06,
+      "loss": 0.1408,
+      "step": 6403
+    },
+    {
+      "epoch": 2.138587410252129,
+      "grad_norm": 0.4718955603006833,
+      "learning_rate": 2.3068355397614313e-06,
+      "loss": 0.1574,
+      "step": 6404
+    },
+    {
+      "epoch": 2.1389213558190017,
+      "grad_norm": 0.5054752478801913,
+      "learning_rate": 2.3051984060573855e-06,
+      "loss": 0.1601,
+      "step": 6405
+    },
+    {
+      "epoch": 2.139255301385874,
+      "grad_norm": 0.4728648535556213,
+      "learning_rate": 2.303561679434849e-06,
+      "loss": 0.1509,
+      "step": 6406
+    },
+    {
+      "epoch": 2.1395892469527467,
+      "grad_norm": 0.527707051909604,
+      "learning_rate": 2.3019253601410725e-06,
+      "loss": 0.1704,
+      "step": 6407
+    },
+    {
+      "epoch": 2.1399231925196194,
+      "grad_norm": 0.47206826548069847,
+      "learning_rate": 2.300289448423237e-06,
+      "loss": 0.1479,
+      "step": 6408
+    },
+    {
+      "epoch": 2.140257138086492,
+      "grad_norm": 0.5071150172469431,
+      "learning_rate": 2.2986539445284705e-06,
+      "loss": 0.1628,
+      "step": 6409
+    },
+    {
+      "epoch": 2.1405910836533644,
+      "grad_norm": 0.5114902739240026,
+      "learning_rate": 2.2970188487038293e-06,
+      "loss": 0.1616,
+      "step": 6410
+    },
+    {
+      "epoch": 2.140925029220237,
+      "grad_norm": 0.5143644585016027,
+      "learning_rate": 2.295384161196321e-06,
+      "loss": 0.1639,
+      "step": 6411
+    },
+    {
+      "epoch": 2.14125897478711,
+      "grad_norm": 0.47396643773274283,
+      "learning_rate": 2.293749882252879e-06,
+      "loss": 0.1453,
+      "step": 6412
+    },
+    {
+      "epoch": 2.1415929203539825,
+      "grad_norm": 0.5070902818006285,
+      "learning_rate": 2.2921160121203847e-06,
+      "loss": 0.1617,
+      "step": 6413
+    },
+    {
+      "epoch": 2.141926865920855,
+      "grad_norm": 0.5206899156340429,
+      "learning_rate": 2.290482551045649e-06,
+      "loss": 0.1673,
+      "step": 6414
+    },
+    {
+      "epoch": 2.1422608114877275,
+      "grad_norm": 0.4910791307640534,
+      "learning_rate": 2.2888494992754294e-06,
+      "loss": 0.163,
+      "step": 6415
+    },
+    {
+      "epoch": 2.1425947570546002,
+      "grad_norm": 0.49078588542590307,
+      "learning_rate": 2.2872168570564136e-06,
+      "loss": 0.1597,
+      "step": 6416
+    },
+    {
+      "epoch": 2.1429287026214725,
+      "grad_norm": 0.48656902115080175,
+      "learning_rate": 2.2855846246352335e-06,
+      "loss": 0.1573,
+      "step": 6417
+    },
+    {
+      "epoch": 2.143262648188345,
+      "grad_norm": 0.5285475116853615,
+      "learning_rate": 2.2839528022584596e-06,
+      "loss": 0.1678,
+      "step": 6418
+    },
+    {
+      "epoch": 2.143596593755218,
+      "grad_norm": 0.5097327251095314,
+      "learning_rate": 2.2823213901725927e-06,
+      "loss": 0.1722,
+      "step": 6419
+    },
+    {
+      "epoch": 2.1439305393220907,
+      "grad_norm": 0.5020115119724882,
+      "learning_rate": 2.2806903886240815e-06,
+      "loss": 0.1567,
+      "step": 6420
+    },
+    {
+      "epoch": 2.144264484888963,
+      "grad_norm": 0.5005087225398663,
+      "learning_rate": 2.2790597978593044e-06,
+      "loss": 0.1608,
+      "step": 6421
+    },
+    {
+      "epoch": 2.1445984304558356,
+      "grad_norm": 0.4972452001072568,
+      "learning_rate": 2.2774296181245825e-06,
+      "loss": 0.1708,
+      "step": 6422
+    },
+    {
+      "epoch": 2.1449323760227084,
+      "grad_norm": 0.5307429371345069,
+      "learning_rate": 2.275799849666174e-06,
+      "loss": 0.1578,
+      "step": 6423
+    },
+    {
+      "epoch": 2.145266321589581,
+      "grad_norm": 0.5104050954727964,
+      "learning_rate": 2.274170492730277e-06,
+      "loss": 0.1669,
+      "step": 6424
+    },
+    {
+      "epoch": 2.1456002671564534,
+      "grad_norm": 0.5447982842445855,
+      "learning_rate": 2.27254154756302e-06,
+      "loss": 0.1671,
+      "step": 6425
+    },
+    {
+      "epoch": 2.145934212723326,
+      "grad_norm": 0.539396360023142,
+      "learning_rate": 2.2709130144104795e-06,
+      "loss": 0.1722,
+      "step": 6426
+    },
+    {
+      "epoch": 2.146268158290199,
+      "grad_norm": 0.48061812916772473,
+      "learning_rate": 2.26928489351866e-06,
+      "loss": 0.1572,
+      "step": 6427
+    },
+    {
+      "epoch": 2.146602103857071,
+      "grad_norm": 0.5558718837443882,
+      "learning_rate": 2.267657185133511e-06,
+      "loss": 0.1679,
+      "step": 6428
+    },
+    {
+      "epoch": 2.146936049423944,
+      "grad_norm": 0.4912517189710352,
+      "learning_rate": 2.2660298895009157e-06,
+      "loss": 0.1533,
+      "step": 6429
+    },
+    {
+      "epoch": 2.1472699949908165,
+      "grad_norm": 0.5449170299577375,
+      "learning_rate": 2.2644030068666993e-06,
+      "loss": 0.1701,
+      "step": 6430
+    },
+    {
+      "epoch": 2.147603940557689,
+      "grad_norm": 0.5249455334713858,
+      "learning_rate": 2.2627765374766175e-06,
+      "loss": 0.1519,
+      "step": 6431
+    },
+    {
+      "epoch": 2.1479378861245615,
+      "grad_norm": 0.5490812081136831,
+      "learning_rate": 2.2611504815763715e-06,
+      "loss": 0.175,
+      "step": 6432
+    },
+    {
+      "epoch": 2.148271831691434,
+      "grad_norm": 0.501323784094873,
+      "learning_rate": 2.259524839411592e-06,
+      "loss": 0.1545,
+      "step": 6433
+    },
+    {
+      "epoch": 2.148605777258307,
+      "grad_norm": 0.46873001297372796,
+      "learning_rate": 2.2578996112278535e-06,
+      "loss": 0.1472,
+      "step": 6434
+    },
+    {
+      "epoch": 2.1489397228251796,
+      "grad_norm": 0.5482354098239963,
+      "learning_rate": 2.2562747972706663e-06,
+      "loss": 0.17,
+      "step": 6435
+    },
+    {
+      "epoch": 2.149273668392052,
+      "grad_norm": 0.5127389817917001,
+      "learning_rate": 2.254650397785479e-06,
+      "loss": 0.1563,
+      "step": 6436
+    },
+    {
+      "epoch": 2.1496076139589246,
+      "grad_norm": 0.5439560842821645,
+      "learning_rate": 2.253026413017672e-06,
+      "loss": 0.1631,
+      "step": 6437
+    },
+    {
+      "epoch": 2.1499415595257974,
+      "grad_norm": 0.5200199301047186,
+      "learning_rate": 2.2514028432125722e-06,
+      "loss": 0.1587,
+      "step": 6438
+    },
+    {
+      "epoch": 2.15027550509267,
+      "grad_norm": 0.48845583828397887,
+      "learning_rate": 2.249779688615435e-06,
+      "loss": 0.1572,
+      "step": 6439
+    },
+    {
+      "epoch": 2.1506094506595423,
+      "grad_norm": 0.4553983917786923,
+      "learning_rate": 2.248156949471459e-06,
+      "loss": 0.1534,
+      "step": 6440
+    },
+    {
+      "epoch": 2.150943396226415,
+      "grad_norm": 0.4848004378929125,
+      "learning_rate": 2.2465346260257786e-06,
+      "loss": 0.1607,
+      "step": 6441
+    },
+    {
+      "epoch": 2.151277341793288,
+      "grad_norm": 0.49347866729377543,
+      "learning_rate": 2.2449127185234626e-06,
+      "loss": 0.1541,
+      "step": 6442
+    },
+    {
+      "epoch": 2.1516112873601605,
+      "grad_norm": 0.5741045672451875,
+      "learning_rate": 2.2432912272095227e-06,
+      "loss": 0.1637,
+      "step": 6443
+    },
+    {
+      "epoch": 2.1519452329270328,
+      "grad_norm": 0.5120871107028994,
+      "learning_rate": 2.2416701523288997e-06,
+      "loss": 0.1607,
+      "step": 6444
+    },
+    {
+      "epoch": 2.1522791784939055,
+      "grad_norm": 0.4437091895118815,
+      "learning_rate": 2.240049494126479e-06,
+      "loss": 0.1463,
+      "step": 6445
+    },
+    {
+      "epoch": 2.152613124060778,
+      "grad_norm": 0.4871523787141582,
+      "learning_rate": 2.238429252847079e-06,
+      "loss": 0.1627,
+      "step": 6446
+    },
+    {
+      "epoch": 2.1529470696276505,
+      "grad_norm": 0.49786741973353504,
+      "learning_rate": 2.2368094287354586e-06,
+      "loss": 0.162,
+      "step": 6447
+    },
+    {
+      "epoch": 2.153281015194523,
+      "grad_norm": 0.5147125301457852,
+      "learning_rate": 2.2351900220363083e-06,
+      "loss": 0.1634,
+      "step": 6448
+    },
+    {
+      "epoch": 2.153614960761396,
+      "grad_norm": 0.5019185934700248,
+      "learning_rate": 2.2335710329942613e-06,
+      "loss": 0.1608,
+      "step": 6449
+    },
+    {
+      "epoch": 2.1539489063282686,
+      "grad_norm": 0.5206661699341952,
+      "learning_rate": 2.2319524618538814e-06,
+      "loss": 0.1593,
+      "step": 6450
+    },
+    {
+      "epoch": 2.154282851895141,
+      "grad_norm": 0.5048674347182872,
+      "learning_rate": 2.2303343088596753e-06,
+      "loss": 0.1581,
+      "step": 6451
+    },
+    {
+      "epoch": 2.1546167974620136,
+      "grad_norm": 0.48713040989661927,
+      "learning_rate": 2.2287165742560828e-06,
+      "loss": 0.1578,
+      "step": 6452
+    },
+    {
+      "epoch": 2.1549507430288863,
+      "grad_norm": 0.49585282458056723,
+      "learning_rate": 2.227099258287485e-06,
+      "loss": 0.157,
+      "step": 6453
+    },
+    {
+      "epoch": 2.155284688595759,
+      "grad_norm": 0.5328333267249918,
+      "learning_rate": 2.2254823611981926e-06,
+      "loss": 0.1658,
+      "step": 6454
+    },
+    {
+      "epoch": 2.1556186341626313,
+      "grad_norm": 0.4931587019632607,
+      "learning_rate": 2.2238658832324593e-06,
+      "loss": 0.1547,
+      "step": 6455
+    },
+    {
+      "epoch": 2.155952579729504,
+      "grad_norm": 0.4506269112296143,
+      "learning_rate": 2.222249824634471e-06,
+      "loss": 0.1439,
+      "step": 6456
+    },
+    {
+      "epoch": 2.1562865252963768,
+      "grad_norm": 0.5495482012766242,
+      "learning_rate": 2.220634185648354e-06,
+      "loss": 0.1773,
+      "step": 6457
+    },
+    {
+      "epoch": 2.1566204708632495,
+      "grad_norm": 0.483556197943497,
+      "learning_rate": 2.2190189665181684e-06,
+      "loss": 0.1548,
+      "step": 6458
+    },
+    {
+      "epoch": 2.1569544164301218,
+      "grad_norm": 0.504658616614687,
+      "learning_rate": 2.2174041674879152e-06,
+      "loss": 0.1605,
+      "step": 6459
+    },
+    {
+      "epoch": 2.1572883619969945,
+      "grad_norm": 0.47666557877987864,
+      "learning_rate": 2.2157897888015247e-06,
+      "loss": 0.1489,
+      "step": 6460
+    },
+    {
+      "epoch": 2.157622307563867,
+      "grad_norm": 0.5125882112106799,
+      "learning_rate": 2.214175830702871e-06,
+      "loss": 0.1576,
+      "step": 6461
+    },
+    {
+      "epoch": 2.15795625313074,
+      "grad_norm": 0.47157304371108405,
+      "learning_rate": 2.2125622934357588e-06,
+      "loss": 0.1528,
+      "step": 6462
+    },
+    {
+      "epoch": 2.158290198697612,
+      "grad_norm": 0.5841106675855774,
+      "learning_rate": 2.210949177243933e-06,
+      "loss": 0.1491,
+      "step": 6463
+    },
+    {
+      "epoch": 2.158624144264485,
+      "grad_norm": 0.4945144495146848,
+      "learning_rate": 2.209336482371076e-06,
+      "loss": 0.161,
+      "step": 6464
+    },
+    {
+      "epoch": 2.1589580898313576,
+      "grad_norm": 0.5245463934666308,
+      "learning_rate": 2.2077242090608e-06,
+      "loss": 0.1638,
+      "step": 6465
+    },
+    {
+      "epoch": 2.15929203539823,
+      "grad_norm": 0.5189920938025878,
+      "learning_rate": 2.206112357556662e-06,
+      "loss": 0.1619,
+      "step": 6466
+    },
+    {
+      "epoch": 2.1596259809651026,
+      "grad_norm": 0.4710356006327489,
+      "learning_rate": 2.2045009281021486e-06,
+      "loss": 0.1492,
+      "step": 6467
+    },
+    {
+      "epoch": 2.1599599265319753,
+      "grad_norm": 0.5391720083168611,
+      "learning_rate": 2.202889920940685e-06,
+      "loss": 0.1673,
+      "step": 6468
+    },
+    {
+      "epoch": 2.160293872098848,
+      "grad_norm": 0.5300999004989326,
+      "learning_rate": 2.2012793363156337e-06,
+      "loss": 0.1662,
+      "step": 6469
+    },
+    {
+      "epoch": 2.1606278176657203,
+      "grad_norm": 0.545770302502262,
+      "learning_rate": 2.199669174470295e-06,
+      "loss": 0.177,
+      "step": 6470
+    },
+    {
+      "epoch": 2.160961763232593,
+      "grad_norm": 0.5544787542265645,
+      "learning_rate": 2.1980594356478977e-06,
+      "loss": 0.1717,
+      "step": 6471
+    },
+    {
+      "epoch": 2.1612957087994658,
+      "grad_norm": 0.47765190280150527,
+      "learning_rate": 2.196450120091617e-06,
+      "loss": 0.1532,
+      "step": 6472
+    },
+    {
+      "epoch": 2.1616296543663385,
+      "grad_norm": 0.48763213938191263,
+      "learning_rate": 2.194841228044554e-06,
+      "loss": 0.1495,
+      "step": 6473
+    },
+    {
+      "epoch": 2.1619635999332107,
+      "grad_norm": 0.47668729251630615,
+      "learning_rate": 2.1932327597497537e-06,
+      "loss": 0.1552,
+      "step": 6474
+    },
+    {
+      "epoch": 2.1622975455000835,
+      "grad_norm": 0.509947791860357,
+      "learning_rate": 2.1916247154501937e-06,
+      "loss": 0.1653,
+      "step": 6475
+    },
+    {
+      "epoch": 2.162631491066956,
+      "grad_norm": 0.5242288336740697,
+      "learning_rate": 2.190017095388789e-06,
+      "loss": 0.167,
+      "step": 6476
+    },
+    {
+      "epoch": 2.1629654366338285,
+      "grad_norm": 0.4842299004185554,
+      "learning_rate": 2.1884098998083867e-06,
+      "loss": 0.1544,
+      "step": 6477
+    },
+    {
+      "epoch": 2.163299382200701,
+      "grad_norm": 0.5657409384245479,
+      "learning_rate": 2.1868031289517773e-06,
+      "loss": 0.17,
+      "step": 6478
+    },
+    {
+      "epoch": 2.163633327767574,
+      "grad_norm": 0.5057846468272276,
+      "learning_rate": 2.1851967830616773e-06,
+      "loss": 0.1608,
+      "step": 6479
+    },
+    {
+      "epoch": 2.1639672733344466,
+      "grad_norm": 0.5045992465263685,
+      "learning_rate": 2.1835908623807462e-06,
+      "loss": 0.1587,
+      "step": 6480
+    },
+    {
+      "epoch": 2.164301218901319,
+      "grad_norm": 0.5337740972894388,
+      "learning_rate": 2.1819853671515774e-06,
+      "loss": 0.1659,
+      "step": 6481
+    },
+    {
+      "epoch": 2.1646351644681916,
+      "grad_norm": 0.5210496803965817,
+      "learning_rate": 2.180380297616702e-06,
+      "loss": 0.1591,
+      "step": 6482
+    },
+    {
+      "epoch": 2.1649691100350643,
+      "grad_norm": 0.5248088157313302,
+      "learning_rate": 2.178775654018581e-06,
+      "loss": 0.1624,
+      "step": 6483
+    },
+    {
+      "epoch": 2.165303055601937,
+      "grad_norm": 0.5262130819709666,
+      "learning_rate": 2.177171436599618e-06,
+      "loss": 0.1688,
+      "step": 6484
+    },
+    {
+      "epoch": 2.1656370011688093,
+      "grad_norm": 0.5361036965685855,
+      "learning_rate": 2.1755676456021454e-06,
+      "loss": 0.1605,
+      "step": 6485
+    },
+    {
+      "epoch": 2.165970946735682,
+      "grad_norm": 0.4938672738227419,
+      "learning_rate": 2.173964281268436e-06,
+      "loss": 0.1545,
+      "step": 6486
+    },
+    {
+      "epoch": 2.1663048923025547,
+      "grad_norm": 0.5366176125755003,
+      "learning_rate": 2.1723613438407e-06,
+      "loss": 0.1582,
+      "step": 6487
+    },
+    {
+      "epoch": 2.1666388378694275,
+      "grad_norm": 0.5677703842430186,
+      "learning_rate": 2.170758833561075e-06,
+      "loss": 0.1607,
+      "step": 6488
+    },
+    {
+      "epoch": 2.1669727834362997,
+      "grad_norm": 0.5831020985138208,
+      "learning_rate": 2.1691567506716433e-06,
+      "loss": 0.1692,
+      "step": 6489
+    },
+    {
+      "epoch": 2.1673067290031724,
+      "grad_norm": 0.5308398773677225,
+      "learning_rate": 2.1675550954144147e-06,
+      "loss": 0.1566,
+      "step": 6490
+    },
+    {
+      "epoch": 2.167640674570045,
+      "grad_norm": 0.47316973548002056,
+      "learning_rate": 2.1659538680313403e-06,
+      "loss": 0.1487,
+      "step": 6491
+    },
+    {
+      "epoch": 2.167974620136918,
+      "grad_norm": 0.5451574177675723,
+      "learning_rate": 2.1643530687643036e-06,
+      "loss": 0.1648,
+      "step": 6492
+    },
+    {
+      "epoch": 2.16830856570379,
+      "grad_norm": 0.559874991831784,
+      "learning_rate": 2.1627526978551265e-06,
+      "loss": 0.1522,
+      "step": 6493
+    },
+    {
+      "epoch": 2.168642511270663,
+      "grad_norm": 0.5653685743132745,
+      "learning_rate": 2.1611527555455604e-06,
+      "loss": 0.1705,
+      "step": 6494
+    },
+    {
+      "epoch": 2.1689764568375356,
+      "grad_norm": 0.4836390569478203,
+      "learning_rate": 2.159553242077298e-06,
+      "loss": 0.1603,
+      "step": 6495
+    },
+    {
+      "epoch": 2.169310402404408,
+      "grad_norm": 0.5285691812855493,
+      "learning_rate": 2.1579541576919624e-06,
+      "loss": 0.1624,
+      "step": 6496
+    },
+    {
+      "epoch": 2.1696443479712806,
+      "grad_norm": 0.5499127677294646,
+      "learning_rate": 2.1563555026311166e-06,
+      "loss": 0.1644,
+      "step": 6497
+    },
+    {
+      "epoch": 2.1699782935381533,
+      "grad_norm": 0.5111505875055791,
+      "learning_rate": 2.154757277136251e-06,
+      "loss": 0.1594,
+      "step": 6498
+    },
+    {
+      "epoch": 2.170312239105026,
+      "grad_norm": 0.4626953791952365,
+      "learning_rate": 2.153159481448805e-06,
+      "loss": 0.1402,
+      "step": 6499
+    },
+    {
+      "epoch": 2.1706461846718983,
+      "grad_norm": 0.5450349768292784,
+      "learning_rate": 2.1515621158101372e-06,
+      "loss": 0.1658,
+      "step": 6500
+    },
+    {
+      "epoch": 2.170980130238771,
+      "grad_norm": 0.4631639748506876,
+      "learning_rate": 2.1499651804615534e-06,
+      "loss": 0.1496,
+      "step": 6501
+    },
+    {
+      "epoch": 2.1713140758056437,
+      "grad_norm": 0.5041192964368789,
+      "learning_rate": 2.148368675644285e-06,
+      "loss": 0.1693,
+      "step": 6502
+    },
+    {
+      "epoch": 2.1716480213725164,
+      "grad_norm": 0.5020294331070693,
+      "learning_rate": 2.146772601599507e-06,
+      "loss": 0.1599,
+      "step": 6503
+    },
+    {
+      "epoch": 2.1719819669393887,
+      "grad_norm": 0.5070209282890003,
+      "learning_rate": 2.1451769585683196e-06,
+      "loss": 0.153,
+      "step": 6504
+    },
+    {
+      "epoch": 2.1723159125062614,
+      "grad_norm": 0.5004602093524981,
+      "learning_rate": 2.14358174679177e-06,
+      "loss": 0.1534,
+      "step": 6505
+    },
+    {
+      "epoch": 2.172649858073134,
+      "grad_norm": 0.4939325771882108,
+      "learning_rate": 2.1419869665108303e-06,
+      "loss": 0.1525,
+      "step": 6506
+    },
+    {
+      "epoch": 2.172983803640007,
+      "grad_norm": 0.5500117190899586,
+      "learning_rate": 2.140392617966412e-06,
+      "loss": 0.1658,
+      "step": 6507
+    },
+    {
+      "epoch": 2.173317749206879,
+      "grad_norm": 0.536575903530373,
+      "learning_rate": 2.1387987013993583e-06,
+      "loss": 0.1588,
+      "step": 6508
+    },
+    {
+      "epoch": 2.173651694773752,
+      "grad_norm": 0.49710428559563785,
+      "learning_rate": 2.137205217050452e-06,
+      "loss": 0.1607,
+      "step": 6509
+    },
+    {
+      "epoch": 2.1739856403406246,
+      "grad_norm": 0.46432099575708613,
+      "learning_rate": 2.135612165160404e-06,
+      "loss": 0.1476,
+      "step": 6510
+    },
+    {
+      "epoch": 2.1743195859074973,
+      "grad_norm": 0.5301977692094965,
+      "learning_rate": 2.1340195459698653e-06,
+      "loss": 0.1652,
+      "step": 6511
+    },
+    {
+      "epoch": 2.1746535314743696,
+      "grad_norm": 0.4947325177627652,
+      "learning_rate": 2.1324273597194223e-06,
+      "loss": 0.159,
+      "step": 6512
+    },
+    {
+      "epoch": 2.1749874770412423,
+      "grad_norm": 0.4855377166862449,
+      "learning_rate": 2.1308356066495893e-06,
+      "loss": 0.1516,
+      "step": 6513
+    },
+    {
+      "epoch": 2.175321422608115,
+      "grad_norm": 0.47997522170058093,
+      "learning_rate": 2.1292442870008213e-06,
+      "loss": 0.155,
+      "step": 6514
+    },
+    {
+      "epoch": 2.1756553681749873,
+      "grad_norm": 0.5003823360484102,
+      "learning_rate": 2.1276534010135053e-06,
+      "loss": 0.1468,
+      "step": 6515
+    },
+    {
+      "epoch": 2.17598931374186,
+      "grad_norm": 0.5190481271339713,
+      "learning_rate": 2.1260629489279662e-06,
+      "loss": 0.1649,
+      "step": 6516
+    },
+    {
+      "epoch": 2.1763232593087327,
+      "grad_norm": 0.4604055659642761,
+      "learning_rate": 2.1244729309844564e-06,
+      "loss": 0.1495,
+      "step": 6517
+    },
+    {
+      "epoch": 2.1766572048756054,
+      "grad_norm": 0.558870729021529,
+      "learning_rate": 2.1228833474231703e-06,
+      "loss": 0.1622,
+      "step": 6518
+    },
+    {
+      "epoch": 2.1769911504424777,
+      "grad_norm": 0.48777199157089124,
+      "learning_rate": 2.1212941984842295e-06,
+      "loss": 0.1514,
+      "step": 6519
+    },
+    {
+      "epoch": 2.1773250960093504,
+      "grad_norm": 0.49435336792484225,
+      "learning_rate": 2.1197054844076975e-06,
+      "loss": 0.1488,
+      "step": 6520
+    },
+    {
+      "epoch": 2.177659041576223,
+      "grad_norm": 0.5453124726321699,
+      "learning_rate": 2.118117205433563e-06,
+      "loss": 0.167,
+      "step": 6521
+    },
+    {
+      "epoch": 2.177992987143096,
+      "grad_norm": 0.4889668124604241,
+      "learning_rate": 2.1165293618017612e-06,
+      "loss": 0.1562,
+      "step": 6522
+    },
+    {
+      "epoch": 2.178326932709968,
+      "grad_norm": 0.5060322916216825,
+      "learning_rate": 2.1149419537521495e-06,
+      "loss": 0.1636,
+      "step": 6523
+    },
+    {
+      "epoch": 2.178660878276841,
+      "grad_norm": 0.572919121563999,
+      "learning_rate": 2.1133549815245273e-06,
+      "loss": 0.1706,
+      "step": 6524
+    },
+    {
+      "epoch": 2.1789948238437136,
+      "grad_norm": 0.5270190608334708,
+      "learning_rate": 2.1117684453586236e-06,
+      "loss": 0.1531,
+      "step": 6525
+    },
+    {
+      "epoch": 2.179328769410586,
+      "grad_norm": 0.5559601743872964,
+      "learning_rate": 2.110182345494105e-06,
+      "loss": 0.1719,
+      "step": 6526
+    },
+    {
+      "epoch": 2.1796627149774586,
+      "grad_norm": 0.5192431372887695,
+      "learning_rate": 2.1085966821705662e-06,
+      "loss": 0.1595,
+      "step": 6527
+    },
+    {
+      "epoch": 2.1799966605443313,
+      "grad_norm": 0.501123467668914,
+      "learning_rate": 2.1070114556275473e-06,
+      "loss": 0.1564,
+      "step": 6528
+    },
+    {
+      "epoch": 2.180330606111204,
+      "grad_norm": 0.4801142572023988,
+      "learning_rate": 2.1054266661045105e-06,
+      "loss": 0.145,
+      "step": 6529
+    },
+    {
+      "epoch": 2.1806645516780763,
+      "grad_norm": 0.5100950044547728,
+      "learning_rate": 2.103842313840859e-06,
+      "loss": 0.1618,
+      "step": 6530
+    },
+    {
+      "epoch": 2.180998497244949,
+      "grad_norm": 0.49876239967556707,
+      "learning_rate": 2.1022583990759265e-06,
+      "loss": 0.1579,
+      "step": 6531
+    },
+    {
+      "epoch": 2.1813324428118217,
+      "grad_norm": 0.5300376773967295,
+      "learning_rate": 2.1006749220489834e-06,
+      "loss": 0.1601,
+      "step": 6532
+    },
+    {
+      "epoch": 2.1816663883786944,
+      "grad_norm": 0.5139517459957516,
+      "learning_rate": 2.0990918829992307e-06,
+      "loss": 0.1638,
+      "step": 6533
+    },
+    {
+      "epoch": 2.1820003339455667,
+      "grad_norm": 0.4956864720621521,
+      "learning_rate": 2.097509282165806e-06,
+      "loss": 0.1678,
+      "step": 6534
+    },
+    {
+      "epoch": 2.1823342795124394,
+      "grad_norm": 0.5514089951815736,
+      "learning_rate": 2.0959271197877816e-06,
+      "loss": 0.1699,
+      "step": 6535
+    },
+    {
+      "epoch": 2.182668225079312,
+      "grad_norm": 0.529402674892044,
+      "learning_rate": 2.0943453961041587e-06,
+      "loss": 0.1583,
+      "step": 6536
+    },
+    {
+      "epoch": 2.183002170646185,
+      "grad_norm": 0.5027930243812394,
+      "learning_rate": 2.0927641113538764e-06,
+      "loss": 0.1558,
+      "step": 6537
+    },
+    {
+      "epoch": 2.183336116213057,
+      "grad_norm": 0.48247284060734513,
+      "learning_rate": 2.0911832657758086e-06,
+      "loss": 0.1504,
+      "step": 6538
+    },
+    {
+      "epoch": 2.18367006177993,
+      "grad_norm": 0.553224289771926,
+      "learning_rate": 2.089602859608757e-06,
+      "loss": 0.1612,
+      "step": 6539
+    },
+    {
+      "epoch": 2.1840040073468026,
+      "grad_norm": 0.5128855380439065,
+      "learning_rate": 2.088022893091462e-06,
+      "loss": 0.1688,
+      "step": 6540
+    },
+    {
+      "epoch": 2.1843379529136753,
+      "grad_norm": 0.4987615340223241,
+      "learning_rate": 2.086443366462598e-06,
+      "loss": 0.1584,
+      "step": 6541
+    },
+    {
+      "epoch": 2.1846718984805475,
+      "grad_norm": 0.45996958264592197,
+      "learning_rate": 2.084864279960768e-06,
+      "loss": 0.1414,
+      "step": 6542
+    },
+    {
+      "epoch": 2.1850058440474203,
+      "grad_norm": 0.5060506307251191,
+      "learning_rate": 2.0832856338245157e-06,
+      "loss": 0.1662,
+      "step": 6543
+    },
+    {
+      "epoch": 2.185339789614293,
+      "grad_norm": 0.5240449344599418,
+      "learning_rate": 2.0817074282923087e-06,
+      "loss": 0.17,
+      "step": 6544
+    },
+    {
+      "epoch": 2.1856737351811653,
+      "grad_norm": 0.48387681849094627,
+      "learning_rate": 2.080129663602557e-06,
+      "loss": 0.1501,
+      "step": 6545
+    },
+    {
+      "epoch": 2.186007680748038,
+      "grad_norm": 0.4837620146066662,
+      "learning_rate": 2.0785523399935996e-06,
+      "loss": 0.1548,
+      "step": 6546
+    },
+    {
+      "epoch": 2.1863416263149107,
+      "grad_norm": 0.5719904583600143,
+      "learning_rate": 2.076975457703712e-06,
+      "loss": 0.1661,
+      "step": 6547
+    },
+    {
+      "epoch": 2.1866755718817834,
+      "grad_norm": 0.5294141126429812,
+      "learning_rate": 2.0753990169710973e-06,
+      "loss": 0.1555,
+      "step": 6548
+    },
+    {
+      "epoch": 2.1870095174486557,
+      "grad_norm": 0.5038994789068824,
+      "learning_rate": 2.0738230180338993e-06,
+      "loss": 0.1579,
+      "step": 6549
+    },
+    {
+      "epoch": 2.1873434630155284,
+      "grad_norm": 0.5180801238233425,
+      "learning_rate": 2.0722474611301868e-06,
+      "loss": 0.1565,
+      "step": 6550
+    },
+    {
+      "epoch": 2.187677408582401,
+      "grad_norm": 0.517550029930567,
+      "learning_rate": 2.0706723464979687e-06,
+      "loss": 0.1631,
+      "step": 6551
+    },
+    {
+      "epoch": 2.188011354149274,
+      "grad_norm": 0.5632560056493515,
+      "learning_rate": 2.0690976743751844e-06,
+      "loss": 0.1737,
+      "step": 6552
+    },
+    {
+      "epoch": 2.188345299716146,
+      "grad_norm": 0.5057496168693179,
+      "learning_rate": 2.0675234449997085e-06,
+      "loss": 0.1558,
+      "step": 6553
+    },
+    {
+      "epoch": 2.188679245283019,
+      "grad_norm": 0.49701932491339085,
+      "learning_rate": 2.065949658609343e-06,
+      "loss": 0.1564,
+      "step": 6554
+    },
+    {
+      "epoch": 2.1890131908498915,
+      "grad_norm": 0.49770206149039964,
+      "learning_rate": 2.0643763154418304e-06,
+      "loss": 0.1545,
+      "step": 6555
+    },
+    {
+      "epoch": 2.1893471364167643,
+      "grad_norm": 0.5063235022252858,
+      "learning_rate": 2.06280341573484e-06,
+      "loss": 0.1538,
+      "step": 6556
+    },
+    {
+      "epoch": 2.1896810819836365,
+      "grad_norm": 0.4810621047438368,
+      "learning_rate": 2.0612309597259776e-06,
+      "loss": 0.1518,
+      "step": 6557
+    },
+    {
+      "epoch": 2.1900150275505093,
+      "grad_norm": 0.5329849457976076,
+      "learning_rate": 2.059658947652784e-06,
+      "loss": 0.1588,
+      "step": 6558
+    },
+    {
+      "epoch": 2.190348973117382,
+      "grad_norm": 0.4806564305506935,
+      "learning_rate": 2.058087379752725e-06,
+      "loss": 0.1555,
+      "step": 6559
+    },
+    {
+      "epoch": 2.1906829186842547,
+      "grad_norm": 0.5063446638868672,
+      "learning_rate": 2.056516256263208e-06,
+      "loss": 0.1567,
+      "step": 6560
+    },
+    {
+      "epoch": 2.191016864251127,
+      "grad_norm": 0.5230077420968994,
+      "learning_rate": 2.0549455774215705e-06,
+      "loss": 0.156,
+      "step": 6561
+    },
+    {
+      "epoch": 2.1913508098179997,
+      "grad_norm": 0.49869886677540565,
+      "learning_rate": 2.0533753434650784e-06,
+      "loss": 0.1652,
+      "step": 6562
+    },
+    {
+      "epoch": 2.1916847553848724,
+      "grad_norm": 0.5172552699731698,
+      "learning_rate": 2.0518055546309362e-06,
+      "loss": 0.1523,
+      "step": 6563
+    },
+    {
+      "epoch": 2.1920187009517447,
+      "grad_norm": 0.4894924237454037,
+      "learning_rate": 2.0502362111562806e-06,
+      "loss": 0.1578,
+      "step": 6564
+    },
+    {
+      "epoch": 2.1923526465186174,
+      "grad_norm": 0.5085326999824232,
+      "learning_rate": 2.048667313278176e-06,
+      "loss": 0.1537,
+      "step": 6565
+    },
+    {
+      "epoch": 2.19268659208549,
+      "grad_norm": 0.49822549747118766,
+      "learning_rate": 2.0470988612336264e-06,
+      "loss": 0.1536,
+      "step": 6566
+    },
+    {
+      "epoch": 2.193020537652363,
+      "grad_norm": 0.5148739261403503,
+      "learning_rate": 2.045530855259561e-06,
+      "loss": 0.1637,
+      "step": 6567
+    },
+    {
+      "epoch": 2.193354483219235,
+      "grad_norm": 0.5168387604417966,
+      "learning_rate": 2.043963295592848e-06,
+      "loss": 0.166,
+      "step": 6568
+    },
+    {
+      "epoch": 2.193688428786108,
+      "grad_norm": 0.5076478055233222,
+      "learning_rate": 2.042396182470285e-06,
+      "loss": 0.1676,
+      "step": 6569
+    },
+    {
+      "epoch": 2.1940223743529805,
+      "grad_norm": 0.4850803716080977,
+      "learning_rate": 2.040829516128605e-06,
+      "loss": 0.1507,
+      "step": 6570
+    },
+    {
+      "epoch": 2.1943563199198532,
+      "grad_norm": 0.5096244454579222,
+      "learning_rate": 2.0392632968044686e-06,
+      "loss": 0.1642,
+      "step": 6571
+    },
+    {
+      "epoch": 2.1946902654867255,
+      "grad_norm": 0.4942711483631219,
+      "learning_rate": 2.0376975247344736e-06,
+      "loss": 0.1559,
+      "step": 6572
+    },
+    {
+      "epoch": 2.1950242110535982,
+      "grad_norm": 0.5207629645400701,
+      "learning_rate": 2.0361322001551466e-06,
+      "loss": 0.1721,
+      "step": 6573
+    },
+    {
+      "epoch": 2.195358156620471,
+      "grad_norm": 0.45191543742406315,
+      "learning_rate": 2.034567323302949e-06,
+      "loss": 0.1337,
+      "step": 6574
+    },
+    {
+      "epoch": 2.1956921021873432,
+      "grad_norm": 0.48335848242486656,
+      "learning_rate": 2.0330028944142736e-06,
+      "loss": 0.1551,
+      "step": 6575
+    },
+    {
+      "epoch": 2.196026047754216,
+      "grad_norm": 0.4956348807447585,
+      "learning_rate": 2.031438913725448e-06,
+      "loss": 0.1504,
+      "step": 6576
+    },
+    {
+      "epoch": 2.1963599933210887,
+      "grad_norm": 0.5681269369080285,
+      "learning_rate": 2.0298753814727267e-06,
+      "loss": 0.1531,
+      "step": 6577
+    },
+    {
+      "epoch": 2.1966939388879614,
+      "grad_norm": 0.5291428022534318,
+      "learning_rate": 2.028312297892303e-06,
+      "loss": 0.1621,
+      "step": 6578
+    },
+    {
+      "epoch": 2.1970278844548337,
+      "grad_norm": 0.5239390275772546,
+      "learning_rate": 2.0267496632202953e-06,
+      "loss": 0.166,
+      "step": 6579
+    },
+    {
+      "epoch": 2.1973618300217064,
+      "grad_norm": 0.5409865518744863,
+      "learning_rate": 2.0251874776927598e-06,
+      "loss": 0.1643,
+      "step": 6580
+    },
+    {
+      "epoch": 2.197695775588579,
+      "grad_norm": 0.5372965064486103,
+      "learning_rate": 2.0236257415456833e-06,
+      "loss": 0.1646,
+      "step": 6581
+    },
+    {
+      "epoch": 2.198029721155452,
+      "grad_norm": 0.509285554214082,
+      "learning_rate": 2.022064455014986e-06,
+      "loss": 0.1623,
+      "step": 6582
+    },
+    {
+      "epoch": 2.198363666722324,
+      "grad_norm": 0.757210617776752,
+      "learning_rate": 2.0205036183365145e-06,
+      "loss": 0.1749,
+      "step": 6583
+    },
+    {
+      "epoch": 2.198697612289197,
+      "grad_norm": 0.4667375146591432,
+      "learning_rate": 2.018943231746056e-06,
+      "loss": 0.1455,
+      "step": 6584
+    },
+    {
+      "epoch": 2.1990315578560695,
+      "grad_norm": 0.5693617648213951,
+      "learning_rate": 2.0173832954793216e-06,
+      "loss": 0.164,
+      "step": 6585
+    },
+    {
+      "epoch": 2.1993655034229422,
+      "grad_norm": 0.49312895306196874,
+      "learning_rate": 2.0158238097719597e-06,
+      "loss": 0.1543,
+      "step": 6586
+    },
+    {
+      "epoch": 2.1996994489898145,
+      "grad_norm": 0.5064476054065244,
+      "learning_rate": 2.0142647748595502e-06,
+      "loss": 0.1651,
+      "step": 6587
+    },
+    {
+      "epoch": 2.2000333945566872,
+      "grad_norm": 0.5332088995867316,
+      "learning_rate": 2.0127061909776e-06,
+      "loss": 0.1661,
+      "step": 6588
+    },
+    {
+      "epoch": 2.20036734012356,
+      "grad_norm": 0.47982702274764294,
+      "learning_rate": 2.0111480583615566e-06,
+      "loss": 0.1466,
+      "step": 6589
+    },
+    {
+      "epoch": 2.2007012856904327,
+      "grad_norm": 0.4910497844263571,
+      "learning_rate": 2.00959037724679e-06,
+      "loss": 0.1515,
+      "step": 6590
+    },
+    {
+      "epoch": 2.201035231257305,
+      "grad_norm": 0.5585938097349513,
+      "learning_rate": 2.0080331478686087e-06,
+      "loss": 0.1732,
+      "step": 6591
+    },
+    {
+      "epoch": 2.2013691768241777,
+      "grad_norm": 0.4896922496577059,
+      "learning_rate": 2.006476370462247e-06,
+      "loss": 0.1557,
+      "step": 6592
+    },
+    {
+      "epoch": 2.2017031223910504,
+      "grad_norm": 0.4576228652187881,
+      "learning_rate": 2.0049200452628803e-06,
+      "loss": 0.1437,
+      "step": 6593
+    },
+    {
+      "epoch": 2.2020370679579226,
+      "grad_norm": 0.4734216591875249,
+      "learning_rate": 2.0033641725056048e-06,
+      "loss": 0.1487,
+      "step": 6594
+    },
+    {
+      "epoch": 2.2023710135247954,
+      "grad_norm": 0.5060810259961187,
+      "learning_rate": 2.001808752425457e-06,
+      "loss": 0.1471,
+      "step": 6595
+    },
+    {
+      "epoch": 2.202704959091668,
+      "grad_norm": 0.48329648438901496,
+      "learning_rate": 2.000253785257398e-06,
+      "loss": 0.1525,
+      "step": 6596
+    },
+    {
+      "epoch": 2.203038904658541,
+      "grad_norm": 0.5357710590885335,
+      "learning_rate": 1.998699271236326e-06,
+      "loss": 0.1612,
+      "step": 6597
+    },
+    {
+      "epoch": 2.203372850225413,
+      "grad_norm": 0.525150917969299,
+      "learning_rate": 1.997145210597068e-06,
+      "loss": 0.1697,
+      "step": 6598
+    },
+    {
+      "epoch": 2.203706795792286,
+      "grad_norm": 0.47723396449168365,
+      "learning_rate": 1.9955916035743855e-06,
+      "loss": 0.1512,
+      "step": 6599
+    },
+    {
+      "epoch": 2.2040407413591585,
+      "grad_norm": 0.47291493505957904,
+      "learning_rate": 1.9940384504029647e-06,
+      "loss": 0.1562,
+      "step": 6600
+    },
+    {
+      "epoch": 2.2043746869260312,
+      "grad_norm": 0.5084750289652352,
+      "learning_rate": 1.9924857513174324e-06,
+      "loss": 0.1588,
+      "step": 6601
+    },
+    {
+      "epoch": 2.2047086324929035,
+      "grad_norm": 0.5015177899122328,
+      "learning_rate": 1.990933506552337e-06,
+      "loss": 0.1591,
+      "step": 6602
+    },
+    {
+      "epoch": 2.205042578059776,
+      "grad_norm": 0.5140862563797094,
+      "learning_rate": 1.989381716342167e-06,
+      "loss": 0.1669,
+      "step": 6603
+    },
+    {
+      "epoch": 2.205376523626649,
+      "grad_norm": 0.5335787711446098,
+      "learning_rate": 1.9878303809213367e-06,
+      "loss": 0.1512,
+      "step": 6604
+    },
+    {
+      "epoch": 2.2057104691935217,
+      "grad_norm": 0.5424751411330692,
+      "learning_rate": 1.986279500524197e-06,
+      "loss": 0.1657,
+      "step": 6605
+    },
+    {
+      "epoch": 2.206044414760394,
+      "grad_norm": 0.5056230951340244,
+      "learning_rate": 1.984729075385022e-06,
+      "loss": 0.1596,
+      "step": 6606
+    },
+    {
+      "epoch": 2.2063783603272666,
+      "grad_norm": 0.47996255878953,
+      "learning_rate": 1.983179105738026e-06,
+      "loss": 0.1582,
+      "step": 6607
+    },
+    {
+      "epoch": 2.2067123058941394,
+      "grad_norm": 0.4911356123723574,
+      "learning_rate": 1.9816295918173462e-06,
+      "loss": 0.1534,
+      "step": 6608
+    },
+    {
+      "epoch": 2.207046251461012,
+      "grad_norm": 0.465393956829766,
+      "learning_rate": 1.9800805338570562e-06,
+      "loss": 0.1601,
+      "step": 6609
+    },
+    {
+      "epoch": 2.2073801970278843,
+      "grad_norm": 0.4979367096913454,
+      "learning_rate": 1.9785319320911623e-06,
+      "loss": 0.1463,
+      "step": 6610
+    },
+    {
+      "epoch": 2.207714142594757,
+      "grad_norm": 0.486772494636018,
+      "learning_rate": 1.9769837867535948e-06,
+      "loss": 0.1557,
+      "step": 6611
+    },
+    {
+      "epoch": 2.20804808816163,
+      "grad_norm": 0.487323805440558,
+      "learning_rate": 1.9754360980782227e-06,
+      "loss": 0.1563,
+      "step": 6612
+    },
+    {
+      "epoch": 2.208382033728502,
+      "grad_norm": 0.5575220348139566,
+      "learning_rate": 1.973888866298839e-06,
+      "loss": 0.1676,
+      "step": 6613
+    },
+    {
+      "epoch": 2.2087159792953748,
+      "grad_norm": 0.537349301156599,
+      "learning_rate": 1.972342091649176e-06,
+      "loss": 0.1667,
+      "step": 6614
+    },
+    {
+      "epoch": 2.2090499248622475,
+      "grad_norm": 0.49476253066855386,
+      "learning_rate": 1.9707957743628854e-06,
+      "loss": 0.1534,
+      "step": 6615
+    },
+    {
+      "epoch": 2.20938387042912,
+      "grad_norm": 0.5405416629754679,
+      "learning_rate": 1.9692499146735646e-06,
+      "loss": 0.1642,
+      "step": 6616
+    },
+    {
+      "epoch": 2.2097178159959925,
+      "grad_norm": 0.5165616973685996,
+      "learning_rate": 1.967704512814728e-06,
+      "loss": 0.164,
+      "step": 6617
+    },
+    {
+      "epoch": 2.210051761562865,
+      "grad_norm": 0.524024480111212,
+      "learning_rate": 1.966159569019831e-06,
+      "loss": 0.1629,
+      "step": 6618
+    },
+    {
+      "epoch": 2.210385707129738,
+      "grad_norm": 0.528009150373322,
+      "learning_rate": 1.9646150835222517e-06,
+      "loss": 0.1641,
+      "step": 6619
+    },
+    {
+      "epoch": 2.2107196526966106,
+      "grad_norm": 0.48772901021730125,
+      "learning_rate": 1.9630710565553063e-06,
+      "loss": 0.1532,
+      "step": 6620
+    },
+    {
+      "epoch": 2.211053598263483,
+      "grad_norm": 0.4371984282049327,
+      "learning_rate": 1.9615274883522327e-06,
+      "loss": 0.1386,
+      "step": 6621
+    },
+    {
+      "epoch": 2.2113875438303556,
+      "grad_norm": 0.5506478629606789,
+      "learning_rate": 1.9599843791462123e-06,
+      "loss": 0.1703,
+      "step": 6622
+    },
+    {
+      "epoch": 2.2117214893972283,
+      "grad_norm": 0.5183244278342396,
+      "learning_rate": 1.958441729170345e-06,
+      "loss": 0.1558,
+      "step": 6623
+    },
+    {
+      "epoch": 2.2120554349641006,
+      "grad_norm": 0.4653062989884138,
+      "learning_rate": 1.9568995386576695e-06,
+      "loss": 0.148,
+      "step": 6624
+    },
+    {
+      "epoch": 2.2123893805309733,
+      "grad_norm": 0.49138499728836343,
+      "learning_rate": 1.9553578078411476e-06,
+      "loss": 0.1541,
+      "step": 6625
+    },
+    {
+      "epoch": 2.212723326097846,
+      "grad_norm": 0.5254863963640038,
+      "learning_rate": 1.953816536953681e-06,
+      "loss": 0.1546,
+      "step": 6626
+    },
+    {
+      "epoch": 2.2130572716647188,
+      "grad_norm": 0.5206964772367224,
+      "learning_rate": 1.95227572622809e-06,
+      "loss": 0.1621,
+      "step": 6627
+    },
+    {
+      "epoch": 2.213391217231591,
+      "grad_norm": 0.563970641509512,
+      "learning_rate": 1.95073537589714e-06,
+      "loss": 0.1722,
+      "step": 6628
+    },
+    {
+      "epoch": 2.2137251627984638,
+      "grad_norm": 0.4877735248249902,
+      "learning_rate": 1.949195486193514e-06,
+      "loss": 0.1466,
+      "step": 6629
+    },
+    {
+      "epoch": 2.2140591083653365,
+      "grad_norm": 0.5032630664511833,
+      "learning_rate": 1.9476560573498332e-06,
+      "loss": 0.1651,
+      "step": 6630
+    },
+    {
+      "epoch": 2.214393053932209,
+      "grad_norm": 0.5375345394822361,
+      "learning_rate": 1.946117089598644e-06,
+      "loss": 0.1737,
+      "step": 6631
+    },
+    {
+      "epoch": 2.2147269994990815,
+      "grad_norm": 0.4730562669425479,
+      "learning_rate": 1.9445785831724274e-06,
+      "loss": 0.1511,
+      "step": 6632
+    },
+    {
+      "epoch": 2.215060945065954,
+      "grad_norm": 0.5274486320082097,
+      "learning_rate": 1.943040538303591e-06,
+      "loss": 0.1596,
+      "step": 6633
+    },
+    {
+      "epoch": 2.215394890632827,
+      "grad_norm": 0.5020720245794152,
+      "learning_rate": 1.9415029552244758e-06,
+      "loss": 0.1585,
+      "step": 6634
+    },
+    {
+      "epoch": 2.2157288361996996,
+      "grad_norm": 0.5225443854090014,
+      "learning_rate": 1.939965834167354e-06,
+      "loss": 0.1642,
+      "step": 6635
+    },
+    {
+      "epoch": 2.216062781766572,
+      "grad_norm": 0.5246272127996099,
+      "learning_rate": 1.9384291753644215e-06,
+      "loss": 0.1531,
+      "step": 6636
+    },
+    {
+      "epoch": 2.2163967273334446,
+      "grad_norm": 0.4890983704136873,
+      "learning_rate": 1.9368929790478126e-06,
+      "loss": 0.155,
+      "step": 6637
+    },
+    {
+      "epoch": 2.2167306729003173,
+      "grad_norm": 0.5091358082023,
+      "learning_rate": 1.935357245449583e-06,
+      "loss": 0.1508,
+      "step": 6638
+    },
+    {
+      "epoch": 2.21706461846719,
+      "grad_norm": 0.5688713170703724,
+      "learning_rate": 1.9338219748017297e-06,
+      "loss": 0.1745,
+      "step": 6639
+    },
+    {
+      "epoch": 2.2173985640340623,
+      "grad_norm": 0.5001120410713611,
+      "learning_rate": 1.932287167336168e-06,
+      "loss": 0.1617,
+      "step": 6640
+    },
+    {
+      "epoch": 2.217732509600935,
+      "grad_norm": 0.5223274403848096,
+      "learning_rate": 1.9307528232847533e-06,
+      "loss": 0.1588,
+      "step": 6641
+    },
+    {
+      "epoch": 2.2180664551678078,
+      "grad_norm": 0.49231056426484826,
+      "learning_rate": 1.9292189428792617e-06,
+      "loss": 0.1453,
+      "step": 6642
+    },
+    {
+      "epoch": 2.21840040073468,
+      "grad_norm": 0.5447393922671808,
+      "learning_rate": 1.927685526351408e-06,
+      "loss": 0.1614,
+      "step": 6643
+    },
+    {
+      "epoch": 2.2187343463015528,
+      "grad_norm": 0.5130642995180773,
+      "learning_rate": 1.9261525739328273e-06,
+      "loss": 0.1539,
+      "step": 6644
+    },
+    {
+      "epoch": 2.2190682918684255,
+      "grad_norm": 0.4673620218126964,
+      "learning_rate": 1.924620085855097e-06,
+      "loss": 0.1416,
+      "step": 6645
+    },
+    {
+      "epoch": 2.219402237435298,
+      "grad_norm": 0.5239963893912938,
+      "learning_rate": 1.923088062349713e-06,
+      "loss": 0.1574,
+      "step": 6646
+    },
+    {
+      "epoch": 2.2197361830021705,
+      "grad_norm": 0.5192341161140694,
+      "learning_rate": 1.9215565036481083e-06,
+      "loss": 0.1606,
+      "step": 6647
+    },
+    {
+      "epoch": 2.220070128569043,
+      "grad_norm": 0.520663322092088,
+      "learning_rate": 1.920025409981639e-06,
+      "loss": 0.1663,
+      "step": 6648
+    },
+    {
+      "epoch": 2.220404074135916,
+      "grad_norm": 0.5032776512637247,
+      "learning_rate": 1.918494781581599e-06,
+      "loss": 0.158,
+      "step": 6649
+    },
+    {
+      "epoch": 2.2207380197027886,
+      "grad_norm": 0.4713101365681985,
+      "learning_rate": 1.9169646186792025e-06,
+      "loss": 0.1467,
+      "step": 6650
+    },
+    {
+      "epoch": 2.221071965269661,
+      "grad_norm": 0.46675389053456917,
+      "learning_rate": 1.9154349215056052e-06,
+      "loss": 0.1546,
+      "step": 6651
+    },
+    {
+      "epoch": 2.2214059108365336,
+      "grad_norm": 0.5200622300655069,
+      "learning_rate": 1.9139056902918805e-06,
+      "loss": 0.1665,
+      "step": 6652
+    },
+    {
+      "epoch": 2.2217398564034063,
+      "grad_norm": 0.48668102810544783,
+      "learning_rate": 1.912376925269041e-06,
+      "loss": 0.1544,
+      "step": 6653
+    },
+    {
+      "epoch": 2.222073801970279,
+      "grad_norm": 0.5121790498856402,
+      "learning_rate": 1.910848626668021e-06,
+      "loss": 0.1533,
+      "step": 6654
+    },
+    {
+      "epoch": 2.2224077475371513,
+      "grad_norm": 0.4908382616540141,
+      "learning_rate": 1.9093207947196908e-06,
+      "loss": 0.1604,
+      "step": 6655
+    },
+    {
+      "epoch": 2.222741693104024,
+      "grad_norm": 0.4999744558438863,
+      "learning_rate": 1.9077934296548445e-06,
+      "loss": 0.1542,
+      "step": 6656
+    },
+    {
+      "epoch": 2.2230756386708967,
+      "grad_norm": 0.5256425135871804,
+      "learning_rate": 1.9062665317042106e-06,
+      "loss": 0.1519,
+      "step": 6657
+    },
+    {
+      "epoch": 2.2234095842377695,
+      "grad_norm": 0.4849691863222047,
+      "learning_rate": 1.9047401010984456e-06,
+      "loss": 0.1504,
+      "step": 6658
+    },
+    {
+      "epoch": 2.2237435298046417,
+      "grad_norm": 0.4791081942937301,
+      "learning_rate": 1.9032141380681329e-06,
+      "loss": 0.1499,
+      "step": 6659
+    },
+    {
+      "epoch": 2.2240774753715145,
+      "grad_norm": 0.5202987778444312,
+      "learning_rate": 1.9016886428437893e-06,
+      "loss": 0.1667,
+      "step": 6660
+    },
+    {
+      "epoch": 2.224411420938387,
+      "grad_norm": 0.4901000495769656,
+      "learning_rate": 1.9001636156558562e-06,
+      "loss": 0.1565,
+      "step": 6661
+    },
+    {
+      "epoch": 2.2247453665052594,
+      "grad_norm": 0.4823202571420205,
+      "learning_rate": 1.8986390567347085e-06,
+      "loss": 0.1529,
+      "step": 6662
+    },
+    {
+      "epoch": 2.225079312072132,
+      "grad_norm": 0.5025149193666405,
+      "learning_rate": 1.8971149663106482e-06,
+      "loss": 0.1572,
+      "step": 6663
+    },
+    {
+      "epoch": 2.225413257639005,
+      "grad_norm": 0.4810470147277487,
+      "learning_rate": 1.8955913446139096e-06,
+      "loss": 0.1569,
+      "step": 6664
+    },
+    {
+      "epoch": 2.2257472032058776,
+      "grad_norm": 0.46343686657363625,
+      "learning_rate": 1.8940681918746495e-06,
+      "loss": 0.1452,
+      "step": 6665
+    },
+    {
+      "epoch": 2.22608114877275,
+      "grad_norm": 0.47263048459042295,
+      "learning_rate": 1.8925455083229622e-06,
+      "loss": 0.1542,
+      "step": 6666
+    },
+    {
+      "epoch": 2.2264150943396226,
+      "grad_norm": 0.48227650724640997,
+      "learning_rate": 1.891023294188863e-06,
+      "loss": 0.1566,
+      "step": 6667
+    },
+    {
+      "epoch": 2.2267490399064953,
+      "grad_norm": 0.5703992870293059,
+      "learning_rate": 1.8895015497023022e-06,
+      "loss": 0.1628,
+      "step": 6668
+    },
+    {
+      "epoch": 2.227082985473368,
+      "grad_norm": 0.5171876577597354,
+      "learning_rate": 1.8879802750931574e-06,
+      "loss": 0.1594,
+      "step": 6669
+    },
+    {
+      "epoch": 2.2274169310402403,
+      "grad_norm": 0.4918589655419636,
+      "learning_rate": 1.886459470591237e-06,
+      "loss": 0.1492,
+      "step": 6670
+    },
+    {
+      "epoch": 2.227750876607113,
+      "grad_norm": 0.535274808727,
+      "learning_rate": 1.8849391364262721e-06,
+      "loss": 0.1616,
+      "step": 6671
+    },
+    {
+      "epoch": 2.2280848221739857,
+      "grad_norm": 0.4918788564052449,
+      "learning_rate": 1.883419272827931e-06,
+      "loss": 0.1544,
+      "step": 6672
+    },
+    {
+      "epoch": 2.228418767740858,
+      "grad_norm": 0.5305209595696436,
+      "learning_rate": 1.881899880025802e-06,
+      "loss": 0.1537,
+      "step": 6673
+    },
+    {
+      "epoch": 2.2287527133077307,
+      "grad_norm": 0.5543131551389201,
+      "learning_rate": 1.8803809582494143e-06,
+      "loss": 0.1598,
+      "step": 6674
+    },
+    {
+      "epoch": 2.2290866588746034,
+      "grad_norm": 0.5080012549587938,
+      "learning_rate": 1.878862507728213e-06,
+      "loss": 0.1602,
+      "step": 6675
+    },
+    {
+      "epoch": 2.229420604441476,
+      "grad_norm": 0.48004398539465265,
+      "learning_rate": 1.877344528691582e-06,
+      "loss": 0.147,
+      "step": 6676
+    },
+    {
+      "epoch": 2.2297545500083484,
+      "grad_norm": 0.5318763956775434,
+      "learning_rate": 1.8758270213688263e-06,
+      "loss": 0.1656,
+      "step": 6677
+    },
+    {
+      "epoch": 2.230088495575221,
+      "grad_norm": 0.527960686138271,
+      "learning_rate": 1.8743099859891866e-06,
+      "loss": 0.1584,
+      "step": 6678
+    },
+    {
+      "epoch": 2.230422441142094,
+      "grad_norm": 0.4634772222288885,
+      "learning_rate": 1.8727934227818255e-06,
+      "loss": 0.1459,
+      "step": 6679
+    },
+    {
+      "epoch": 2.2307563867089666,
+      "grad_norm": 0.5262489510400422,
+      "learning_rate": 1.8712773319758398e-06,
+      "loss": 0.1623,
+      "step": 6680
+    },
+    {
+      "epoch": 2.231090332275839,
+      "grad_norm": 0.49489280932867763,
+      "learning_rate": 1.8697617138002545e-06,
+      "loss": 0.1478,
+      "step": 6681
+    },
+    {
+      "epoch": 2.2314242778427116,
+      "grad_norm": 0.5031062159733811,
+      "learning_rate": 1.8682465684840178e-06,
+      "loss": 0.1564,
+      "step": 6682
+    },
+    {
+      "epoch": 2.2317582234095843,
+      "grad_norm": 0.4953788465486049,
+      "learning_rate": 1.8667318962560137e-06,
+      "loss": 0.1626,
+      "step": 6683
+    },
+    {
+      "epoch": 2.232092168976457,
+      "grad_norm": 0.5329151547726576,
+      "learning_rate": 1.865217697345048e-06,
+      "loss": 0.1674,
+      "step": 6684
+    },
+    {
+      "epoch": 2.2324261145433293,
+      "grad_norm": 0.5510845744808649,
+      "learning_rate": 1.86370397197986e-06,
+      "loss": 0.1663,
+      "step": 6685
+    },
+    {
+      "epoch": 2.232760060110202,
+      "grad_norm": 0.5461167706937765,
+      "learning_rate": 1.8621907203891159e-06,
+      "loss": 0.1704,
+      "step": 6686
+    },
+    {
+      "epoch": 2.2330940056770747,
+      "grad_norm": 0.5425557341551985,
+      "learning_rate": 1.8606779428014116e-06,
+      "loss": 0.1778,
+      "step": 6687
+    },
+    {
+      "epoch": 2.2334279512439474,
+      "grad_norm": 0.5394947897686756,
+      "learning_rate": 1.8591656394452667e-06,
+      "loss": 0.1649,
+      "step": 6688
+    },
+    {
+      "epoch": 2.2337618968108197,
+      "grad_norm": 0.4755957408400034,
+      "learning_rate": 1.8576538105491359e-06,
+      "loss": 0.1519,
+      "step": 6689
+    },
+    {
+      "epoch": 2.2340958423776924,
+      "grad_norm": 0.49279914369472505,
+      "learning_rate": 1.8561424563413949e-06,
+      "loss": 0.1529,
+      "step": 6690
+    },
+    {
+      "epoch": 2.234429787944565,
+      "grad_norm": 0.495172322606566,
+      "learning_rate": 1.8546315770503537e-06,
+      "loss": 0.153,
+      "step": 6691
+    },
+    {
+      "epoch": 2.2347637335114374,
+      "grad_norm": 0.4707511082921486,
+      "learning_rate": 1.8531211729042486e-06,
+      "loss": 0.1505,
+      "step": 6692
+    },
+    {
+      "epoch": 2.23509767907831,
+      "grad_norm": 0.5004010993757639,
+      "learning_rate": 1.8516112441312451e-06,
+      "loss": 0.1559,
+      "step": 6693
+    },
+    {
+      "epoch": 2.235431624645183,
+      "grad_norm": 0.5388923994203958,
+      "learning_rate": 1.8501017909594327e-06,
+      "loss": 0.1623,
+      "step": 6694
+    },
+    {
+      "epoch": 2.2357655702120556,
+      "grad_norm": 0.5725007414524714,
+      "learning_rate": 1.8485928136168353e-06,
+      "loss": 0.1733,
+      "step": 6695
+    },
+    {
+      "epoch": 2.236099515778928,
+      "grad_norm": 0.5413884294057945,
+      "learning_rate": 1.8470843123313982e-06,
+      "loss": 0.1526,
+      "step": 6696
+    },
+    {
+      "epoch": 2.2364334613458006,
+      "grad_norm": 0.5215501259534671,
+      "learning_rate": 1.8455762873309995e-06,
+      "loss": 0.1599,
+      "step": 6697
+    },
+    {
+      "epoch": 2.2367674069126733,
+      "grad_norm": 0.5434730278673744,
+      "learning_rate": 1.844068738843446e-06,
+      "loss": 0.1697,
+      "step": 6698
+    },
+    {
+      "epoch": 2.237101352479546,
+      "grad_norm": 0.8447780480576517,
+      "learning_rate": 1.8425616670964702e-06,
+      "loss": 0.1591,
+      "step": 6699
+    },
+    {
+      "epoch": 2.2374352980464183,
+      "grad_norm": 0.4889453910112141,
+      "learning_rate": 1.8410550723177306e-06,
+      "loss": 0.148,
+      "step": 6700
+    },
+    {
+      "epoch": 2.237769243613291,
+      "grad_norm": 0.5348087835055323,
+      "learning_rate": 1.8395489547348193e-06,
+      "loss": 0.1586,
+      "step": 6701
+    },
+    {
+      "epoch": 2.2381031891801637,
+      "grad_norm": 0.509290022703682,
+      "learning_rate": 1.8380433145752502e-06,
+      "loss": 0.159,
+      "step": 6702
+    },
+    {
+      "epoch": 2.2384371347470364,
+      "grad_norm": 0.5065409444970693,
+      "learning_rate": 1.8365381520664695e-06,
+      "loss": 0.1551,
+      "step": 6703
+    },
+    {
+      "epoch": 2.2387710803139087,
+      "grad_norm": 0.5002223281275832,
+      "learning_rate": 1.8350334674358505e-06,
+      "loss": 0.1524,
+      "step": 6704
+    },
+    {
+      "epoch": 2.2391050258807814,
+      "grad_norm": 0.521909205621699,
+      "learning_rate": 1.8335292609106914e-06,
+      "loss": 0.1507,
+      "step": 6705
+    },
+    {
+      "epoch": 2.239438971447654,
+      "grad_norm": 0.5051898762790804,
+      "learning_rate": 1.8320255327182224e-06,
+      "loss": 0.1597,
+      "step": 6706
+    },
+    {
+      "epoch": 2.239772917014527,
+      "grad_norm": 0.46375293585113725,
+      "learning_rate": 1.8305222830855973e-06,
+      "loss": 0.1426,
+      "step": 6707
+    },
+    {
+      "epoch": 2.240106862581399,
+      "grad_norm": 0.45694266168008607,
+      "learning_rate": 1.8290195122399007e-06,
+      "loss": 0.141,
+      "step": 6708
+    },
+    {
+      "epoch": 2.240440808148272,
+      "grad_norm": 0.5090966358483844,
+      "learning_rate": 1.8275172204081437e-06,
+      "loss": 0.1543,
+      "step": 6709
+    },
+    {
+      "epoch": 2.2407747537151446,
+      "grad_norm": 0.5076709222116111,
+      "learning_rate": 1.826015407817267e-06,
+      "loss": 0.1548,
+      "step": 6710
+    },
+    {
+      "epoch": 2.241108699282017,
+      "grad_norm": 0.553645802415818,
+      "learning_rate": 1.8245140746941336e-06,
+      "loss": 0.1673,
+      "step": 6711
+    },
+    {
+      "epoch": 2.2414426448488896,
+      "grad_norm": 0.5243245790384291,
+      "learning_rate": 1.823013221265541e-06,
+      "loss": 0.1592,
+      "step": 6712
+    },
+    {
+      "epoch": 2.2417765904157623,
+      "grad_norm": 0.5330623020965718,
+      "learning_rate": 1.8215128477582077e-06,
+      "loss": 0.1623,
+      "step": 6713
+    },
+    {
+      "epoch": 2.242110535982635,
+      "grad_norm": 0.5257624493735769,
+      "learning_rate": 1.8200129543987843e-06,
+      "loss": 0.1633,
+      "step": 6714
+    },
+    {
+      "epoch": 2.2424444815495073,
+      "grad_norm": 0.5021319140625929,
+      "learning_rate": 1.818513541413847e-06,
+      "loss": 0.1648,
+      "step": 6715
+    },
+    {
+      "epoch": 2.24277842711638,
+      "grad_norm": 0.5189532423799506,
+      "learning_rate": 1.8170146090299018e-06,
+      "loss": 0.1596,
+      "step": 6716
+    },
+    {
+      "epoch": 2.2431123726832527,
+      "grad_norm": 0.5187290309316134,
+      "learning_rate": 1.8155161574733772e-06,
+      "loss": 0.1553,
+      "step": 6717
+    },
+    {
+      "epoch": 2.2434463182501254,
+      "grad_norm": 0.5275964325368073,
+      "learning_rate": 1.8140181869706341e-06,
+      "loss": 0.1555,
+      "step": 6718
+    },
+    {
+      "epoch": 2.2437802638169977,
+      "grad_norm": 0.5053082731154795,
+      "learning_rate": 1.812520697747956e-06,
+      "loss": 0.1592,
+      "step": 6719
+    },
+    {
+      "epoch": 2.2441142093838704,
+      "grad_norm": 0.539300185313967,
+      "learning_rate": 1.8110236900315582e-06,
+      "loss": 0.1582,
+      "step": 6720
+    },
+    {
+      "epoch": 2.244448154950743,
+      "grad_norm": 0.5128000481869541,
+      "learning_rate": 1.8095271640475802e-06,
+      "loss": 0.1615,
+      "step": 6721
+    },
+    {
+      "epoch": 2.2447821005176154,
+      "grad_norm": 0.5339868466354106,
+      "learning_rate": 1.8080311200220935e-06,
+      "loss": 0.1661,
+      "step": 6722
+    },
+    {
+      "epoch": 2.245116046084488,
+      "grad_norm": 0.5083433012779742,
+      "learning_rate": 1.8065355581810878e-06,
+      "loss": 0.1551,
+      "step": 6723
+    },
+    {
+      "epoch": 2.245449991651361,
+      "grad_norm": 0.5628766905570722,
+      "learning_rate": 1.80504047875049e-06,
+      "loss": 0.1724,
+      "step": 6724
+    },
+    {
+      "epoch": 2.2457839372182336,
+      "grad_norm": 0.5084207068862423,
+      "learning_rate": 1.8035458819561453e-06,
+      "loss": 0.158,
+      "step": 6725
+    },
+    {
+      "epoch": 2.246117882785106,
+      "grad_norm": 0.5065025201452854,
+      "learning_rate": 1.8020517680238326e-06,
+      "loss": 0.1598,
+      "step": 6726
+    },
+    {
+      "epoch": 2.2464518283519785,
+      "grad_norm": 0.4990795520486874,
+      "learning_rate": 1.8005581371792564e-06,
+      "loss": 0.1525,
+      "step": 6727
+    },
+    {
+      "epoch": 2.2467857739188513,
+      "grad_norm": 0.5080788566646063,
+      "learning_rate": 1.799064989648044e-06,
+      "loss": 0.1528,
+      "step": 6728
+    },
+    {
+      "epoch": 2.247119719485724,
+      "grad_norm": 0.4974786247986726,
+      "learning_rate": 1.797572325655756e-06,
+      "loss": 0.1557,
+      "step": 6729
+    },
+    {
+      "epoch": 2.2474536650525963,
+      "grad_norm": 0.47239598611834804,
+      "learning_rate": 1.7960801454278742e-06,
+      "loss": 0.1466,
+      "step": 6730
+    },
+    {
+      "epoch": 2.247787610619469,
+      "grad_norm": 0.4902956696179493,
+      "learning_rate": 1.7945884491898119e-06,
+      "loss": 0.1496,
+      "step": 6731
+    },
+    {
+      "epoch": 2.2481215561863417,
+      "grad_norm": 0.4867643676693888,
+      "learning_rate": 1.7930972371669064e-06,
+      "loss": 0.153,
+      "step": 6732
+    },
+    {
+      "epoch": 2.2484555017532144,
+      "grad_norm": 0.49740646202701094,
+      "learning_rate": 1.791606509584425e-06,
+      "loss": 0.1581,
+      "step": 6733
+    },
+    {
+      "epoch": 2.2487894473200867,
+      "grad_norm": 0.4636440286036032,
+      "learning_rate": 1.7901162666675564e-06,
+      "loss": 0.1425,
+      "step": 6734
+    },
+    {
+      "epoch": 2.2491233928869594,
+      "grad_norm": 0.5257749950431144,
+      "learning_rate": 1.7886265086414222e-06,
+      "loss": 0.1714,
+      "step": 6735
+    },
+    {
+      "epoch": 2.249457338453832,
+      "grad_norm": 0.5635419552815703,
+      "learning_rate": 1.7871372357310651e-06,
+      "loss": 0.1699,
+      "step": 6736
+    },
+    {
+      "epoch": 2.249791284020705,
+      "grad_norm": 0.6170420468842089,
+      "learning_rate": 1.7856484481614605e-06,
+      "loss": 0.1536,
+      "step": 6737
+    },
+    {
+      "epoch": 2.250125229587577,
+      "grad_norm": 0.5167276951617383,
+      "learning_rate": 1.784160146157502e-06,
+      "loss": 0.1624,
+      "step": 6738
+    },
+    {
+      "epoch": 2.25045917515445,
+      "grad_norm": 0.54841599016874,
+      "learning_rate": 1.7826723299440224e-06,
+      "loss": 0.1693,
+      "step": 6739
+    },
+    {
+      "epoch": 2.2507931207213225,
+      "grad_norm": 0.45416893488770504,
+      "learning_rate": 1.7811849997457681e-06,
+      "loss": 0.1435,
+      "step": 6740
+    },
+    {
+      "epoch": 2.251127066288195,
+      "grad_norm": 0.5136555475743064,
+      "learning_rate": 1.779698155787422e-06,
+      "loss": 0.1577,
+      "step": 6741
+    },
+    {
+      "epoch": 2.2514610118550675,
+      "grad_norm": 0.4719315935157619,
+      "learning_rate": 1.7782117982935854e-06,
+      "loss": 0.1522,
+      "step": 6742
+    },
+    {
+      "epoch": 2.2517949574219402,
+      "grad_norm": 0.5301261871177757,
+      "learning_rate": 1.7767259274887937e-06,
+      "loss": 0.163,
+      "step": 6743
+    },
+    {
+      "epoch": 2.252128902988813,
+      "grad_norm": 0.5431816850628656,
+      "learning_rate": 1.7752405435975002e-06,
+      "loss": 0.1704,
+      "step": 6744
+    },
+    {
+      "epoch": 2.2524628485556852,
+      "grad_norm": 0.49674970018012937,
+      "learning_rate": 1.7737556468440964e-06,
+      "loss": 0.1541,
+      "step": 6745
+    },
+    {
+      "epoch": 2.252796794122558,
+      "grad_norm": 0.5211417957528897,
+      "learning_rate": 1.7722712374528877e-06,
+      "loss": 0.1613,
+      "step": 6746
+    },
+    {
+      "epoch": 2.2531307396894307,
+      "grad_norm": 0.5097312100588599,
+      "learning_rate": 1.7707873156481158e-06,
+      "loss": 0.1553,
+      "step": 6747
+    },
+    {
+      "epoch": 2.2534646852563034,
+      "grad_norm": 0.5066032252735806,
+      "learning_rate": 1.7693038816539416e-06,
+      "loss": 0.1571,
+      "step": 6748
+    },
+    {
+      "epoch": 2.2537986308231757,
+      "grad_norm": 0.5454789116485014,
+      "learning_rate": 1.767820935694457e-06,
+      "loss": 0.1717,
+      "step": 6749
+    },
+    {
+      "epoch": 2.2541325763900484,
+      "grad_norm": 0.524425253837954,
+      "learning_rate": 1.7663384779936764e-06,
+      "loss": 0.1659,
+      "step": 6750
+    },
+    {
+      "epoch": 2.254466521956921,
+      "grad_norm": 0.49825581432779875,
+      "learning_rate": 1.7648565087755442e-06,
+      "loss": 0.1494,
+      "step": 6751
+    },
+    {
+      "epoch": 2.2548004675237934,
+      "grad_norm": 0.5309709923110758,
+      "learning_rate": 1.76337502826393e-06,
+      "loss": 0.1589,
+      "step": 6752
+    },
+    {
+      "epoch": 2.255134413090666,
+      "grad_norm": 0.5173890025549714,
+      "learning_rate": 1.7618940366826266e-06,
+      "loss": 0.1596,
+      "step": 6753
+    },
+    {
+      "epoch": 2.255468358657539,
+      "grad_norm": 0.538202002454078,
+      "learning_rate": 1.7604135342553564e-06,
+      "loss": 0.1688,
+      "step": 6754
+    },
+    {
+      "epoch": 2.2558023042244115,
+      "grad_norm": 0.5511066879412111,
+      "learning_rate": 1.7589335212057663e-06,
+      "loss": 0.1716,
+      "step": 6755
+    },
+    {
+      "epoch": 2.2561362497912842,
+      "grad_norm": 0.49185964257268533,
+      "learning_rate": 1.7574539977574323e-06,
+      "loss": 0.1553,
+      "step": 6756
+    },
+    {
+      "epoch": 2.2564701953581565,
+      "grad_norm": 0.5284689435240054,
+      "learning_rate": 1.7559749641338497e-06,
+      "loss": 0.157,
+      "step": 6757
+    },
+    {
+      "epoch": 2.2568041409250292,
+      "grad_norm": 0.5137649551069989,
+      "learning_rate": 1.7544964205584476e-06,
+      "loss": 0.1509,
+      "step": 6758
+    },
+    {
+      "epoch": 2.257138086491902,
+      "grad_norm": 0.5155409321355809,
+      "learning_rate": 1.7530183672545743e-06,
+      "loss": 0.1557,
+      "step": 6759
+    },
+    {
+      "epoch": 2.2574720320587742,
+      "grad_norm": 0.5312821280939001,
+      "learning_rate": 1.7515408044455102e-06,
+      "loss": 0.1541,
+      "step": 6760
+    },
+    {
+      "epoch": 2.257805977625647,
+      "grad_norm": 0.5410274974402473,
+      "learning_rate": 1.7500637323544534e-06,
+      "loss": 0.1625,
+      "step": 6761
+    },
+    {
+      "epoch": 2.2581399231925197,
+      "grad_norm": 0.5369298073405431,
+      "learning_rate": 1.74858715120454e-06,
+      "loss": 0.1623,
+      "step": 6762
+    },
+    {
+      "epoch": 2.2584738687593924,
+      "grad_norm": 0.5330721160665443,
+      "learning_rate": 1.7471110612188203e-06,
+      "loss": 0.1667,
+      "step": 6763
+    },
+    {
+      "epoch": 2.2588078143262647,
+      "grad_norm": 0.5564739677028375,
+      "learning_rate": 1.7456354626202775e-06,
+      "loss": 0.1622,
+      "step": 6764
+    },
+    {
+      "epoch": 2.2591417598931374,
+      "grad_norm": 0.5025246524131759,
+      "learning_rate": 1.7441603556318155e-06,
+      "loss": 0.1591,
+      "step": 6765
+    },
+    {
+      "epoch": 2.25947570546001,
+      "grad_norm": 0.5102844772368513,
+      "learning_rate": 1.74268574047627e-06,
+      "loss": 0.1605,
+      "step": 6766
+    },
+    {
+      "epoch": 2.259809651026883,
+      "grad_norm": 0.528306743042957,
+      "learning_rate": 1.7412116173763931e-06,
+      "loss": 0.1564,
+      "step": 6767
+    },
+    {
+      "epoch": 2.260143596593755,
+      "grad_norm": 0.5076408569071325,
+      "learning_rate": 1.7397379865548758e-06,
+      "loss": 0.1603,
+      "step": 6768
+    },
+    {
+      "epoch": 2.260477542160628,
+      "grad_norm": 0.5293441364353967,
+      "learning_rate": 1.7382648482343229e-06,
+      "loss": 0.1626,
+      "step": 6769
+    },
+    {
+      "epoch": 2.2608114877275005,
+      "grad_norm": 0.5160283033863646,
+      "learning_rate": 1.7367922026372713e-06,
+      "loss": 0.1491,
+      "step": 6770
+    },
+    {
+      "epoch": 2.261145433294373,
+      "grad_norm": 0.5243537166726169,
+      "learning_rate": 1.7353200499861794e-06,
+      "loss": 0.1659,
+      "step": 6771
+    },
+    {
+      "epoch": 2.2614793788612455,
+      "grad_norm": 0.46787144815206605,
+      "learning_rate": 1.733848390503436e-06,
+      "loss": 0.1473,
+      "step": 6772
+    },
+    {
+      "epoch": 2.261813324428118,
+      "grad_norm": 0.5408034010548239,
+      "learning_rate": 1.732377224411349e-06,
+      "loss": 0.1628,
+      "step": 6773
+    },
+    {
+      "epoch": 2.262147269994991,
+      "grad_norm": 0.5580667063910909,
+      "learning_rate": 1.7309065519321572e-06,
+      "loss": 0.1675,
+      "step": 6774
+    },
+    {
+      "epoch": 2.2624812155618637,
+      "grad_norm": 0.5029465093966058,
+      "learning_rate": 1.729436373288025e-06,
+      "loss": 0.1554,
+      "step": 6775
+    },
+    {
+      "epoch": 2.262815161128736,
+      "grad_norm": 0.515217227344929,
+      "learning_rate": 1.7279666887010361e-06,
+      "loss": 0.1618,
+      "step": 6776
+    },
+    {
+      "epoch": 2.2631491066956086,
+      "grad_norm": 0.5002295957612295,
+      "learning_rate": 1.726497498393206e-06,
+      "loss": 0.1531,
+      "step": 6777
+    },
+    {
+      "epoch": 2.2634830522624814,
+      "grad_norm": 0.5206621476137773,
+      "learning_rate": 1.7250288025864747e-06,
+      "loss": 0.17,
+      "step": 6778
+    },
+    {
+      "epoch": 2.2638169978293536,
+      "grad_norm": 0.5521129732067034,
+      "learning_rate": 1.7235606015027029e-06,
+      "loss": 0.1707,
+      "step": 6779
+    },
+    {
+      "epoch": 2.2641509433962264,
+      "grad_norm": 0.5102176164588842,
+      "learning_rate": 1.7220928953636812e-06,
+      "loss": 0.1572,
+      "step": 6780
+    },
+    {
+      "epoch": 2.264484888963099,
+      "grad_norm": 0.5209070051551632,
+      "learning_rate": 1.7206256843911252e-06,
+      "loss": 0.1541,
+      "step": 6781
+    },
+    {
+      "epoch": 2.264818834529972,
+      "grad_norm": 0.48372276583702023,
+      "learning_rate": 1.7191589688066706e-06,
+      "loss": 0.1577,
+      "step": 6782
+    },
+    {
+      "epoch": 2.265152780096844,
+      "grad_norm": 0.5580522540630803,
+      "learning_rate": 1.7176927488318868e-06,
+      "loss": 0.1672,
+      "step": 6783
+    },
+    {
+      "epoch": 2.265486725663717,
+      "grad_norm": 0.5402796189908685,
+      "learning_rate": 1.7162270246882595e-06,
+      "loss": 0.1607,
+      "step": 6784
+    },
+    {
+      "epoch": 2.2658206712305895,
+      "grad_norm": 0.48998808382575204,
+      "learning_rate": 1.7147617965972052e-06,
+      "loss": 0.1558,
+      "step": 6785
+    },
+    {
+      "epoch": 2.266154616797462,
+      "grad_norm": 0.4631863279760277,
+      "learning_rate": 1.7132970647800639e-06,
+      "loss": 0.1402,
+      "step": 6786
+    },
+    {
+      "epoch": 2.2664885623643345,
+      "grad_norm": 0.5319179193859932,
+      "learning_rate": 1.7118328294581028e-06,
+      "loss": 0.1637,
+      "step": 6787
+    },
+    {
+      "epoch": 2.266822507931207,
+      "grad_norm": 0.5057429440108493,
+      "learning_rate": 1.7103690908525072e-06,
+      "loss": 0.1508,
+      "step": 6788
+    },
+    {
+      "epoch": 2.26715645349808,
+      "grad_norm": 0.5396050693212004,
+      "learning_rate": 1.7089058491843967e-06,
+      "loss": 0.1644,
+      "step": 6789
+    },
+    {
+      "epoch": 2.267490399064952,
+      "grad_norm": 0.5070077732229445,
+      "learning_rate": 1.7074431046748075e-06,
+      "loss": 0.1552,
+      "step": 6790
+    },
+    {
+      "epoch": 2.267824344631825,
+      "grad_norm": 0.5506435678104178,
+      "learning_rate": 1.7059808575447057e-06,
+      "loss": 0.1587,
+      "step": 6791
+    },
+    {
+      "epoch": 2.2681582901986976,
+      "grad_norm": 0.5180851984442151,
+      "learning_rate": 1.7045191080149815e-06,
+      "loss": 0.1545,
+      "step": 6792
+    },
+    {
+      "epoch": 2.2684922357655704,
+      "grad_norm": 0.503753635614517,
+      "learning_rate": 1.7030578563064504e-06,
+      "loss": 0.1575,
+      "step": 6793
+    },
+    {
+      "epoch": 2.2688261813324426,
+      "grad_norm": 0.5181212723409958,
+      "learning_rate": 1.7015971026398487e-06,
+      "loss": 0.1563,
+      "step": 6794
+    },
+    {
+      "epoch": 2.2691601268993153,
+      "grad_norm": 0.47584629353603225,
+      "learning_rate": 1.7001368472358442e-06,
+      "loss": 0.149,
+      "step": 6795
+    },
+    {
+      "epoch": 2.269494072466188,
+      "grad_norm": 0.5001463375155726,
+      "learning_rate": 1.6986770903150213e-06,
+      "loss": 0.1497,
+      "step": 6796
+    },
+    {
+      "epoch": 2.269828018033061,
+      "grad_norm": 0.5233528316290307,
+      "learning_rate": 1.697217832097896e-06,
+      "loss": 0.1543,
+      "step": 6797
+    },
+    {
+      "epoch": 2.270161963599933,
+      "grad_norm": 0.5574253388658429,
+      "learning_rate": 1.6957590728049078e-06,
+      "loss": 0.1696,
+      "step": 6798
+    },
+    {
+      "epoch": 2.2704959091668058,
+      "grad_norm": 0.5284625902947807,
+      "learning_rate": 1.6943008126564164e-06,
+      "loss": 0.1522,
+      "step": 6799
+    },
+    {
+      "epoch": 2.2708298547336785,
+      "grad_norm": 0.5126830539761008,
+      "learning_rate": 1.6928430518727102e-06,
+      "loss": 0.159,
+      "step": 6800
+    },
+    {
+      "epoch": 2.2711638003005508,
+      "grad_norm": 0.5004642345379817,
+      "learning_rate": 1.6913857906740033e-06,
+      "loss": 0.1561,
+      "step": 6801
+    },
+    {
+      "epoch": 2.2714977458674235,
+      "grad_norm": 0.5114750316659729,
+      "learning_rate": 1.6899290292804288e-06,
+      "loss": 0.1614,
+      "step": 6802
+    },
+    {
+      "epoch": 2.271831691434296,
+      "grad_norm": 0.5395301012866076,
+      "learning_rate": 1.6884727679120493e-06,
+      "loss": 0.1575,
+      "step": 6803
+    },
+    {
+      "epoch": 2.272165637001169,
+      "grad_norm": 0.5343051941291189,
+      "learning_rate": 1.687017006788852e-06,
+      "loss": 0.1649,
+      "step": 6804
+    },
+    {
+      "epoch": 2.2724995825680416,
+      "grad_norm": 0.49336411451085405,
+      "learning_rate": 1.6855617461307427e-06,
+      "loss": 0.1535,
+      "step": 6805
+    },
+    {
+      "epoch": 2.272833528134914,
+      "grad_norm": 0.5168391221141233,
+      "learning_rate": 1.6841069861575598e-06,
+      "loss": 0.1534,
+      "step": 6806
+    },
+    {
+      "epoch": 2.2731674737017866,
+      "grad_norm": 0.49208671517873626,
+      "learning_rate": 1.6826527270890587e-06,
+      "loss": 0.1514,
+      "step": 6807
+    },
+    {
+      "epoch": 2.2735014192686593,
+      "grad_norm": 0.4972768325081571,
+      "learning_rate": 1.6811989691449232e-06,
+      "loss": 0.1572,
+      "step": 6808
+    },
+    {
+      "epoch": 2.2738353648355316,
+      "grad_norm": 0.48863264662669803,
+      "learning_rate": 1.6797457125447614e-06,
+      "loss": 0.1481,
+      "step": 6809
+    },
+    {
+      "epoch": 2.2741693104024043,
+      "grad_norm": 0.5221620981515211,
+      "learning_rate": 1.678292957508106e-06,
+      "loss": 0.1627,
+      "step": 6810
+    },
+    {
+      "epoch": 2.274503255969277,
+      "grad_norm": 0.519752893108589,
+      "learning_rate": 1.6768407042544093e-06,
+      "loss": 0.1641,
+      "step": 6811
+    },
+    {
+      "epoch": 2.2748372015361498,
+      "grad_norm": 0.5210756938106716,
+      "learning_rate": 1.6753889530030554e-06,
+      "loss": 0.156,
+      "step": 6812
+    },
+    {
+      "epoch": 2.275171147103022,
+      "grad_norm": 0.5504910072599424,
+      "learning_rate": 1.673937703973344e-06,
+      "loss": 0.166,
+      "step": 6813
+    },
+    {
+      "epoch": 2.2755050926698948,
+      "grad_norm": 0.4910304113541981,
+      "learning_rate": 1.6724869573845054e-06,
+      "loss": 0.1502,
+      "step": 6814
+    },
+    {
+      "epoch": 2.2758390382367675,
+      "grad_norm": 0.5560014050143508,
+      "learning_rate": 1.6710367134556926e-06,
+      "loss": 0.1667,
+      "step": 6815
+    },
+    {
+      "epoch": 2.27617298380364,
+      "grad_norm": 0.5421536902759155,
+      "learning_rate": 1.6695869724059827e-06,
+      "loss": 0.1513,
+      "step": 6816
+    },
+    {
+      "epoch": 2.2765069293705125,
+      "grad_norm": 0.5104732647871095,
+      "learning_rate": 1.6681377344543737e-06,
+      "loss": 0.1632,
+      "step": 6817
+    },
+    {
+      "epoch": 2.276840874937385,
+      "grad_norm": 0.5161082704973411,
+      "learning_rate": 1.6666889998197927e-06,
+      "loss": 0.1588,
+      "step": 6818
+    },
+    {
+      "epoch": 2.277174820504258,
+      "grad_norm": 0.5211932452386149,
+      "learning_rate": 1.6652407687210853e-06,
+      "loss": 0.1645,
+      "step": 6819
+    },
+    {
+      "epoch": 2.27750876607113,
+      "grad_norm": 0.47349743298012514,
+      "learning_rate": 1.6637930413770249e-06,
+      "loss": 0.1438,
+      "step": 6820
+    },
+    {
+      "epoch": 2.277842711638003,
+      "grad_norm": 0.5511636785563299,
+      "learning_rate": 1.6623458180063084e-06,
+      "loss": 0.1632,
+      "step": 6821
+    },
+    {
+      "epoch": 2.2781766572048756,
+      "grad_norm": 0.5180653642051466,
+      "learning_rate": 1.6608990988275575e-06,
+      "loss": 0.1581,
+      "step": 6822
+    },
+    {
+      "epoch": 2.2785106027717483,
+      "grad_norm": 0.48902595955559913,
+      "learning_rate": 1.6594528840593128e-06,
+      "loss": 0.1577,
+      "step": 6823
+    },
+    {
+      "epoch": 2.278844548338621,
+      "grad_norm": 0.5283771989671097,
+      "learning_rate": 1.6580071739200448e-06,
+      "loss": 0.1637,
+      "step": 6824
+    },
+    {
+      "epoch": 2.2791784939054933,
+      "grad_norm": 0.4798685050156958,
+      "learning_rate": 1.6565619686281425e-06,
+      "loss": 0.1425,
+      "step": 6825
+    },
+    {
+      "epoch": 2.279512439472366,
+      "grad_norm": 0.5262706305004466,
+      "learning_rate": 1.6551172684019224e-06,
+      "loss": 0.1572,
+      "step": 6826
+    },
+    {
+      "epoch": 2.2798463850392388,
+      "grad_norm": 0.5158131133227103,
+      "learning_rate": 1.6536730734596257e-06,
+      "loss": 0.1578,
+      "step": 6827
+    },
+    {
+      "epoch": 2.280180330606111,
+      "grad_norm": 0.5469011331347828,
+      "learning_rate": 1.652229384019411e-06,
+      "loss": 0.171,
+      "step": 6828
+    },
+    {
+      "epoch": 2.2805142761729837,
+      "grad_norm": 0.5075499952775651,
+      "learning_rate": 1.650786200299368e-06,
+      "loss": 0.1542,
+      "step": 6829
+    },
+    {
+      "epoch": 2.2808482217398565,
+      "grad_norm": 0.5425927659062526,
+      "learning_rate": 1.6493435225175042e-06,
+      "loss": 0.1612,
+      "step": 6830
+    },
+    {
+      "epoch": 2.281182167306729,
+      "grad_norm": 0.46335339234518974,
+      "learning_rate": 1.6479013508917552e-06,
+      "loss": 0.1502,
+      "step": 6831
+    },
+    {
+      "epoch": 2.2815161128736015,
+      "grad_norm": 0.48052008185076805,
+      "learning_rate": 1.6464596856399734e-06,
+      "loss": 0.1493,
+      "step": 6832
+    },
+    {
+      "epoch": 2.281850058440474,
+      "grad_norm": 0.48377398516358,
+      "learning_rate": 1.6450185269799462e-06,
+      "loss": 0.1529,
+      "step": 6833
+    },
+    {
+      "epoch": 2.282184004007347,
+      "grad_norm": 0.5249199169347003,
+      "learning_rate": 1.6435778751293723e-06,
+      "loss": 0.1536,
+      "step": 6834
+    },
+    {
+      "epoch": 2.2825179495742196,
+      "grad_norm": 0.49911936444525085,
+      "learning_rate": 1.6421377303058829e-06,
+      "loss": 0.1509,
+      "step": 6835
+    },
+    {
+      "epoch": 2.282851895141092,
+      "grad_norm": 0.5020572435479131,
+      "learning_rate": 1.640698092727025e-06,
+      "loss": 0.1572,
+      "step": 6836
+    },
+    {
+      "epoch": 2.2831858407079646,
+      "grad_norm": 0.5029497729673867,
+      "learning_rate": 1.639258962610275e-06,
+      "loss": 0.1536,
+      "step": 6837
+    },
+    {
+      "epoch": 2.2835197862748373,
+      "grad_norm": 0.4852289148568792,
+      "learning_rate": 1.6378203401730303e-06,
+      "loss": 0.163,
+      "step": 6838
+    },
+    {
+      "epoch": 2.2838537318417096,
+      "grad_norm": 0.5788851829372151,
+      "learning_rate": 1.6363822256326128e-06,
+      "loss": 0.1653,
+      "step": 6839
+    },
+    {
+      "epoch": 2.2841876774085823,
+      "grad_norm": 0.5116983762288065,
+      "learning_rate": 1.6349446192062635e-06,
+      "loss": 0.1602,
+      "step": 6840
+    },
+    {
+      "epoch": 2.284521622975455,
+      "grad_norm": 0.5586774089313628,
+      "learning_rate": 1.633507521111154e-06,
+      "loss": 0.1619,
+      "step": 6841
+    },
+    {
+      "epoch": 2.2848555685423277,
+      "grad_norm": 0.5331816877090615,
+      "learning_rate": 1.6320709315643708e-06,
+      "loss": 0.1701,
+      "step": 6842
+    },
+    {
+      "epoch": 2.2851895141092,
+      "grad_norm": 0.49596637489137746,
+      "learning_rate": 1.6306348507829294e-06,
+      "loss": 0.1521,
+      "step": 6843
+    },
+    {
+      "epoch": 2.2855234596760727,
+      "grad_norm": 0.5525180134705676,
+      "learning_rate": 1.6291992789837669e-06,
+      "loss": 0.1614,
+      "step": 6844
+    },
+    {
+      "epoch": 2.2858574052429455,
+      "grad_norm": 0.5308161541932059,
+      "learning_rate": 1.6277642163837444e-06,
+      "loss": 0.1607,
+      "step": 6845
+    },
+    {
+      "epoch": 2.286191350809818,
+      "grad_norm": 0.49825308016711456,
+      "learning_rate": 1.6263296631996422e-06,
+      "loss": 0.1527,
+      "step": 6846
+    },
+    {
+      "epoch": 2.2865252963766904,
+      "grad_norm": 0.4718547788653712,
+      "learning_rate": 1.6248956196481701e-06,
+      "loss": 0.148,
+      "step": 6847
+    },
+    {
+      "epoch": 2.286859241943563,
+      "grad_norm": 0.5382050275600502,
+      "learning_rate": 1.6234620859459537e-06,
+      "loss": 0.1613,
+      "step": 6848
+    },
+    {
+      "epoch": 2.287193187510436,
+      "grad_norm": 0.49960602289495415,
+      "learning_rate": 1.6220290623095463e-06,
+      "loss": 0.1491,
+      "step": 6849
+    },
+    {
+      "epoch": 2.287527133077308,
+      "grad_norm": 0.5156720699510375,
+      "learning_rate": 1.6205965489554248e-06,
+      "loss": 0.1561,
+      "step": 6850
+    },
+    {
+      "epoch": 2.287861078644181,
+      "grad_norm": 0.5025226699485877,
+      "learning_rate": 1.619164546099985e-06,
+      "loss": 0.1593,
+      "step": 6851
+    },
+    {
+      "epoch": 2.2881950242110536,
+      "grad_norm": 0.5185730184204812,
+      "learning_rate": 1.6177330539595493e-06,
+      "loss": 0.15,
+      "step": 6852
+    },
+    {
+      "epoch": 2.2885289697779263,
+      "grad_norm": 0.48338359238850986,
+      "learning_rate": 1.6163020727503592e-06,
+      "loss": 0.1486,
+      "step": 6853
+    },
+    {
+      "epoch": 2.288862915344799,
+      "grad_norm": 0.5446527503447174,
+      "learning_rate": 1.6148716026885847e-06,
+      "loss": 0.1715,
+      "step": 6854
+    },
+    {
+      "epoch": 2.2891968609116713,
+      "grad_norm": 0.5200794562046593,
+      "learning_rate": 1.61344164399031e-06,
+      "loss": 0.1543,
+      "step": 6855
+    },
+    {
+      "epoch": 2.289530806478544,
+      "grad_norm": 0.47904725129674774,
+      "learning_rate": 1.6120121968715535e-06,
+      "loss": 0.1538,
+      "step": 6856
+    },
+    {
+      "epoch": 2.2898647520454167,
+      "grad_norm": 0.544473388151401,
+      "learning_rate": 1.6105832615482453e-06,
+      "loss": 0.1712,
+      "step": 6857
+    },
+    {
+      "epoch": 2.290198697612289,
+      "grad_norm": 0.4891963514385171,
+      "learning_rate": 1.609154838236246e-06,
+      "loss": 0.15,
+      "step": 6858
+    },
+    {
+      "epoch": 2.2905326431791617,
+      "grad_norm": 0.5130192652682918,
+      "learning_rate": 1.6077269271513328e-06,
+      "loss": 0.1616,
+      "step": 6859
+    },
+    {
+      "epoch": 2.2908665887460344,
+      "grad_norm": 0.5118177941331495,
+      "learning_rate": 1.606299528509212e-06,
+      "loss": 0.1592,
+      "step": 6860
+    },
+    {
+      "epoch": 2.291200534312907,
+      "grad_norm": 0.49636585798317306,
+      "learning_rate": 1.604872642525503e-06,
+      "loss": 0.1486,
+      "step": 6861
+    },
+    {
+      "epoch": 2.2915344798797794,
+      "grad_norm": 0.5108213993992322,
+      "learning_rate": 1.6034462694157615e-06,
+      "loss": 0.1589,
+      "step": 6862
+    },
+    {
+      "epoch": 2.291868425446652,
+      "grad_norm": 0.5216001516467106,
+      "learning_rate": 1.6020204093954523e-06,
+      "loss": 0.1618,
+      "step": 6863
+    },
+    {
+      "epoch": 2.292202371013525,
+      "grad_norm": 0.4612499122336571,
+      "learning_rate": 1.6005950626799716e-06,
+      "loss": 0.1478,
+      "step": 6864
+    },
+    {
+      "epoch": 2.2925363165803976,
+      "grad_norm": 0.5496928863070641,
+      "learning_rate": 1.5991702294846318e-06,
+      "loss": 0.169,
+      "step": 6865
+    },
+    {
+      "epoch": 2.29287026214727,
+      "grad_norm": 0.519491010757705,
+      "learning_rate": 1.597745910024674e-06,
+      "loss": 0.1535,
+      "step": 6866
+    },
+    {
+      "epoch": 2.2932042077141426,
+      "grad_norm": 0.5395166624198875,
+      "learning_rate": 1.5963221045152537e-06,
+      "loss": 0.1679,
+      "step": 6867
+    },
+    {
+      "epoch": 2.2935381532810153,
+      "grad_norm": 0.49544413691093875,
+      "learning_rate": 1.5948988131714594e-06,
+      "loss": 0.1502,
+      "step": 6868
+    },
+    {
+      "epoch": 2.2938720988478876,
+      "grad_norm": 0.5089430397982071,
+      "learning_rate": 1.593476036208292e-06,
+      "loss": 0.1584,
+      "step": 6869
+    },
+    {
+      "epoch": 2.2942060444147603,
+      "grad_norm": 0.4761749830623971,
+      "learning_rate": 1.5920537738406811e-06,
+      "loss": 0.1438,
+      "step": 6870
+    },
+    {
+      "epoch": 2.294539989981633,
+      "grad_norm": 0.5128220624654717,
+      "learning_rate": 1.5906320262834735e-06,
+      "loss": 0.1582,
+      "step": 6871
+    },
+    {
+      "epoch": 2.2948739355485057,
+      "grad_norm": 0.4769933038550351,
+      "learning_rate": 1.5892107937514424e-06,
+      "loss": 0.1482,
+      "step": 6872
+    },
+    {
+      "epoch": 2.2952078811153784,
+      "grad_norm": 0.49727500860357027,
+      "learning_rate": 1.587790076459283e-06,
+      "loss": 0.1569,
+      "step": 6873
+    },
+    {
+      "epoch": 2.2955418266822507,
+      "grad_norm": 0.5198335573936426,
+      "learning_rate": 1.5863698746216082e-06,
+      "loss": 0.1573,
+      "step": 6874
+    },
+    {
+      "epoch": 2.2958757722491234,
+      "grad_norm": 0.5199506171376683,
+      "learning_rate": 1.58495018845296e-06,
+      "loss": 0.1652,
+      "step": 6875
+    },
+    {
+      "epoch": 2.296209717815996,
+      "grad_norm": 0.5041888828958186,
+      "learning_rate": 1.5835310181677954e-06,
+      "loss": 0.1588,
+      "step": 6876
+    },
+    {
+      "epoch": 2.2965436633828684,
+      "grad_norm": 0.5595007893689801,
+      "learning_rate": 1.5821123639804992e-06,
+      "loss": 0.1639,
+      "step": 6877
+    },
+    {
+      "epoch": 2.296877608949741,
+      "grad_norm": 0.49220921014348185,
+      "learning_rate": 1.5806942261053715e-06,
+      "loss": 0.1582,
+      "step": 6878
+    },
+    {
+      "epoch": 2.297211554516614,
+      "grad_norm": 0.5223549427049144,
+      "learning_rate": 1.5792766047566455e-06,
+      "loss": 0.1577,
+      "step": 6879
+    },
+    {
+      "epoch": 2.2975455000834866,
+      "grad_norm": 0.4778172847309024,
+      "learning_rate": 1.5778595001484648e-06,
+      "loss": 0.1468,
+      "step": 6880
+    },
+    {
+      "epoch": 2.297879445650359,
+      "grad_norm": 0.5503150749267659,
+      "learning_rate": 1.5764429124949022e-06,
+      "loss": 0.1622,
+      "step": 6881
+    },
+    {
+      "epoch": 2.2982133912172316,
+      "grad_norm": 0.48813966328418384,
+      "learning_rate": 1.5750268420099468e-06,
+      "loss": 0.1537,
+      "step": 6882
+    },
+    {
+      "epoch": 2.2985473367841043,
+      "grad_norm": 0.48710079923452626,
+      "learning_rate": 1.5736112889075167e-06,
+      "loss": 0.1499,
+      "step": 6883
+    },
+    {
+      "epoch": 2.298881282350977,
+      "grad_norm": 0.48893154112588416,
+      "learning_rate": 1.5721962534014424e-06,
+      "loss": 0.1566,
+      "step": 6884
+    },
+    {
+      "epoch": 2.2992152279178493,
+      "grad_norm": 0.4801053992347205,
+      "learning_rate": 1.5707817357054882e-06,
+      "loss": 0.15,
+      "step": 6885
+    },
+    {
+      "epoch": 2.299549173484722,
+      "grad_norm": 0.5580430430657689,
+      "learning_rate": 1.5693677360333293e-06,
+      "loss": 0.1615,
+      "step": 6886
+    },
+    {
+      "epoch": 2.2998831190515947,
+      "grad_norm": 0.530284558249108,
+      "learning_rate": 1.56795425459857e-06,
+      "loss": 0.1566,
+      "step": 6887
+    },
+    {
+      "epoch": 2.300217064618467,
+      "grad_norm": 0.5378779347275838,
+      "learning_rate": 1.5665412916147298e-06,
+      "loss": 0.1532,
+      "step": 6888
+    },
+    {
+      "epoch": 2.3005510101853397,
+      "grad_norm": 0.4847888072829989,
+      "learning_rate": 1.5651288472952564e-06,
+      "loss": 0.1509,
+      "step": 6889
+    },
+    {
+      "epoch": 2.3008849557522124,
+      "grad_norm": 0.5658711305785874,
+      "learning_rate": 1.563716921853512e-06,
+      "loss": 0.162,
+      "step": 6890
+    },
+    {
+      "epoch": 2.301218901319085,
+      "grad_norm": 0.5439624589912487,
+      "learning_rate": 1.562305515502791e-06,
+      "loss": 0.1656,
+      "step": 6891
+    },
+    {
+      "epoch": 2.3015528468859574,
+      "grad_norm": 0.555316715494243,
+      "learning_rate": 1.5608946284562977e-06,
+      "loss": 0.1656,
+      "step": 6892
+    },
+    {
+      "epoch": 2.30188679245283,
+      "grad_norm": 0.5347395221649406,
+      "learning_rate": 1.559484260927166e-06,
+      "loss": 0.1592,
+      "step": 6893
+    },
+    {
+      "epoch": 2.302220738019703,
+      "grad_norm": 0.557941240584311,
+      "learning_rate": 1.5580744131284464e-06,
+      "loss": 0.1677,
+      "step": 6894
+    },
+    {
+      "epoch": 2.3025546835865756,
+      "grad_norm": 0.5276421760511721,
+      "learning_rate": 1.5566650852731151e-06,
+      "loss": 0.1685,
+      "step": 6895
+    },
+    {
+      "epoch": 2.302888629153448,
+      "grad_norm": 0.5020349892960838,
+      "learning_rate": 1.5552562775740654e-06,
+      "loss": 0.1594,
+      "step": 6896
+    },
+    {
+      "epoch": 2.3032225747203205,
+      "grad_norm": 0.5293814651715417,
+      "learning_rate": 1.5538479902441156e-06,
+      "loss": 0.1678,
+      "step": 6897
+    },
+    {
+      "epoch": 2.3035565202871933,
+      "grad_norm": 0.48716428596425015,
+      "learning_rate": 1.5524402234960056e-06,
+      "loss": 0.147,
+      "step": 6898
+    },
+    {
+      "epoch": 2.3038904658540655,
+      "grad_norm": 0.48127955691998964,
+      "learning_rate": 1.5510329775423916e-06,
+      "loss": 0.1542,
+      "step": 6899
+    },
+    {
+      "epoch": 2.3042244114209383,
+      "grad_norm": 0.49897847992147765,
+      "learning_rate": 1.5496262525958583e-06,
+      "loss": 0.1628,
+      "step": 6900
+    },
+    {
+      "epoch": 2.304558356987811,
+      "grad_norm": 0.48569413278480206,
+      "learning_rate": 1.5482200488689054e-06,
+      "loss": 0.1399,
+      "step": 6901
+    },
+    {
+      "epoch": 2.3048923025546837,
+      "grad_norm": 0.5315498993933447,
+      "learning_rate": 1.5468143665739565e-06,
+      "loss": 0.1608,
+      "step": 6902
+    },
+    {
+      "epoch": 2.3052262481215564,
+      "grad_norm": 0.5276387365311821,
+      "learning_rate": 1.5454092059233583e-06,
+      "loss": 0.1563,
+      "step": 6903
+    },
+    {
+      "epoch": 2.3055601936884287,
+      "grad_norm": 0.5473811887845702,
+      "learning_rate": 1.5440045671293774e-06,
+      "loss": 0.1657,
+      "step": 6904
+    },
+    {
+      "epoch": 2.3058941392553014,
+      "grad_norm": 0.519692883836073,
+      "learning_rate": 1.542600450404198e-06,
+      "loss": 0.1621,
+      "step": 6905
+    },
+    {
+      "epoch": 2.306228084822174,
+      "grad_norm": 0.5274237692456049,
+      "learning_rate": 1.5411968559599317e-06,
+      "loss": 0.1502,
+      "step": 6906
+    },
+    {
+      "epoch": 2.3065620303890464,
+      "grad_norm": 0.5165181375891817,
+      "learning_rate": 1.5397937840086048e-06,
+      "loss": 0.167,
+      "step": 6907
+    },
+    {
+      "epoch": 2.306895975955919,
+      "grad_norm": 0.5392613904904744,
+      "learning_rate": 1.5383912347621693e-06,
+      "loss": 0.1609,
+      "step": 6908
+    },
+    {
+      "epoch": 2.307229921522792,
+      "grad_norm": 0.5352550993399279,
+      "learning_rate": 1.5369892084324972e-06,
+      "loss": 0.157,
+      "step": 6909
+    },
+    {
+      "epoch": 2.3075638670896645,
+      "grad_norm": 0.5132398793438658,
+      "learning_rate": 1.5355877052313822e-06,
+      "loss": 0.162,
+      "step": 6910
+    },
+    {
+      "epoch": 2.307897812656537,
+      "grad_norm": 0.5059045512819826,
+      "learning_rate": 1.534186725370535e-06,
+      "loss": 0.1569,
+      "step": 6911
+    },
+    {
+      "epoch": 2.3082317582234095,
+      "grad_norm": 0.5241489880536726,
+      "learning_rate": 1.532786269061593e-06,
+      "loss": 0.1645,
+      "step": 6912
+    },
+    {
+      "epoch": 2.3085657037902823,
+      "grad_norm": 0.5862344966512784,
+      "learning_rate": 1.531386336516107e-06,
+      "loss": 0.1482,
+      "step": 6913
+    },
+    {
+      "epoch": 2.308899649357155,
+      "grad_norm": 0.4937850033656664,
+      "learning_rate": 1.52998692794556e-06,
+      "loss": 0.16,
+      "step": 6914
+    },
+    {
+      "epoch": 2.3092335949240272,
+      "grad_norm": 0.5025478647760935,
+      "learning_rate": 1.5285880435613438e-06,
+      "loss": 0.1536,
+      "step": 6915
+    },
+    {
+      "epoch": 2.3095675404909,
+      "grad_norm": 0.5395976195462507,
+      "learning_rate": 1.5271896835747795e-06,
+      "loss": 0.1602,
+      "step": 6916
+    },
+    {
+      "epoch": 2.3099014860577727,
+      "grad_norm": 0.50885180620015,
+      "learning_rate": 1.5257918481971028e-06,
+      "loss": 0.1584,
+      "step": 6917
+    },
+    {
+      "epoch": 2.310235431624645,
+      "grad_norm": 0.49907541237904407,
+      "learning_rate": 1.524394537639477e-06,
+      "loss": 0.1468,
+      "step": 6918
+    },
+    {
+      "epoch": 2.3105693771915177,
+      "grad_norm": 0.520098742948566,
+      "learning_rate": 1.5229977521129785e-06,
+      "loss": 0.152,
+      "step": 6919
+    },
+    {
+      "epoch": 2.3109033227583904,
+      "grad_norm": 0.555034188223172,
+      "learning_rate": 1.5216014918286097e-06,
+      "loss": 0.1728,
+      "step": 6920
+    },
+    {
+      "epoch": 2.311237268325263,
+      "grad_norm": 0.5277476198952158,
+      "learning_rate": 1.5202057569972945e-06,
+      "loss": 0.1559,
+      "step": 6921
+    },
+    {
+      "epoch": 2.311571213892136,
+      "grad_norm": 0.5082237210587632,
+      "learning_rate": 1.518810547829871e-06,
+      "loss": 0.1546,
+      "step": 6922
+    },
+    {
+      "epoch": 2.311905159459008,
+      "grad_norm": 0.5498238792531575,
+      "learning_rate": 1.517415864537105e-06,
+      "loss": 0.1672,
+      "step": 6923
+    },
+    {
+      "epoch": 2.312239105025881,
+      "grad_norm": 0.5102413698777625,
+      "learning_rate": 1.516021707329678e-06,
+      "loss": 0.1639,
+      "step": 6924
+    },
+    {
+      "epoch": 2.3125730505927535,
+      "grad_norm": 0.49159724301878027,
+      "learning_rate": 1.5146280764181942e-06,
+      "loss": 0.1535,
+      "step": 6925
+    },
+    {
+      "epoch": 2.312906996159626,
+      "grad_norm": 0.5166047110747541,
+      "learning_rate": 1.5132349720131783e-06,
+      "loss": 0.1588,
+      "step": 6926
+    },
+    {
+      "epoch": 2.3132409417264985,
+      "grad_norm": 0.5055406381133966,
+      "learning_rate": 1.511842394325077e-06,
+      "loss": 0.1483,
+      "step": 6927
+    },
+    {
+      "epoch": 2.3135748872933712,
+      "grad_norm": 0.5191101090446806,
+      "learning_rate": 1.5104503435642526e-06,
+      "loss": 0.158,
+      "step": 6928
+    },
+    {
+      "epoch": 2.313908832860244,
+      "grad_norm": 0.5014849155315904,
+      "learning_rate": 1.5090588199409927e-06,
+      "loss": 0.159,
+      "step": 6929
+    },
+    {
+      "epoch": 2.3142427784271162,
+      "grad_norm": 0.5270760259726192,
+      "learning_rate": 1.5076678236655018e-06,
+      "loss": 0.1663,
+      "step": 6930
+    },
+    {
+      "epoch": 2.314576723993989,
+      "grad_norm": 0.5265525170252408,
+      "learning_rate": 1.5062773549479064e-06,
+      "loss": 0.1674,
+      "step": 6931
+    },
+    {
+      "epoch": 2.3149106695608617,
+      "grad_norm": 0.4754623491282869,
+      "learning_rate": 1.504887413998254e-06,
+      "loss": 0.1517,
+      "step": 6932
+    },
+    {
+      "epoch": 2.3152446151277344,
+      "grad_norm": 0.4902506423363311,
+      "learning_rate": 1.5034980010265127e-06,
+      "loss": 0.1553,
+      "step": 6933
+    },
+    {
+      "epoch": 2.3155785606946067,
+      "grad_norm": 0.5575953129911351,
+      "learning_rate": 1.5021091162425672e-06,
+      "loss": 0.1666,
+      "step": 6934
+    },
+    {
+      "epoch": 2.3159125062614794,
+      "grad_norm": 0.5099824545247369,
+      "learning_rate": 1.5007207598562268e-06,
+      "loss": 0.1579,
+      "step": 6935
+    },
+    {
+      "epoch": 2.316246451828352,
+      "grad_norm": 0.539874027836098,
+      "learning_rate": 1.4993329320772177e-06,
+      "loss": 0.1662,
+      "step": 6936
+    },
+    {
+      "epoch": 2.3165803973952244,
+      "grad_norm": 0.5295612734416608,
+      "learning_rate": 1.4979456331151875e-06,
+      "loss": 0.1597,
+      "step": 6937
+    },
+    {
+      "epoch": 2.316914342962097,
+      "grad_norm": 0.5894263333785696,
+      "learning_rate": 1.4965588631797052e-06,
+      "loss": 0.159,
+      "step": 6938
+    },
+    {
+      "epoch": 2.31724828852897,
+      "grad_norm": 0.4741039633929316,
+      "learning_rate": 1.4951726224802593e-06,
+      "loss": 0.1494,
+      "step": 6939
+    },
+    {
+      "epoch": 2.3175822340958425,
+      "grad_norm": 0.4862856197350301,
+      "learning_rate": 1.493786911226256e-06,
+      "loss": 0.1569,
+      "step": 6940
+    },
+    {
+      "epoch": 2.317916179662715,
+      "grad_norm": 0.5529752111878695,
+      "learning_rate": 1.492401729627025e-06,
+      "loss": 0.1608,
+      "step": 6941
+    },
+    {
+      "epoch": 2.3182501252295875,
+      "grad_norm": 0.5507021790387356,
+      "learning_rate": 1.491017077891812e-06,
+      "loss": 0.168,
+      "step": 6942
+    },
+    {
+      "epoch": 2.3185840707964602,
+      "grad_norm": 0.5125603052330294,
+      "learning_rate": 1.4896329562297863e-06,
+      "loss": 0.1602,
+      "step": 6943
+    },
+    {
+      "epoch": 2.318918016363333,
+      "grad_norm": 0.5257969566496764,
+      "learning_rate": 1.4882493648500373e-06,
+      "loss": 0.1576,
+      "step": 6944
+    },
+    {
+      "epoch": 2.319251961930205,
+      "grad_norm": 0.5381499105664779,
+      "learning_rate": 1.48686630396157e-06,
+      "loss": 0.1683,
+      "step": 6945
+    },
+    {
+      "epoch": 2.319585907497078,
+      "grad_norm": 0.5422363917617385,
+      "learning_rate": 1.4854837737733147e-06,
+      "loss": 0.1679,
+      "step": 6946
+    },
+    {
+      "epoch": 2.3199198530639507,
+      "grad_norm": 0.5641296285837942,
+      "learning_rate": 1.484101774494116e-06,
+      "loss": 0.1722,
+      "step": 6947
+    },
+    {
+      "epoch": 2.320253798630823,
+      "grad_norm": 0.5253053775187845,
+      "learning_rate": 1.4827203063327427e-06,
+      "loss": 0.1608,
+      "step": 6948
+    },
+    {
+      "epoch": 2.3205877441976956,
+      "grad_norm": 0.4501898716795521,
+      "learning_rate": 1.4813393694978812e-06,
+      "loss": 0.1527,
+      "step": 6949
+    },
+    {
+      "epoch": 2.3209216897645684,
+      "grad_norm": 0.4766070843342987,
+      "learning_rate": 1.479958964198141e-06,
+      "loss": 0.1473,
+      "step": 6950
+    },
+    {
+      "epoch": 2.321255635331441,
+      "grad_norm": 0.5543755055251879,
+      "learning_rate": 1.4785790906420445e-06,
+      "loss": 0.1729,
+      "step": 6951
+    },
+    {
+      "epoch": 2.321589580898314,
+      "grad_norm": 0.47207588022242153,
+      "learning_rate": 1.4771997490380414e-06,
+      "loss": 0.1449,
+      "step": 6952
+    },
+    {
+      "epoch": 2.321923526465186,
+      "grad_norm": 0.5489402475099499,
+      "learning_rate": 1.4758209395944945e-06,
+      "loss": 0.1666,
+      "step": 6953
+    },
+    {
+      "epoch": 2.322257472032059,
+      "grad_norm": 0.4807129427731627,
+      "learning_rate": 1.47444266251969e-06,
+      "loss": 0.1518,
+      "step": 6954
+    },
+    {
+      "epoch": 2.3225914175989315,
+      "grad_norm": 0.554604022659153,
+      "learning_rate": 1.4730649180218337e-06,
+      "loss": 0.1757,
+      "step": 6955
+    },
+    {
+      "epoch": 2.322925363165804,
+      "grad_norm": 0.5348203406510965,
+      "learning_rate": 1.4716877063090517e-06,
+      "loss": 0.1606,
+      "step": 6956
+    },
+    {
+      "epoch": 2.3232593087326765,
+      "grad_norm": 0.525792610223351,
+      "learning_rate": 1.4703110275893846e-06,
+      "loss": 0.1645,
+      "step": 6957
+    },
+    {
+      "epoch": 2.323593254299549,
+      "grad_norm": 0.5191996867761393,
+      "learning_rate": 1.4689348820707988e-06,
+      "loss": 0.1601,
+      "step": 6958
+    },
+    {
+      "epoch": 2.323927199866422,
+      "grad_norm": 0.5021670260891927,
+      "learning_rate": 1.4675592699611741e-06,
+      "loss": 0.157,
+      "step": 6959
+    },
+    {
+      "epoch": 2.324261145433294,
+      "grad_norm": 0.5189366470501476,
+      "learning_rate": 1.4661841914683156e-06,
+      "loss": 0.1561,
+      "step": 6960
+    },
+    {
+      "epoch": 2.324595091000167,
+      "grad_norm": 0.5367046655871626,
+      "learning_rate": 1.464809646799944e-06,
+      "loss": 0.1705,
+      "step": 6961
+    },
+    {
+      "epoch": 2.3249290365670396,
+      "grad_norm": 0.46937388491530546,
+      "learning_rate": 1.463435636163702e-06,
+      "loss": 0.1535,
+      "step": 6962
+    },
+    {
+      "epoch": 2.3252629821339124,
+      "grad_norm": 0.482195068948941,
+      "learning_rate": 1.4620621597671476e-06,
+      "loss": 0.1498,
+      "step": 6963
+    },
+    {
+      "epoch": 2.3255969277007846,
+      "grad_norm": 0.4782325808602372,
+      "learning_rate": 1.4606892178177633e-06,
+      "loss": 0.1492,
+      "step": 6964
+    },
+    {
+      "epoch": 2.3259308732676574,
+      "grad_norm": 0.49949896473890126,
+      "learning_rate": 1.459316810522945e-06,
+      "loss": 0.1541,
+      "step": 6965
+    },
+    {
+      "epoch": 2.32626481883453,
+      "grad_norm": 0.5514134284700918,
+      "learning_rate": 1.457944938090013e-06,
+      "loss": 0.1627,
+      "step": 6966
+    },
+    {
+      "epoch": 2.3265987644014023,
+      "grad_norm": 0.5122505150543959,
+      "learning_rate": 1.456573600726206e-06,
+      "loss": 0.1582,
+      "step": 6967
+    },
+    {
+      "epoch": 2.326932709968275,
+      "grad_norm": 0.5155388868093768,
+      "learning_rate": 1.4552027986386775e-06,
+      "loss": 0.1605,
+      "step": 6968
+    },
+    {
+      "epoch": 2.3272666555351478,
+      "grad_norm": 0.5319645910926085,
+      "learning_rate": 1.453832532034506e-06,
+      "loss": 0.1672,
+      "step": 6969
+    },
+    {
+      "epoch": 2.3276006011020205,
+      "grad_norm": 0.5212564670307227,
+      "learning_rate": 1.4524628011206843e-06,
+      "loss": 0.1606,
+      "step": 6970
+    },
+    {
+      "epoch": 2.327934546668893,
+      "grad_norm": 0.5197062994619243,
+      "learning_rate": 1.4510936061041269e-06,
+      "loss": 0.1586,
+      "step": 6971
+    },
+    {
+      "epoch": 2.3282684922357655,
+      "grad_norm": 0.5793340057434153,
+      "learning_rate": 1.449724947191668e-06,
+      "loss": 0.1625,
+      "step": 6972
+    },
+    {
+      "epoch": 2.328602437802638,
+      "grad_norm": 0.5080639803068403,
+      "learning_rate": 1.4483568245900597e-06,
+      "loss": 0.1554,
+      "step": 6973
+    },
+    {
+      "epoch": 2.328936383369511,
+      "grad_norm": 0.5268549200084457,
+      "learning_rate": 1.4469892385059713e-06,
+      "loss": 0.1655,
+      "step": 6974
+    },
+    {
+      "epoch": 2.329270328936383,
+      "grad_norm": 0.4921015168951849,
+      "learning_rate": 1.4456221891459953e-06,
+      "loss": 0.1559,
+      "step": 6975
+    },
+    {
+      "epoch": 2.329604274503256,
+      "grad_norm": 0.507333813102178,
+      "learning_rate": 1.4442556767166371e-06,
+      "loss": 0.1478,
+      "step": 6976
+    },
+    {
+      "epoch": 2.3299382200701286,
+      "grad_norm": 0.4805039087114265,
+      "learning_rate": 1.4428897014243288e-06,
+      "loss": 0.143,
+      "step": 6977
+    },
+    {
+      "epoch": 2.3302721656370013,
+      "grad_norm": 0.525864995462506,
+      "learning_rate": 1.4415242634754107e-06,
+      "loss": 0.1564,
+      "step": 6978
+    },
+    {
+      "epoch": 2.3306061112038736,
+      "grad_norm": 0.48453778085656507,
+      "learning_rate": 1.4401593630761562e-06,
+      "loss": 0.153,
+      "step": 6979
+    },
+    {
+      "epoch": 2.3309400567707463,
+      "grad_norm": 0.5560825048911766,
+      "learning_rate": 1.4387950004327434e-06,
+      "loss": 0.1696,
+      "step": 6980
+    },
+    {
+      "epoch": 2.331274002337619,
+      "grad_norm": 0.4999034443705413,
+      "learning_rate": 1.4374311757512798e-06,
+      "loss": 0.1489,
+      "step": 6981
+    },
+    {
+      "epoch": 2.3316079479044918,
+      "grad_norm": 0.48674712053155916,
+      "learning_rate": 1.4360678892377833e-06,
+      "loss": 0.1558,
+      "step": 6982
+    },
+    {
+      "epoch": 2.331941893471364,
+      "grad_norm": 0.5376662163779152,
+      "learning_rate": 1.434705141098197e-06,
+      "loss": 0.1618,
+      "step": 6983
+    },
+    {
+      "epoch": 2.3322758390382368,
+      "grad_norm": 0.49634944316675383,
+      "learning_rate": 1.4333429315383768e-06,
+      "loss": 0.1536,
+      "step": 6984
+    },
+    {
+      "epoch": 2.3326097846051095,
+      "grad_norm": 0.5261178930730273,
+      "learning_rate": 1.4319812607641055e-06,
+      "loss": 0.1491,
+      "step": 6985
+    },
+    {
+      "epoch": 2.3329437301719818,
+      "grad_norm": 0.4925740465169854,
+      "learning_rate": 1.4306201289810756e-06,
+      "loss": 0.1542,
+      "step": 6986
+    },
+    {
+      "epoch": 2.3332776757388545,
+      "grad_norm": 0.5190410270947525,
+      "learning_rate": 1.4292595363949047e-06,
+      "loss": 0.1499,
+      "step": 6987
+    },
+    {
+      "epoch": 2.333611621305727,
+      "grad_norm": 0.5062847873051785,
+      "learning_rate": 1.4278994832111232e-06,
+      "loss": 0.1559,
+      "step": 6988
+    },
+    {
+      "epoch": 2.3339455668726,
+      "grad_norm": 0.5152470669811914,
+      "learning_rate": 1.4265399696351867e-06,
+      "loss": 0.1543,
+      "step": 6989
+    },
+    {
+      "epoch": 2.334279512439472,
+      "grad_norm": 0.5110787108439315,
+      "learning_rate": 1.4251809958724623e-06,
+      "loss": 0.1607,
+      "step": 6990
+    },
+    {
+      "epoch": 2.334613458006345,
+      "grad_norm": 0.4988509374702272,
+      "learning_rate": 1.4238225621282403e-06,
+      "loss": 0.1588,
+      "step": 6991
+    },
+    {
+      "epoch": 2.3349474035732176,
+      "grad_norm": 0.5005557600957211,
+      "learning_rate": 1.4224646686077303e-06,
+      "loss": 0.1569,
+      "step": 6992
+    },
+    {
+      "epoch": 2.3352813491400903,
+      "grad_norm": 0.48763438079425747,
+      "learning_rate": 1.4211073155160544e-06,
+      "loss": 0.158,
+      "step": 6993
+    },
+    {
+      "epoch": 2.3356152947069626,
+      "grad_norm": 0.5314112983450026,
+      "learning_rate": 1.4197505030582588e-06,
+      "loss": 0.164,
+      "step": 6994
+    },
+    {
+      "epoch": 2.3359492402738353,
+      "grad_norm": 0.502506767280615,
+      "learning_rate": 1.4183942314393056e-06,
+      "loss": 0.1553,
+      "step": 6995
+    },
+    {
+      "epoch": 2.336283185840708,
+      "grad_norm": 0.514455411162069,
+      "learning_rate": 1.4170385008640774e-06,
+      "loss": 0.1561,
+      "step": 6996
+    },
+    {
+      "epoch": 2.3366171314075803,
+      "grad_norm": 0.5256261869279392,
+      "learning_rate": 1.4156833115373702e-06,
+      "loss": 0.1522,
+      "step": 6997
+    },
+    {
+      "epoch": 2.336951076974453,
+      "grad_norm": 0.5419599369164668,
+      "learning_rate": 1.4143286636639043e-06,
+      "loss": 0.1648,
+      "step": 6998
+    },
+    {
+      "epoch": 2.3372850225413258,
+      "grad_norm": 0.4543607056514028,
+      "learning_rate": 1.4129745574483123e-06,
+      "loss": 0.1443,
+      "step": 6999
+    },
+    {
+      "epoch": 2.3376189681081985,
+      "grad_norm": 0.5119302019907201,
+      "learning_rate": 1.4116209930951508e-06,
+      "loss": 0.1589,
+      "step": 7000
+    },
+    {
+      "epoch": 2.337952913675071,
+      "grad_norm": 0.5143893370915871,
+      "learning_rate": 1.4102679708088867e-06,
+      "loss": 0.1617,
+      "step": 7001
+    },
+    {
+      "epoch": 2.3382868592419435,
+      "grad_norm": 0.4951445956972631,
+      "learning_rate": 1.4089154907939162e-06,
+      "loss": 0.1552,
+      "step": 7002
+    },
+    {
+      "epoch": 2.338620804808816,
+      "grad_norm": 0.4944981712221224,
+      "learning_rate": 1.4075635532545435e-06,
+      "loss": 0.1471,
+      "step": 7003
+    },
+    {
+      "epoch": 2.338954750375689,
+      "grad_norm": 0.5217606879242871,
+      "learning_rate": 1.4062121583949967e-06,
+      "loss": 0.1527,
+      "step": 7004
+    },
+    {
+      "epoch": 2.339288695942561,
+      "grad_norm": 0.5372747665208512,
+      "learning_rate": 1.4048613064194178e-06,
+      "loss": 0.1639,
+      "step": 7005
+    },
+    {
+      "epoch": 2.339622641509434,
+      "grad_norm": 0.5084540072164366,
+      "learning_rate": 1.4035109975318712e-06,
+      "loss": 0.1522,
+      "step": 7006
+    },
+    {
+      "epoch": 2.3399565870763066,
+      "grad_norm": 0.4915624674536933,
+      "learning_rate": 1.4021612319363326e-06,
+      "loss": 0.1475,
+      "step": 7007
+    },
+    {
+      "epoch": 2.3402905326431793,
+      "grad_norm": 0.5141788309732952,
+      "learning_rate": 1.4008120098367062e-06,
+      "loss": 0.1611,
+      "step": 7008
+    },
+    {
+      "epoch": 2.3406244782100516,
+      "grad_norm": 0.5340670488732328,
+      "learning_rate": 1.3994633314368034e-06,
+      "loss": 0.1635,
+      "step": 7009
+    },
+    {
+      "epoch": 2.3409584237769243,
+      "grad_norm": 0.4839970244154489,
+      "learning_rate": 1.3981151969403606e-06,
+      "loss": 0.1458,
+      "step": 7010
+    },
+    {
+      "epoch": 2.341292369343797,
+      "grad_norm": 0.4786078131814078,
+      "learning_rate": 1.3967676065510266e-06,
+      "loss": 0.145,
+      "step": 7011
+    },
+    {
+      "epoch": 2.3416263149106697,
+      "grad_norm": 0.47065456298226127,
+      "learning_rate": 1.3954205604723742e-06,
+      "loss": 0.1455,
+      "step": 7012
+    },
+    {
+      "epoch": 2.341960260477542,
+      "grad_norm": 0.550572446756861,
+      "learning_rate": 1.3940740589078872e-06,
+      "loss": 0.1606,
+      "step": 7013
+    },
+    {
+      "epoch": 2.3422942060444147,
+      "grad_norm": 0.5654689482348924,
+      "learning_rate": 1.3927281020609712e-06,
+      "loss": 0.172,
+      "step": 7014
+    },
+    {
+      "epoch": 2.3426281516112875,
+      "grad_norm": 0.5893082288878831,
+      "learning_rate": 1.391382690134952e-06,
+      "loss": 0.1689,
+      "step": 7015
+    },
+    {
+      "epoch": 2.3429620971781597,
+      "grad_norm": 0.532586165834147,
+      "learning_rate": 1.3900378233330658e-06,
+      "loss": 0.1589,
+      "step": 7016
+    },
+    {
+      "epoch": 2.3432960427450324,
+      "grad_norm": 0.5434759431816953,
+      "learning_rate": 1.3886935018584719e-06,
+      "loss": 0.1561,
+      "step": 7017
+    },
+    {
+      "epoch": 2.343629988311905,
+      "grad_norm": 0.4986753275861255,
+      "learning_rate": 1.3873497259142483e-06,
+      "loss": 0.1576,
+      "step": 7018
+    },
+    {
+      "epoch": 2.343963933878778,
+      "grad_norm": 0.4850699716446126,
+      "learning_rate": 1.3860064957033847e-06,
+      "loss": 0.1565,
+      "step": 7019
+    },
+    {
+      "epoch": 2.3442978794456506,
+      "grad_norm": 0.4626717680688309,
+      "learning_rate": 1.384663811428793e-06,
+      "loss": 0.1403,
+      "step": 7020
+    },
+    {
+      "epoch": 2.344631825012523,
+      "grad_norm": 0.5748882353001997,
+      "learning_rate": 1.3833216732933035e-06,
+      "loss": 0.1677,
+      "step": 7021
+    },
+    {
+      "epoch": 2.3449657705793956,
+      "grad_norm": 0.4886924233962931,
+      "learning_rate": 1.3819800814996587e-06,
+      "loss": 0.1476,
+      "step": 7022
+    },
+    {
+      "epoch": 2.3452997161462683,
+      "grad_norm": 0.5189394050352707,
+      "learning_rate": 1.3806390362505251e-06,
+      "loss": 0.1625,
+      "step": 7023
+    },
+    {
+      "epoch": 2.3456336617131406,
+      "grad_norm": 0.4929703711875753,
+      "learning_rate": 1.3792985377484796e-06,
+      "loss": 0.153,
+      "step": 7024
+    },
+    {
+      "epoch": 2.3459676072800133,
+      "grad_norm": 0.5074258924655396,
+      "learning_rate": 1.3779585861960226e-06,
+      "loss": 0.1479,
+      "step": 7025
+    },
+    {
+      "epoch": 2.346301552846886,
+      "grad_norm": 0.5344322232369463,
+      "learning_rate": 1.3766191817955699e-06,
+      "loss": 0.1666,
+      "step": 7026
+    },
+    {
+      "epoch": 2.3466354984137587,
+      "grad_norm": 0.5450383787251549,
+      "learning_rate": 1.3752803247494545e-06,
+      "loss": 0.1697,
+      "step": 7027
+    },
+    {
+      "epoch": 2.346969443980631,
+      "grad_norm": 0.5356447852167889,
+      "learning_rate": 1.3739420152599247e-06,
+      "loss": 0.1625,
+      "step": 7028
+    },
+    {
+      "epoch": 2.3473033895475037,
+      "grad_norm": 0.49739392807914606,
+      "learning_rate": 1.37260425352915e-06,
+      "loss": 0.1442,
+      "step": 7029
+    },
+    {
+      "epoch": 2.3476373351143764,
+      "grad_norm": 0.4677708828080009,
+      "learning_rate": 1.3712670397592127e-06,
+      "loss": 0.1455,
+      "step": 7030
+    },
+    {
+      "epoch": 2.347971280681249,
+      "grad_norm": 0.531422867823814,
+      "learning_rate": 1.3699303741521158e-06,
+      "loss": 0.1503,
+      "step": 7031
+    },
+    {
+      "epoch": 2.3483052262481214,
+      "grad_norm": 0.5219314248338266,
+      "learning_rate": 1.3685942569097793e-06,
+      "loss": 0.1666,
+      "step": 7032
+    },
+    {
+      "epoch": 2.348639171814994,
+      "grad_norm": 0.5289721391712382,
+      "learning_rate": 1.3672586882340393e-06,
+      "loss": 0.1646,
+      "step": 7033
+    },
+    {
+      "epoch": 2.348973117381867,
+      "grad_norm": 0.4755425996576312,
+      "learning_rate": 1.3659236683266475e-06,
+      "loss": 0.1514,
+      "step": 7034
+    },
+    {
+      "epoch": 2.349307062948739,
+      "grad_norm": 0.49561200390500754,
+      "learning_rate": 1.3645891973892772e-06,
+      "loss": 0.1493,
+      "step": 7035
+    },
+    {
+      "epoch": 2.349641008515612,
+      "grad_norm": 0.5361377733253501,
+      "learning_rate": 1.3632552756235124e-06,
+      "loss": 0.1635,
+      "step": 7036
+    },
+    {
+      "epoch": 2.3499749540824846,
+      "grad_norm": 0.5245007107369157,
+      "learning_rate": 1.3619219032308594e-06,
+      "loss": 0.1621,
+      "step": 7037
+    },
+    {
+      "epoch": 2.3503088996493573,
+      "grad_norm": 0.5233641249374307,
+      "learning_rate": 1.3605890804127415e-06,
+      "loss": 0.1567,
+      "step": 7038
+    },
+    {
+      "epoch": 2.3506428452162296,
+      "grad_norm": 0.529716471906781,
+      "learning_rate": 1.3592568073704943e-06,
+      "loss": 0.1567,
+      "step": 7039
+    },
+    {
+      "epoch": 2.3509767907831023,
+      "grad_norm": 0.5282219003779071,
+      "learning_rate": 1.3579250843053747e-06,
+      "loss": 0.1522,
+      "step": 7040
+    },
+    {
+      "epoch": 2.351310736349975,
+      "grad_norm": 0.5234598148920355,
+      "learning_rate": 1.3565939114185568e-06,
+      "loss": 0.1503,
+      "step": 7041
+    },
+    {
+      "epoch": 2.3516446819168477,
+      "grad_norm": 0.6006249908501967,
+      "learning_rate": 1.3552632889111266e-06,
+      "loss": 0.1674,
+      "step": 7042
+    },
+    {
+      "epoch": 2.35197862748372,
+      "grad_norm": 0.5012523499275392,
+      "learning_rate": 1.3539332169840918e-06,
+      "loss": 0.1537,
+      "step": 7043
+    },
+    {
+      "epoch": 2.3523125730505927,
+      "grad_norm": 0.47883312000552264,
+      "learning_rate": 1.3526036958383777e-06,
+      "loss": 0.1444,
+      "step": 7044
+    },
+    {
+      "epoch": 2.3526465186174654,
+      "grad_norm": 0.5046804575270472,
+      "learning_rate": 1.35127472567482e-06,
+      "loss": 0.1516,
+      "step": 7045
+    },
+    {
+      "epoch": 2.3529804641843377,
+      "grad_norm": 0.4807311309572237,
+      "learning_rate": 1.3499463066941787e-06,
+      "loss": 0.1466,
+      "step": 7046
+    },
+    {
+      "epoch": 2.3533144097512104,
+      "grad_norm": 0.5251926780869718,
+      "learning_rate": 1.3486184390971246e-06,
+      "loss": 0.1545,
+      "step": 7047
+    },
+    {
+      "epoch": 2.353648355318083,
+      "grad_norm": 0.5442022137332256,
+      "learning_rate": 1.347291123084249e-06,
+      "loss": 0.1689,
+      "step": 7048
+    },
+    {
+      "epoch": 2.353982300884956,
+      "grad_norm": 0.5188350766445398,
+      "learning_rate": 1.3459643588560583e-06,
+      "loss": 0.1597,
+      "step": 7049
+    },
+    {
+      "epoch": 2.3543162464518286,
+      "grad_norm": 0.47155341824463987,
+      "learning_rate": 1.3446381466129777e-06,
+      "loss": 0.1477,
+      "step": 7050
+    },
+    {
+      "epoch": 2.354650192018701,
+      "grad_norm": 0.5474209459200585,
+      "learning_rate": 1.3433124865553437e-06,
+      "loss": 0.1697,
+      "step": 7051
+    },
+    {
+      "epoch": 2.3549841375855736,
+      "grad_norm": 0.5271510936720203,
+      "learning_rate": 1.3419873788834164e-06,
+      "loss": 0.1657,
+      "step": 7052
+    },
+    {
+      "epoch": 2.3553180831524463,
+      "grad_norm": 0.5277962735942314,
+      "learning_rate": 1.3406628237973662e-06,
+      "loss": 0.1608,
+      "step": 7053
+    },
+    {
+      "epoch": 2.3556520287193186,
+      "grad_norm": 0.5556972174352737,
+      "learning_rate": 1.339338821497283e-06,
+      "loss": 0.1625,
+      "step": 7054
+    },
+    {
+      "epoch": 2.3559859742861913,
+      "grad_norm": 0.5248699594279046,
+      "learning_rate": 1.3380153721831745e-06,
+      "loss": 0.1619,
+      "step": 7055
+    },
+    {
+      "epoch": 2.356319919853064,
+      "grad_norm": 0.505305439890515,
+      "learning_rate": 1.3366924760549632e-06,
+      "loss": 0.1605,
+      "step": 7056
+    },
+    {
+      "epoch": 2.3566538654199367,
+      "grad_norm": 0.509624361874994,
+      "learning_rate": 1.3353701333124863e-06,
+      "loss": 0.1558,
+      "step": 7057
+    },
+    {
+      "epoch": 2.356987810986809,
+      "grad_norm": 0.5139996192629374,
+      "learning_rate": 1.3340483441555024e-06,
+      "loss": 0.1569,
+      "step": 7058
+    },
+    {
+      "epoch": 2.3573217565536817,
+      "grad_norm": 0.50576323330744,
+      "learning_rate": 1.3327271087836792e-06,
+      "loss": 0.1528,
+      "step": 7059
+    },
+    {
+      "epoch": 2.3576557021205544,
+      "grad_norm": 0.4936700282126162,
+      "learning_rate": 1.331406427396607e-06,
+      "loss": 0.1513,
+      "step": 7060
+    },
+    {
+      "epoch": 2.357989647687427,
+      "grad_norm": 0.5475069928862222,
+      "learning_rate": 1.3300863001937902e-06,
+      "loss": 0.1588,
+      "step": 7061
+    },
+    {
+      "epoch": 2.3583235932542994,
+      "grad_norm": 0.5259688821837174,
+      "learning_rate": 1.3287667273746513e-06,
+      "loss": 0.1646,
+      "step": 7062
+    },
+    {
+      "epoch": 2.358657538821172,
+      "grad_norm": 0.6091072760868147,
+      "learning_rate": 1.3274477091385241e-06,
+      "loss": 0.1632,
+      "step": 7063
+    },
+    {
+      "epoch": 2.358991484388045,
+      "grad_norm": 0.49664916653441177,
+      "learning_rate": 1.3261292456846648e-06,
+      "loss": 0.1483,
+      "step": 7064
+    },
+    {
+      "epoch": 2.359325429954917,
+      "grad_norm": 0.5428777150582343,
+      "learning_rate": 1.3248113372122395e-06,
+      "loss": 0.1596,
+      "step": 7065
+    },
+    {
+      "epoch": 2.35965937552179,
+      "grad_norm": 0.5426577923297033,
+      "learning_rate": 1.3234939839203358e-06,
+      "loss": 0.1602,
+      "step": 7066
+    },
+    {
+      "epoch": 2.3599933210886626,
+      "grad_norm": 0.5188785999223405,
+      "learning_rate": 1.3221771860079569e-06,
+      "loss": 0.1591,
+      "step": 7067
+    },
+    {
+      "epoch": 2.3603272666555353,
+      "grad_norm": 0.5086994931226321,
+      "learning_rate": 1.3208609436740178e-06,
+      "loss": 0.1522,
+      "step": 7068
+    },
+    {
+      "epoch": 2.360661212222408,
+      "grad_norm": 0.47362864053642506,
+      "learning_rate": 1.3195452571173551e-06,
+      "loss": 0.1467,
+      "step": 7069
+    },
+    {
+      "epoch": 2.3609951577892803,
+      "grad_norm": 0.5264254283456309,
+      "learning_rate": 1.3182301265367154e-06,
+      "loss": 0.1555,
+      "step": 7070
+    },
+    {
+      "epoch": 2.361329103356153,
+      "grad_norm": 0.4742773192417172,
+      "learning_rate": 1.3169155521307664e-06,
+      "loss": 0.1433,
+      "step": 7071
+    },
+    {
+      "epoch": 2.3616630489230257,
+      "grad_norm": 0.5145066895595479,
+      "learning_rate": 1.3156015340980904e-06,
+      "loss": 0.1574,
+      "step": 7072
+    },
+    {
+      "epoch": 2.361996994489898,
+      "grad_norm": 0.5604203763870581,
+      "learning_rate": 1.3142880726371865e-06,
+      "loss": 0.1657,
+      "step": 7073
+    },
+    {
+      "epoch": 2.3623309400567707,
+      "grad_norm": 0.53274867372277,
+      "learning_rate": 1.312975167946466e-06,
+      "loss": 0.1718,
+      "step": 7074
+    },
+    {
+      "epoch": 2.3626648856236434,
+      "grad_norm": 0.53957363798226,
+      "learning_rate": 1.3116628202242603e-06,
+      "loss": 0.1594,
+      "step": 7075
+    },
+    {
+      "epoch": 2.362998831190516,
+      "grad_norm": 0.5006375891688938,
+      "learning_rate": 1.3103510296688137e-06,
+      "loss": 0.1598,
+      "step": 7076
+    },
+    {
+      "epoch": 2.3633327767573884,
+      "grad_norm": 0.48159573025330465,
+      "learning_rate": 1.309039796478288e-06,
+      "loss": 0.1495,
+      "step": 7077
+    },
+    {
+      "epoch": 2.363666722324261,
+      "grad_norm": 0.5301067970927092,
+      "learning_rate": 1.307729120850761e-06,
+      "loss": 0.1578,
+      "step": 7078
+    },
+    {
+      "epoch": 2.364000667891134,
+      "grad_norm": 0.5357552915413548,
+      "learning_rate": 1.306419002984226e-06,
+      "loss": 0.1566,
+      "step": 7079
+    },
+    {
+      "epoch": 2.3643346134580066,
+      "grad_norm": 0.5482638058241193,
+      "learning_rate": 1.3051094430765905e-06,
+      "loss": 0.1611,
+      "step": 7080
+    },
+    {
+      "epoch": 2.364668559024879,
+      "grad_norm": 0.5096740195373038,
+      "learning_rate": 1.3038004413256805e-06,
+      "loss": 0.1604,
+      "step": 7081
+    },
+    {
+      "epoch": 2.3650025045917515,
+      "grad_norm": 0.470098411066193,
+      "learning_rate": 1.3024919979292338e-06,
+      "loss": 0.1474,
+      "step": 7082
+    },
+    {
+      "epoch": 2.3653364501586243,
+      "grad_norm": 0.538932422734086,
+      "learning_rate": 1.3011841130849079e-06,
+      "loss": 0.1491,
+      "step": 7083
+    },
+    {
+      "epoch": 2.3656703957254965,
+      "grad_norm": 0.48937676072754,
+      "learning_rate": 1.2998767869902733e-06,
+      "loss": 0.159,
+      "step": 7084
+    },
+    {
+      "epoch": 2.3660043412923693,
+      "grad_norm": 0.49651132883192295,
+      "learning_rate": 1.2985700198428197e-06,
+      "loss": 0.1492,
+      "step": 7085
+    },
+    {
+      "epoch": 2.366338286859242,
+      "grad_norm": 0.5336066250058544,
+      "learning_rate": 1.2972638118399456e-06,
+      "loss": 0.1509,
+      "step": 7086
+    },
+    {
+      "epoch": 2.3666722324261147,
+      "grad_norm": 0.5022650378372203,
+      "learning_rate": 1.2959581631789725e-06,
+      "loss": 0.1521,
+      "step": 7087
+    },
+    {
+      "epoch": 2.367006177992987,
+      "grad_norm": 0.5256152756288829,
+      "learning_rate": 1.2946530740571316e-06,
+      "loss": 0.1561,
+      "step": 7088
+    },
+    {
+      "epoch": 2.3673401235598597,
+      "grad_norm": 0.5106806485581995,
+      "learning_rate": 1.293348544671572e-06,
+      "loss": 0.1584,
+      "step": 7089
+    },
+    {
+      "epoch": 2.3676740691267324,
+      "grad_norm": 0.4815349175488414,
+      "learning_rate": 1.2920445752193617e-06,
+      "loss": 0.1494,
+      "step": 7090
+    },
+    {
+      "epoch": 2.368008014693605,
+      "grad_norm": 0.49516867886375865,
+      "learning_rate": 1.2907411658974756e-06,
+      "loss": 0.1536,
+      "step": 7091
+    },
+    {
+      "epoch": 2.3683419602604774,
+      "grad_norm": 0.5133737571155627,
+      "learning_rate": 1.2894383169028134e-06,
+      "loss": 0.1514,
+      "step": 7092
+    },
+    {
+      "epoch": 2.36867590582735,
+      "grad_norm": 0.5047295137569292,
+      "learning_rate": 1.2881360284321825e-06,
+      "loss": 0.1537,
+      "step": 7093
+    },
+    {
+      "epoch": 2.369009851394223,
+      "grad_norm": 0.5233980435478595,
+      "learning_rate": 1.2868343006823113e-06,
+      "loss": 0.1617,
+      "step": 7094
+    },
+    {
+      "epoch": 2.369343796961095,
+      "grad_norm": 0.47283027887814555,
+      "learning_rate": 1.2855331338498377e-06,
+      "loss": 0.1528,
+      "step": 7095
+    },
+    {
+      "epoch": 2.369677742527968,
+      "grad_norm": 0.5649361822819848,
+      "learning_rate": 1.2842325281313233e-06,
+      "loss": 0.1699,
+      "step": 7096
+    },
+    {
+      "epoch": 2.3700116880948405,
+      "grad_norm": 0.5228406368655593,
+      "learning_rate": 1.282932483723236e-06,
+      "loss": 0.1571,
+      "step": 7097
+    },
+    {
+      "epoch": 2.3703456336617132,
+      "grad_norm": 0.528258600013772,
+      "learning_rate": 1.2816330008219656e-06,
+      "loss": 0.1673,
+      "step": 7098
+    },
+    {
+      "epoch": 2.370679579228586,
+      "grad_norm": 0.5126074736580588,
+      "learning_rate": 1.280334079623811e-06,
+      "loss": 0.1586,
+      "step": 7099
+    },
+    {
+      "epoch": 2.3710135247954582,
+      "grad_norm": 0.5209165501480114,
+      "learning_rate": 1.2790357203249931e-06,
+      "loss": 0.1612,
+      "step": 7100
+    },
+    {
+      "epoch": 2.371347470362331,
+      "grad_norm": 0.5650474597014316,
+      "learning_rate": 1.2777379231216391e-06,
+      "loss": 0.1616,
+      "step": 7101
+    },
+    {
+      "epoch": 2.3716814159292037,
+      "grad_norm": 0.5315493028470518,
+      "learning_rate": 1.2764406882098035e-06,
+      "loss": 0.1576,
+      "step": 7102
+    },
+    {
+      "epoch": 2.372015361496076,
+      "grad_norm": 0.511469997004763,
+      "learning_rate": 1.2751440157854439e-06,
+      "loss": 0.1546,
+      "step": 7103
+    },
+    {
+      "epoch": 2.3723493070629487,
+      "grad_norm": 0.5397007074652154,
+      "learning_rate": 1.2738479060444408e-06,
+      "loss": 0.1623,
+      "step": 7104
+    },
+    {
+      "epoch": 2.3726832526298214,
+      "grad_norm": 0.47300263395113906,
+      "learning_rate": 1.2725523591825845e-06,
+      "loss": 0.1443,
+      "step": 7105
+    },
+    {
+      "epoch": 2.373017198196694,
+      "grad_norm": 0.49067687314657127,
+      "learning_rate": 1.2712573753955842e-06,
+      "loss": 0.1523,
+      "step": 7106
+    },
+    {
+      "epoch": 2.3733511437635664,
+      "grad_norm": 0.5354579285385755,
+      "learning_rate": 1.2699629548790599e-06,
+      "loss": 0.1634,
+      "step": 7107
+    },
+    {
+      "epoch": 2.373685089330439,
+      "grad_norm": 0.5125476972638018,
+      "learning_rate": 1.2686690978285533e-06,
+      "loss": 0.1618,
+      "step": 7108
+    },
+    {
+      "epoch": 2.374019034897312,
+      "grad_norm": 0.5010154309189855,
+      "learning_rate": 1.267375804439513e-06,
+      "loss": 0.1514,
+      "step": 7109
+    },
+    {
+      "epoch": 2.3743529804641845,
+      "grad_norm": 0.4951422731618649,
+      "learning_rate": 1.2660830749073093e-06,
+      "loss": 0.1505,
+      "step": 7110
+    },
+    {
+      "epoch": 2.374686926031057,
+      "grad_norm": 0.5662049963465066,
+      "learning_rate": 1.2647909094272215e-06,
+      "loss": 0.1686,
+      "step": 7111
+    },
+    {
+      "epoch": 2.3750208715979295,
+      "grad_norm": 0.496822761224137,
+      "learning_rate": 1.2634993081944469e-06,
+      "loss": 0.15,
+      "step": 7112
+    },
+    {
+      "epoch": 2.3753548171648022,
+      "grad_norm": 0.48684393051910746,
+      "learning_rate": 1.2622082714040995e-06,
+      "loss": 0.1555,
+      "step": 7113
+    },
+    {
+      "epoch": 2.3756887627316745,
+      "grad_norm": 0.51894267837415,
+      "learning_rate": 1.2609177992512022e-06,
+      "loss": 0.1644,
+      "step": 7114
+    },
+    {
+      "epoch": 2.3760227082985472,
+      "grad_norm": 0.5292677495026286,
+      "learning_rate": 1.2596278919306993e-06,
+      "loss": 0.1585,
+      "step": 7115
+    },
+    {
+      "epoch": 2.37635665386542,
+      "grad_norm": 0.48845207606014346,
+      "learning_rate": 1.2583385496374428e-06,
+      "loss": 0.152,
+      "step": 7116
+    },
+    {
+      "epoch": 2.3766905994322927,
+      "grad_norm": 0.5379775111433409,
+      "learning_rate": 1.2570497725662067e-06,
+      "loss": 0.1645,
+      "step": 7117
+    },
+    {
+      "epoch": 2.3770245449991654,
+      "grad_norm": 0.5055506542870268,
+      "learning_rate": 1.2557615609116713e-06,
+      "loss": 0.1555,
+      "step": 7118
+    },
+    {
+      "epoch": 2.3773584905660377,
+      "grad_norm": 0.4733144294957525,
+      "learning_rate": 1.254473914868442e-06,
+      "loss": 0.1484,
+      "step": 7119
+    },
+    {
+      "epoch": 2.3776924361329104,
+      "grad_norm": 0.5299212758598031,
+      "learning_rate": 1.2531868346310288e-06,
+      "loss": 0.16,
+      "step": 7120
+    },
+    {
+      "epoch": 2.378026381699783,
+      "grad_norm": 0.5284052191730892,
+      "learning_rate": 1.2519003203938628e-06,
+      "loss": 0.1529,
+      "step": 7121
+    },
+    {
+      "epoch": 2.3783603272666554,
+      "grad_norm": 0.5238906543650232,
+      "learning_rate": 1.2506143723512842e-06,
+      "loss": 0.1557,
+      "step": 7122
+    },
+    {
+      "epoch": 2.378694272833528,
+      "grad_norm": 0.5335370952252194,
+      "learning_rate": 1.2493289906975543e-06,
+      "loss": 0.1533,
+      "step": 7123
+    },
+    {
+      "epoch": 2.379028218400401,
+      "grad_norm": 0.5786503237493286,
+      "learning_rate": 1.2480441756268397e-06,
+      "loss": 0.1715,
+      "step": 7124
+    },
+    {
+      "epoch": 2.3793621639672735,
+      "grad_norm": 0.4843155928013435,
+      "learning_rate": 1.2467599273332332e-06,
+      "loss": 0.156,
+      "step": 7125
+    },
+    {
+      "epoch": 2.379696109534146,
+      "grad_norm": 0.5227159067590956,
+      "learning_rate": 1.245476246010731e-06,
+      "loss": 0.1642,
+      "step": 7126
+    },
+    {
+      "epoch": 2.3800300551010185,
+      "grad_norm": 0.4944621882634703,
+      "learning_rate": 1.244193131853252e-06,
+      "loss": 0.1475,
+      "step": 7127
+    },
+    {
+      "epoch": 2.3803640006678912,
+      "grad_norm": 0.5358983175601851,
+      "learning_rate": 1.2429105850546213e-06,
+      "loss": 0.1548,
+      "step": 7128
+    },
+    {
+      "epoch": 2.380697946234764,
+      "grad_norm": 0.5032449469880587,
+      "learning_rate": 1.241628605808587e-06,
+      "loss": 0.1589,
+      "step": 7129
+    },
+    {
+      "epoch": 2.381031891801636,
+      "grad_norm": 0.5243288444558761,
+      "learning_rate": 1.2403471943088018e-06,
+      "loss": 0.1556,
+      "step": 7130
+    },
+    {
+      "epoch": 2.381365837368509,
+      "grad_norm": 0.5599001289068225,
+      "learning_rate": 1.239066350748845e-06,
+      "loss": 0.1747,
+      "step": 7131
+    },
+    {
+      "epoch": 2.3816997829353816,
+      "grad_norm": 0.5533477090829723,
+      "learning_rate": 1.2377860753221976e-06,
+      "loss": 0.1721,
+      "step": 7132
+    },
+    {
+      "epoch": 2.382033728502254,
+      "grad_norm": 0.551804825022015,
+      "learning_rate": 1.236506368222264e-06,
+      "loss": 0.1633,
+      "step": 7133
+    },
+    {
+      "epoch": 2.3823676740691266,
+      "grad_norm": 0.5217476920386144,
+      "learning_rate": 1.235227229642355e-06,
+      "loss": 0.156,
+      "step": 7134
+    },
+    {
+      "epoch": 2.3827016196359994,
+      "grad_norm": 0.5338330593096421,
+      "learning_rate": 1.2339486597757038e-06,
+      "loss": 0.1665,
+      "step": 7135
+    },
+    {
+      "epoch": 2.383035565202872,
+      "grad_norm": 0.5231766549666796,
+      "learning_rate": 1.2326706588154496e-06,
+      "loss": 0.1622,
+      "step": 7136
+    },
+    {
+      "epoch": 2.3833695107697443,
+      "grad_norm": 0.5193547058090123,
+      "learning_rate": 1.2313932269546518e-06,
+      "loss": 0.1553,
+      "step": 7137
+    },
+    {
+      "epoch": 2.383703456336617,
+      "grad_norm": 0.5268208854231688,
+      "learning_rate": 1.2301163643862817e-06,
+      "loss": 0.1584,
+      "step": 7138
+    },
+    {
+      "epoch": 2.38403740190349,
+      "grad_norm": 0.509012014482607,
+      "learning_rate": 1.2288400713032227e-06,
+      "loss": 0.1524,
+      "step": 7139
+    },
+    {
+      "epoch": 2.3843713474703625,
+      "grad_norm": 0.5645944044633189,
+      "learning_rate": 1.2275643478982762e-06,
+      "loss": 0.1684,
+      "step": 7140
+    },
+    {
+      "epoch": 2.3847052930372348,
+      "grad_norm": 0.5293226597996646,
+      "learning_rate": 1.2262891943641526e-06,
+      "loss": 0.1536,
+      "step": 7141
+    },
+    {
+      "epoch": 2.3850392386041075,
+      "grad_norm": 0.567921584560172,
+      "learning_rate": 1.2250146108934802e-06,
+      "loss": 0.1659,
+      "step": 7142
+    },
+    {
+      "epoch": 2.38537318417098,
+      "grad_norm": 0.5097541913627487,
+      "learning_rate": 1.2237405976787997e-06,
+      "loss": 0.1544,
+      "step": 7143
+    },
+    {
+      "epoch": 2.3857071297378525,
+      "grad_norm": 0.5253260119018623,
+      "learning_rate": 1.2224671549125673e-06,
+      "loss": 0.1627,
+      "step": 7144
+    },
+    {
+      "epoch": 2.386041075304725,
+      "grad_norm": 0.5148031038721778,
+      "learning_rate": 1.2211942827871486e-06,
+      "loss": 0.1573,
+      "step": 7145
+    },
+    {
+      "epoch": 2.386375020871598,
+      "grad_norm": 0.515311953727318,
+      "learning_rate": 1.2199219814948294e-06,
+      "loss": 0.154,
+      "step": 7146
+    },
+    {
+      "epoch": 2.3867089664384706,
+      "grad_norm": 0.5236022358304645,
+      "learning_rate": 1.218650251227802e-06,
+      "loss": 0.1574,
+      "step": 7147
+    },
+    {
+      "epoch": 2.3870429120053434,
+      "grad_norm": 0.5614360461378991,
+      "learning_rate": 1.2173790921781786e-06,
+      "loss": 0.1605,
+      "step": 7148
+    },
+    {
+      "epoch": 2.3873768575722156,
+      "grad_norm": 0.5241197638766266,
+      "learning_rate": 1.2161085045379818e-06,
+      "loss": 0.155,
+      "step": 7149
+    },
+    {
+      "epoch": 2.3877108031390883,
+      "grad_norm": 0.5220473602267451,
+      "learning_rate": 1.214838488499151e-06,
+      "loss": 0.1545,
+      "step": 7150
+    },
+    {
+      "epoch": 2.388044748705961,
+      "grad_norm": 0.5446184370250129,
+      "learning_rate": 1.2135690442535335e-06,
+      "loss": 0.1576,
+      "step": 7151
+    },
+    {
+      "epoch": 2.3883786942728333,
+      "grad_norm": 0.5369404959958738,
+      "learning_rate": 1.2123001719928972e-06,
+      "loss": 0.1645,
+      "step": 7152
+    },
+    {
+      "epoch": 2.388712639839706,
+      "grad_norm": 0.5412541896274068,
+      "learning_rate": 1.211031871908916e-06,
+      "loss": 0.1587,
+      "step": 7153
+    },
+    {
+      "epoch": 2.3890465854065788,
+      "grad_norm": 0.5880386251671798,
+      "learning_rate": 1.2097641441931868e-06,
+      "loss": 0.1785,
+      "step": 7154
+    },
+    {
+      "epoch": 2.3893805309734515,
+      "grad_norm": 0.5251849578213053,
+      "learning_rate": 1.2084969890372111e-06,
+      "loss": 0.1567,
+      "step": 7155
+    },
+    {
+      "epoch": 2.3897144765403238,
+      "grad_norm": 0.5583506662971707,
+      "learning_rate": 1.2072304066324103e-06,
+      "loss": 0.1698,
+      "step": 7156
+    },
+    {
+      "epoch": 2.3900484221071965,
+      "grad_norm": 0.5335369655122735,
+      "learning_rate": 1.205964397170113e-06,
+      "loss": 0.1592,
+      "step": 7157
+    },
+    {
+      "epoch": 2.390382367674069,
+      "grad_norm": 0.5418466727615724,
+      "learning_rate": 1.2046989608415682e-06,
+      "loss": 0.1598,
+      "step": 7158
+    },
+    {
+      "epoch": 2.390716313240942,
+      "grad_norm": 0.4969750469808351,
+      "learning_rate": 1.2034340978379328e-06,
+      "loss": 0.1478,
+      "step": 7159
+    },
+    {
+      "epoch": 2.391050258807814,
+      "grad_norm": 0.6032598631872019,
+      "learning_rate": 1.2021698083502797e-06,
+      "loss": 0.172,
+      "step": 7160
+    },
+    {
+      "epoch": 2.391384204374687,
+      "grad_norm": 0.5109947419002258,
+      "learning_rate": 1.2009060925695965e-06,
+      "loss": 0.146,
+      "step": 7161
+    },
+    {
+      "epoch": 2.3917181499415596,
+      "grad_norm": 0.5009788689546751,
+      "learning_rate": 1.1996429506867797e-06,
+      "loss": 0.1501,
+      "step": 7162
+    },
+    {
+      "epoch": 2.392052095508432,
+      "grad_norm": 0.5335564451937221,
+      "learning_rate": 1.1983803828926438e-06,
+      "loss": 0.1633,
+      "step": 7163
+    },
+    {
+      "epoch": 2.3923860410753046,
+      "grad_norm": 0.5471286776914387,
+      "learning_rate": 1.1971183893779125e-06,
+      "loss": 0.1637,
+      "step": 7164
+    },
+    {
+      "epoch": 2.3927199866421773,
+      "grad_norm": 0.5296431778636541,
+      "learning_rate": 1.1958569703332262e-06,
+      "loss": 0.1612,
+      "step": 7165
+    },
+    {
+      "epoch": 2.39305393220905,
+      "grad_norm": 0.5313318060463695,
+      "learning_rate": 1.1945961259491368e-06,
+      "loss": 0.1544,
+      "step": 7166
+    },
+    {
+      "epoch": 2.3933878777759228,
+      "grad_norm": 0.5573335599867503,
+      "learning_rate": 1.1933358564161108e-06,
+      "loss": 0.1561,
+      "step": 7167
+    },
+    {
+      "epoch": 2.393721823342795,
+      "grad_norm": 0.5278835623914241,
+      "learning_rate": 1.1920761619245246e-06,
+      "loss": 0.1661,
+      "step": 7168
+    },
+    {
+      "epoch": 2.3940557689096678,
+      "grad_norm": 0.5286543182285065,
+      "learning_rate": 1.1908170426646726e-06,
+      "loss": 0.1576,
+      "step": 7169
+    },
+    {
+      "epoch": 2.3943897144765405,
+      "grad_norm": 0.4797109888500716,
+      "learning_rate": 1.189558498826756e-06,
+      "loss": 0.1383,
+      "step": 7170
+    },
+    {
+      "epoch": 2.3947236600434127,
+      "grad_norm": 0.5215268741066937,
+      "learning_rate": 1.1883005306008955e-06,
+      "loss": 0.1531,
+      "step": 7171
+    },
+    {
+      "epoch": 2.3950576056102855,
+      "grad_norm": 0.4951612997717505,
+      "learning_rate": 1.1870431381771203e-06,
+      "loss": 0.152,
+      "step": 7172
+    },
+    {
+      "epoch": 2.395391551177158,
+      "grad_norm": 0.5105466271334883,
+      "learning_rate": 1.185786321745377e-06,
+      "loss": 0.1489,
+      "step": 7173
+    },
+    {
+      "epoch": 2.395725496744031,
+      "grad_norm": 0.557204326375292,
+      "learning_rate": 1.1845300814955192e-06,
+      "loss": 0.1576,
+      "step": 7174
+    },
+    {
+      "epoch": 2.396059442310903,
+      "grad_norm": 0.5630209033895921,
+      "learning_rate": 1.18327441761732e-06,
+      "loss": 0.1585,
+      "step": 7175
+    },
+    {
+      "epoch": 2.396393387877776,
+      "grad_norm": 0.5298445211408632,
+      "learning_rate": 1.1820193303004584e-06,
+      "loss": 0.1627,
+      "step": 7176
+    },
+    {
+      "epoch": 2.3967273334446486,
+      "grad_norm": 0.46722816889909324,
+      "learning_rate": 1.1807648197345327e-06,
+      "loss": 0.1376,
+      "step": 7177
+    },
+    {
+      "epoch": 2.3970612790115213,
+      "grad_norm": 0.5325098603893473,
+      "learning_rate": 1.1795108861090515e-06,
+      "loss": 0.1551,
+      "step": 7178
+    },
+    {
+      "epoch": 2.3973952245783936,
+      "grad_norm": 0.5301349354335696,
+      "learning_rate": 1.1782575296134363e-06,
+      "loss": 0.1519,
+      "step": 7179
+    },
+    {
+      "epoch": 2.3977291701452663,
+      "grad_norm": 0.5145757838651948,
+      "learning_rate": 1.1770047504370197e-06,
+      "loss": 0.1538,
+      "step": 7180
+    },
+    {
+      "epoch": 2.398063115712139,
+      "grad_norm": 0.5300739981678517,
+      "learning_rate": 1.1757525487690513e-06,
+      "loss": 0.1541,
+      "step": 7181
+    },
+    {
+      "epoch": 2.3983970612790113,
+      "grad_norm": 0.5601765735049783,
+      "learning_rate": 1.1745009247986882e-06,
+      "loss": 0.1669,
+      "step": 7182
+    },
+    {
+      "epoch": 2.398731006845884,
+      "grad_norm": 0.5090285498231324,
+      "learning_rate": 1.1732498787150044e-06,
+      "loss": 0.1448,
+      "step": 7183
+    },
+    {
+      "epoch": 2.3990649524127567,
+      "grad_norm": 0.5003809889361199,
+      "learning_rate": 1.171999410706986e-06,
+      "loss": 0.1451,
+      "step": 7184
+    },
+    {
+      "epoch": 2.3993988979796295,
+      "grad_norm": 0.5417432938953705,
+      "learning_rate": 1.1707495209635283e-06,
+      "loss": 0.1557,
+      "step": 7185
+    },
+    {
+      "epoch": 2.3997328435465017,
+      "grad_norm": 0.5604087075100392,
+      "learning_rate": 1.1695002096734454e-06,
+      "loss": 0.1657,
+      "step": 7186
+    },
+    {
+      "epoch": 2.4000667891133745,
+      "grad_norm": 0.5283839371353753,
+      "learning_rate": 1.1682514770254567e-06,
+      "loss": 0.1552,
+      "step": 7187
+    },
+    {
+      "epoch": 2.400400734680247,
+      "grad_norm": 0.5167426973502384,
+      "learning_rate": 1.1670033232081995e-06,
+      "loss": 0.1596,
+      "step": 7188
+    },
+    {
+      "epoch": 2.40073468024712,
+      "grad_norm": 0.4966963801218271,
+      "learning_rate": 1.1657557484102228e-06,
+      "loss": 0.1551,
+      "step": 7189
+    },
+    {
+      "epoch": 2.401068625813992,
+      "grad_norm": 0.5768210088793992,
+      "learning_rate": 1.1645087528199883e-06,
+      "loss": 0.1669,
+      "step": 7190
+    },
+    {
+      "epoch": 2.401402571380865,
+      "grad_norm": 0.5000691629042625,
+      "learning_rate": 1.1632623366258666e-06,
+      "loss": 0.1472,
+      "step": 7191
+    },
+    {
+      "epoch": 2.4017365169477376,
+      "grad_norm": 0.5345663390184785,
+      "learning_rate": 1.162016500016147e-06,
+      "loss": 0.1532,
+      "step": 7192
+    },
+    {
+      "epoch": 2.40207046251461,
+      "grad_norm": 0.5230805429006006,
+      "learning_rate": 1.1607712431790242e-06,
+      "loss": 0.1613,
+      "step": 7193
+    },
+    {
+      "epoch": 2.4024044080814826,
+      "grad_norm": 0.47784139221190725,
+      "learning_rate": 1.15952656630261e-06,
+      "loss": 0.1378,
+      "step": 7194
+    },
+    {
+      "epoch": 2.4027383536483553,
+      "grad_norm": 0.5413001588245746,
+      "learning_rate": 1.158282469574929e-06,
+      "loss": 0.16,
+      "step": 7195
+    },
+    {
+      "epoch": 2.403072299215228,
+      "grad_norm": 0.5172176939927573,
+      "learning_rate": 1.1570389531839165e-06,
+      "loss": 0.1449,
+      "step": 7196
+    },
+    {
+      "epoch": 2.4034062447821007,
+      "grad_norm": 0.5126788675251615,
+      "learning_rate": 1.1557960173174183e-06,
+      "loss": 0.1583,
+      "step": 7197
+    },
+    {
+      "epoch": 2.403740190348973,
+      "grad_norm": 0.5071441617182192,
+      "learning_rate": 1.154553662163197e-06,
+      "loss": 0.1485,
+      "step": 7198
+    },
+    {
+      "epoch": 2.4040741359158457,
+      "grad_norm": 0.5710809229470581,
+      "learning_rate": 1.1533118879089227e-06,
+      "loss": 0.1681,
+      "step": 7199
+    },
+    {
+      "epoch": 2.4044080814827185,
+      "grad_norm": 0.5293626105834608,
+      "learning_rate": 1.1520706947421806e-06,
+      "loss": 0.149,
+      "step": 7200
+    },
+    {
+      "epoch": 2.4047420270495907,
+      "grad_norm": 0.5452941336660879,
+      "learning_rate": 1.1508300828504682e-06,
+      "loss": 0.1697,
+      "step": 7201
+    },
+    {
+      "epoch": 2.4050759726164634,
+      "grad_norm": 0.5166114823495399,
+      "learning_rate": 1.1495900524211955e-06,
+      "loss": 0.1532,
+      "step": 7202
+    },
+    {
+      "epoch": 2.405409918183336,
+      "grad_norm": 0.5471123464355846,
+      "learning_rate": 1.1483506036416814e-06,
+      "loss": 0.1664,
+      "step": 7203
+    },
+    {
+      "epoch": 2.405743863750209,
+      "grad_norm": 0.5141430169803471,
+      "learning_rate": 1.1471117366991613e-06,
+      "loss": 0.1555,
+      "step": 7204
+    },
+    {
+      "epoch": 2.406077809317081,
+      "grad_norm": 0.46243845794224187,
+      "learning_rate": 1.1458734517807785e-06,
+      "loss": 0.1418,
+      "step": 7205
+    },
+    {
+      "epoch": 2.406411754883954,
+      "grad_norm": 0.5162303747719997,
+      "learning_rate": 1.1446357490735921e-06,
+      "loss": 0.1535,
+      "step": 7206
+    },
+    {
+      "epoch": 2.4067457004508266,
+      "grad_norm": 0.48927363828819215,
+      "learning_rate": 1.143398628764572e-06,
+      "loss": 0.1525,
+      "step": 7207
+    },
+    {
+      "epoch": 2.4070796460176993,
+      "grad_norm": 0.5390800009438081,
+      "learning_rate": 1.1421620910405977e-06,
+      "loss": 0.1674,
+      "step": 7208
+    },
+    {
+      "epoch": 2.4074135915845716,
+      "grad_norm": 0.4994447846500062,
+      "learning_rate": 1.1409261360884661e-06,
+      "loss": 0.1584,
+      "step": 7209
+    },
+    {
+      "epoch": 2.4077475371514443,
+      "grad_norm": 0.5197712049684485,
+      "learning_rate": 1.1396907640948785e-06,
+      "loss": 0.1478,
+      "step": 7210
+    },
+    {
+      "epoch": 2.408081482718317,
+      "grad_norm": 0.5135963024209774,
+      "learning_rate": 1.1384559752464553e-06,
+      "loss": 0.1488,
+      "step": 7211
+    },
+    {
+      "epoch": 2.4084154282851893,
+      "grad_norm": 0.531386611449885,
+      "learning_rate": 1.137221769729725e-06,
+      "loss": 0.153,
+      "step": 7212
+    },
+    {
+      "epoch": 2.408749373852062,
+      "grad_norm": 0.5578052539668118,
+      "learning_rate": 1.1359881477311301e-06,
+      "loss": 0.1591,
+      "step": 7213
+    },
+    {
+      "epoch": 2.4090833194189347,
+      "grad_norm": 0.5529075972559461,
+      "learning_rate": 1.1347551094370224e-06,
+      "loss": 0.1704,
+      "step": 7214
+    },
+    {
+      "epoch": 2.4094172649858074,
+      "grad_norm": 0.5441030100011386,
+      "learning_rate": 1.1335226550336676e-06,
+      "loss": 0.1587,
+      "step": 7215
+    },
+    {
+      "epoch": 2.40975121055268,
+      "grad_norm": 0.485355416601119,
+      "learning_rate": 1.1322907847072411e-06,
+      "loss": 0.1419,
+      "step": 7216
+    },
+    {
+      "epoch": 2.4100851561195524,
+      "grad_norm": 0.5148035517585307,
+      "learning_rate": 1.1310594986438339e-06,
+      "loss": 0.1503,
+      "step": 7217
+    },
+    {
+      "epoch": 2.410419101686425,
+      "grad_norm": 0.5021996340564928,
+      "learning_rate": 1.129828797029442e-06,
+      "loss": 0.1545,
+      "step": 7218
+    },
+    {
+      "epoch": 2.410753047253298,
+      "grad_norm": 0.5582216534507548,
+      "learning_rate": 1.128598680049982e-06,
+      "loss": 0.1584,
+      "step": 7219
+    },
+    {
+      "epoch": 2.41108699282017,
+      "grad_norm": 0.5244566063420066,
+      "learning_rate": 1.1273691478912752e-06,
+      "loss": 0.1464,
+      "step": 7220
+    },
+    {
+      "epoch": 2.411420938387043,
+      "grad_norm": 0.5195593318131434,
+      "learning_rate": 1.1261402007390587e-06,
+      "loss": 0.1592,
+      "step": 7221
+    },
+    {
+      "epoch": 2.4117548839539156,
+      "grad_norm": 0.5493008998470231,
+      "learning_rate": 1.1249118387789764e-06,
+      "loss": 0.1627,
+      "step": 7222
+    },
+    {
+      "epoch": 2.4120888295207883,
+      "grad_norm": 0.525214705071319,
+      "learning_rate": 1.12368406219659e-06,
+      "loss": 0.1606,
+      "step": 7223
+    },
+    {
+      "epoch": 2.4124227750876606,
+      "grad_norm": 0.4939289053178912,
+      "learning_rate": 1.1224568711773653e-06,
+      "loss": 0.151,
+      "step": 7224
+    },
+    {
+      "epoch": 2.4127567206545333,
+      "grad_norm": 0.48455079432812664,
+      "learning_rate": 1.1212302659066898e-06,
+      "loss": 0.1501,
+      "step": 7225
+    },
+    {
+      "epoch": 2.413090666221406,
+      "grad_norm": 0.5156418088336056,
+      "learning_rate": 1.1200042465698518e-06,
+      "loss": 0.151,
+      "step": 7226
+    },
+    {
+      "epoch": 2.4134246117882787,
+      "grad_norm": 0.5148384016761979,
+      "learning_rate": 1.1187788133520594e-06,
+      "loss": 0.1642,
+      "step": 7227
+    },
+    {
+      "epoch": 2.413758557355151,
+      "grad_norm": 0.5138997748746464,
+      "learning_rate": 1.1175539664384261e-06,
+      "loss": 0.1587,
+      "step": 7228
+    },
+    {
+      "epoch": 2.4140925029220237,
+      "grad_norm": 0.519066569925353,
+      "learning_rate": 1.1163297060139815e-06,
+      "loss": 0.1533,
+      "step": 7229
+    },
+    {
+      "epoch": 2.4144264484888964,
+      "grad_norm": 0.49248902007645573,
+      "learning_rate": 1.1151060322636625e-06,
+      "loss": 0.1506,
+      "step": 7230
+    },
+    {
+      "epoch": 2.4147603940557687,
+      "grad_norm": 0.5312541778053379,
+      "learning_rate": 1.1138829453723204e-06,
+      "loss": 0.1593,
+      "step": 7231
+    },
+    {
+      "epoch": 2.4150943396226414,
+      "grad_norm": 0.5018480303953734,
+      "learning_rate": 1.112660445524718e-06,
+      "loss": 0.1461,
+      "step": 7232
+    },
+    {
+      "epoch": 2.415428285189514,
+      "grad_norm": 0.5226338743612154,
+      "learning_rate": 1.1114385329055262e-06,
+      "loss": 0.1608,
+      "step": 7233
+    },
+    {
+      "epoch": 2.415762230756387,
+      "grad_norm": 0.5635333501894039,
+      "learning_rate": 1.1102172076993301e-06,
+      "loss": 0.1633,
+      "step": 7234
+    },
+    {
+      "epoch": 2.416096176323259,
+      "grad_norm": 0.5602180181468333,
+      "learning_rate": 1.1089964700906257e-06,
+      "loss": 0.1655,
+      "step": 7235
+    },
+    {
+      "epoch": 2.416430121890132,
+      "grad_norm": 0.5756321012974656,
+      "learning_rate": 1.1077763202638208e-06,
+      "loss": 0.1613,
+      "step": 7236
+    },
+    {
+      "epoch": 2.4167640674570046,
+      "grad_norm": 0.5029427815181746,
+      "learning_rate": 1.106556758403231e-06,
+      "loss": 0.1515,
+      "step": 7237
+    },
+    {
+      "epoch": 2.4170980130238773,
+      "grad_norm": 0.5262956548357789,
+      "learning_rate": 1.105337784693088e-06,
+      "loss": 0.1634,
+      "step": 7238
+    },
+    {
+      "epoch": 2.4174319585907496,
+      "grad_norm": 0.5626068139042897,
+      "learning_rate": 1.1041193993175293e-06,
+      "loss": 0.1579,
+      "step": 7239
+    },
+    {
+      "epoch": 2.4177659041576223,
+      "grad_norm": 0.5082947467016112,
+      "learning_rate": 1.1029016024606093e-06,
+      "loss": 0.1532,
+      "step": 7240
+    },
+    {
+      "epoch": 2.418099849724495,
+      "grad_norm": 0.5297964029071952,
+      "learning_rate": 1.101684394306286e-06,
+      "loss": 0.1582,
+      "step": 7241
+    },
+    {
+      "epoch": 2.4184337952913673,
+      "grad_norm": 0.5296061030264333,
+      "learning_rate": 1.100467775038439e-06,
+      "loss": 0.1555,
+      "step": 7242
+    },
+    {
+      "epoch": 2.41876774085824,
+      "grad_norm": 0.5951302915676688,
+      "learning_rate": 1.099251744840849e-06,
+      "loss": 0.1609,
+      "step": 7243
+    },
+    {
+      "epoch": 2.4191016864251127,
+      "grad_norm": 0.5448858703349814,
+      "learning_rate": 1.0980363038972141e-06,
+      "loss": 0.1512,
+      "step": 7244
+    },
+    {
+      "epoch": 2.4194356319919854,
+      "grad_norm": 0.5240412587293782,
+      "learning_rate": 1.096821452391138e-06,
+      "loss": 0.1472,
+      "step": 7245
+    },
+    {
+      "epoch": 2.419769577558858,
+      "grad_norm": 0.5340001181165048,
+      "learning_rate": 1.0956071905061415e-06,
+      "loss": 0.1524,
+      "step": 7246
+    },
+    {
+      "epoch": 2.4201035231257304,
+      "grad_norm": 0.5621964258233278,
+      "learning_rate": 1.0943935184256487e-06,
+      "loss": 0.162,
+      "step": 7247
+    },
+    {
+      "epoch": 2.420437468692603,
+      "grad_norm": 0.49430836108102416,
+      "learning_rate": 1.093180436333005e-06,
+      "loss": 0.1499,
+      "step": 7248
+    },
+    {
+      "epoch": 2.420771414259476,
+      "grad_norm": 0.49277358371660024,
+      "learning_rate": 1.091967944411456e-06,
+      "loss": 0.1503,
+      "step": 7249
+    },
+    {
+      "epoch": 2.421105359826348,
+      "grad_norm": 0.5566152601905626,
+      "learning_rate": 1.0907560428441666e-06,
+      "loss": 0.163,
+      "step": 7250
+    },
+    {
+      "epoch": 2.421439305393221,
+      "grad_norm": 0.5354411967166895,
+      "learning_rate": 1.0895447318142043e-06,
+      "loss": 0.1592,
+      "step": 7251
+    },
+    {
+      "epoch": 2.4217732509600935,
+      "grad_norm": 0.5178968638492157,
+      "learning_rate": 1.0883340115045566e-06,
+      "loss": 0.1555,
+      "step": 7252
+    },
+    {
+      "epoch": 2.4221071965269663,
+      "grad_norm": 0.5064556197403924,
+      "learning_rate": 1.0871238820981133e-06,
+      "loss": 0.1553,
+      "step": 7253
+    },
+    {
+      "epoch": 2.4224411420938385,
+      "grad_norm": 0.513639948844901,
+      "learning_rate": 1.0859143437776803e-06,
+      "loss": 0.1611,
+      "step": 7254
+    },
+    {
+      "epoch": 2.4227750876607113,
+      "grad_norm": 0.539271199077283,
+      "learning_rate": 1.0847053967259736e-06,
+      "loss": 0.1529,
+      "step": 7255
+    },
+    {
+      "epoch": 2.423109033227584,
+      "grad_norm": 0.52303917750466,
+      "learning_rate": 1.0834970411256167e-06,
+      "loss": 0.1611,
+      "step": 7256
+    },
+    {
+      "epoch": 2.4234429787944567,
+      "grad_norm": 0.5294641295234923,
+      "learning_rate": 1.082289277159147e-06,
+      "loss": 0.1598,
+      "step": 7257
+    },
+    {
+      "epoch": 2.423776924361329,
+      "grad_norm": 0.550231135188987,
+      "learning_rate": 1.0810821050090132e-06,
+      "loss": 0.1721,
+      "step": 7258
+    },
+    {
+      "epoch": 2.4241108699282017,
+      "grad_norm": 0.6018154425280414,
+      "learning_rate": 1.0798755248575694e-06,
+      "loss": 0.1592,
+      "step": 7259
+    },
+    {
+      "epoch": 2.4244448154950744,
+      "grad_norm": 0.5102466501385734,
+      "learning_rate": 1.078669536887086e-06,
+      "loss": 0.1615,
+      "step": 7260
+    },
+    {
+      "epoch": 2.4247787610619467,
+      "grad_norm": 0.5042309590141728,
+      "learning_rate": 1.077464141279742e-06,
+      "loss": 0.1518,
+      "step": 7261
+    },
+    {
+      "epoch": 2.4251127066288194,
+      "grad_norm": 0.5388065888923106,
+      "learning_rate": 1.0762593382176244e-06,
+      "loss": 0.1687,
+      "step": 7262
+    },
+    {
+      "epoch": 2.425446652195692,
+      "grad_norm": 0.48205881440780235,
+      "learning_rate": 1.0750551278827365e-06,
+      "loss": 0.1533,
+      "step": 7263
+    },
+    {
+      "epoch": 2.425780597762565,
+      "grad_norm": 0.5594387794673621,
+      "learning_rate": 1.073851510456984e-06,
+      "loss": 0.1672,
+      "step": 7264
+    },
+    {
+      "epoch": 2.4261145433294375,
+      "grad_norm": 0.4905063271692703,
+      "learning_rate": 1.0726484861221902e-06,
+      "loss": 0.149,
+      "step": 7265
+    },
+    {
+      "epoch": 2.42644848889631,
+      "grad_norm": 0.5205108609752576,
+      "learning_rate": 1.0714460550600859e-06,
+      "loss": 0.1485,
+      "step": 7266
+    },
+    {
+      "epoch": 2.4267824344631825,
+      "grad_norm": 0.5231560925068052,
+      "learning_rate": 1.0702442174523132e-06,
+      "loss": 0.1552,
+      "step": 7267
+    },
+    {
+      "epoch": 2.4271163800300553,
+      "grad_norm": 0.4769453332959195,
+      "learning_rate": 1.0690429734804214e-06,
+      "loss": 0.1434,
+      "step": 7268
+    },
+    {
+      "epoch": 2.4274503255969275,
+      "grad_norm": 0.5312419128658573,
+      "learning_rate": 1.0678423233258755e-06,
+      "loss": 0.1514,
+      "step": 7269
+    },
+    {
+      "epoch": 2.4277842711638002,
+      "grad_norm": 0.5457120778252545,
+      "learning_rate": 1.0666422671700438e-06,
+      "loss": 0.1529,
+      "step": 7270
+    },
+    {
+      "epoch": 2.428118216730673,
+      "grad_norm": 0.5443040289592945,
+      "learning_rate": 1.065442805194214e-06,
+      "loss": 0.1643,
+      "step": 7271
+    },
+    {
+      "epoch": 2.4284521622975457,
+      "grad_norm": 0.5274373283039757,
+      "learning_rate": 1.0642439375795748e-06,
+      "loss": 0.1488,
+      "step": 7272
+    },
+    {
+      "epoch": 2.428786107864418,
+      "grad_norm": 0.5390642593642723,
+      "learning_rate": 1.0630456645072324e-06,
+      "loss": 0.1586,
+      "step": 7273
+    },
+    {
+      "epoch": 2.4291200534312907,
+      "grad_norm": 0.5089151199607201,
+      "learning_rate": 1.0618479861581971e-06,
+      "loss": 0.1537,
+      "step": 7274
+    },
+    {
+      "epoch": 2.4294539989981634,
+      "grad_norm": 0.5258576186084203,
+      "learning_rate": 1.060650902713395e-06,
+      "loss": 0.1556,
+      "step": 7275
+    },
+    {
+      "epoch": 2.429787944565036,
+      "grad_norm": 0.5113060552599415,
+      "learning_rate": 1.0594544143536572e-06,
+      "loss": 0.1517,
+      "step": 7276
+    },
+    {
+      "epoch": 2.4301218901319084,
+      "grad_norm": 0.5072388702252578,
+      "learning_rate": 1.0582585212597286e-06,
+      "loss": 0.1499,
+      "step": 7277
+    },
+    {
+      "epoch": 2.430455835698781,
+      "grad_norm": 0.5557065639821541,
+      "learning_rate": 1.0570632236122641e-06,
+      "loss": 0.1621,
+      "step": 7278
+    },
+    {
+      "epoch": 2.430789781265654,
+      "grad_norm": 0.4741104501250161,
+      "learning_rate": 1.0558685215918246e-06,
+      "loss": 0.139,
+      "step": 7279
+    },
+    {
+      "epoch": 2.431123726832526,
+      "grad_norm": 0.5313610251633167,
+      "learning_rate": 1.0546744153788858e-06,
+      "loss": 0.1554,
+      "step": 7280
+    },
+    {
+      "epoch": 2.431457672399399,
+      "grad_norm": 0.5144511403714003,
+      "learning_rate": 1.0534809051538324e-06,
+      "loss": 0.1471,
+      "step": 7281
+    },
+    {
+      "epoch": 2.4317916179662715,
+      "grad_norm": 0.5905335390702079,
+      "learning_rate": 1.0522879910969563e-06,
+      "loss": 0.1742,
+      "step": 7282
+    },
+    {
+      "epoch": 2.4321255635331442,
+      "grad_norm": 0.49916027008308816,
+      "learning_rate": 1.0510956733884614e-06,
+      "loss": 0.1487,
+      "step": 7283
+    },
+    {
+      "epoch": 2.4324595091000165,
+      "grad_norm": 0.5277947791001935,
+      "learning_rate": 1.0499039522084637e-06,
+      "loss": 0.1534,
+      "step": 7284
+    },
+    {
+      "epoch": 2.4327934546668892,
+      "grad_norm": 0.5430657315225573,
+      "learning_rate": 1.0487128277369829e-06,
+      "loss": 0.1649,
+      "step": 7285
+    },
+    {
+      "epoch": 2.433127400233762,
+      "grad_norm": 0.539630403103098,
+      "learning_rate": 1.0475223001539564e-06,
+      "loss": 0.1602,
+      "step": 7286
+    },
+    {
+      "epoch": 2.4334613458006347,
+      "grad_norm": 0.5264937727044785,
+      "learning_rate": 1.0463323696392236e-06,
+      "loss": 0.1561,
+      "step": 7287
+    },
+    {
+      "epoch": 2.433795291367507,
+      "grad_norm": 0.47187721488243567,
+      "learning_rate": 1.0451430363725395e-06,
+      "loss": 0.1526,
+      "step": 7288
+    },
+    {
+      "epoch": 2.4341292369343797,
+      "grad_norm": 0.532862403776023,
+      "learning_rate": 1.043954300533566e-06,
+      "loss": 0.1615,
+      "step": 7289
+    },
+    {
+      "epoch": 2.4344631825012524,
+      "grad_norm": 0.540835650588637,
+      "learning_rate": 1.0427661623018786e-06,
+      "loss": 0.1648,
+      "step": 7290
+    },
+    {
+      "epoch": 2.4347971280681246,
+      "grad_norm": 0.5395899185409637,
+      "learning_rate": 1.0415786218569557e-06,
+      "loss": 0.1655,
+      "step": 7291
+    },
+    {
+      "epoch": 2.4351310736349974,
+      "grad_norm": 0.5003448121019961,
+      "learning_rate": 1.0403916793781922e-06,
+      "loss": 0.1498,
+      "step": 7292
+    },
+    {
+      "epoch": 2.43546501920187,
+      "grad_norm": 0.4713027523101309,
+      "learning_rate": 1.0392053350448867e-06,
+      "loss": 0.14,
+      "step": 7293
+    },
+    {
+      "epoch": 2.435798964768743,
+      "grad_norm": 0.49901626838835894,
+      "learning_rate": 1.0380195890362527e-06,
+      "loss": 0.1591,
+      "step": 7294
+    },
+    {
+      "epoch": 2.4361329103356155,
+      "grad_norm": 0.4962709060760146,
+      "learning_rate": 1.0368344415314101e-06,
+      "loss": 0.1559,
+      "step": 7295
+    },
+    {
+      "epoch": 2.436466855902488,
+      "grad_norm": 0.5031119480742257,
+      "learning_rate": 1.0356498927093916e-06,
+      "loss": 0.1496,
+      "step": 7296
+    },
+    {
+      "epoch": 2.4368008014693605,
+      "grad_norm": 0.4904804267379514,
+      "learning_rate": 1.0344659427491343e-06,
+      "loss": 0.1566,
+      "step": 7297
+    },
+    {
+      "epoch": 2.4371347470362332,
+      "grad_norm": 0.5219797130767747,
+      "learning_rate": 1.0332825918294898e-06,
+      "loss": 0.1554,
+      "step": 7298
+    },
+    {
+      "epoch": 2.4374686926031055,
+      "grad_norm": 0.5189968707324387,
+      "learning_rate": 1.0320998401292154e-06,
+      "loss": 0.1472,
+      "step": 7299
+    },
+    {
+      "epoch": 2.437802638169978,
+      "grad_norm": 0.5253702042666821,
+      "learning_rate": 1.0309176878269806e-06,
+      "loss": 0.1489,
+      "step": 7300
+    },
+    {
+      "epoch": 2.438136583736851,
+      "grad_norm": 0.49138094461283727,
+      "learning_rate": 1.0297361351013646e-06,
+      "loss": 0.1465,
+      "step": 7301
+    },
+    {
+      "epoch": 2.4384705293037237,
+      "grad_norm": 0.5309848647784869,
+      "learning_rate": 1.028555182130853e-06,
+      "loss": 0.1605,
+      "step": 7302
+    },
+    {
+      "epoch": 2.438804474870596,
+      "grad_norm": 0.538951743332172,
+      "learning_rate": 1.027374829093843e-06,
+      "loss": 0.1658,
+      "step": 7303
+    },
+    {
+      "epoch": 2.4391384204374686,
+      "grad_norm": 0.5411757444422154,
+      "learning_rate": 1.0261950761686423e-06,
+      "loss": 0.1581,
+      "step": 7304
+    },
+    {
+      "epoch": 2.4394723660043414,
+      "grad_norm": 0.5200715406012469,
+      "learning_rate": 1.0250159235334645e-06,
+      "loss": 0.158,
+      "step": 7305
+    },
+    {
+      "epoch": 2.439806311571214,
+      "grad_norm": 0.558830853232633,
+      "learning_rate": 1.0238373713664351e-06,
+      "loss": 0.1662,
+      "step": 7306
+    },
+    {
+      "epoch": 2.4401402571380864,
+      "grad_norm": 0.4868370076213395,
+      "learning_rate": 1.0226594198455903e-06,
+      "loss": 0.146,
+      "step": 7307
+    },
+    {
+      "epoch": 2.440474202704959,
+      "grad_norm": 0.49711149600547216,
+      "learning_rate": 1.0214820691488698e-06,
+      "loss": 0.1514,
+      "step": 7308
+    },
+    {
+      "epoch": 2.440808148271832,
+      "grad_norm": 0.5147765431980813,
+      "learning_rate": 1.02030531945413e-06,
+      "loss": 0.1631,
+      "step": 7309
+    },
+    {
+      "epoch": 2.441142093838704,
+      "grad_norm": 0.5618697583716376,
+      "learning_rate": 1.0191291709391298e-06,
+      "loss": 0.1437,
+      "step": 7310
+    },
+    {
+      "epoch": 2.441476039405577,
+      "grad_norm": 0.5719539944950716,
+      "learning_rate": 1.0179536237815413e-06,
+      "loss": 0.1691,
+      "step": 7311
+    },
+    {
+      "epoch": 2.4418099849724495,
+      "grad_norm": 0.5234878850474092,
+      "learning_rate": 1.016778678158945e-06,
+      "loss": 0.155,
+      "step": 7312
+    },
+    {
+      "epoch": 2.442143930539322,
+      "grad_norm": 0.5452399489884633,
+      "learning_rate": 1.015604334248832e-06,
+      "loss": 0.1589,
+      "step": 7313
+    },
+    {
+      "epoch": 2.442477876106195,
+      "grad_norm": 0.5534133198761949,
+      "learning_rate": 1.0144305922285975e-06,
+      "loss": 0.1523,
+      "step": 7314
+    },
+    {
+      "epoch": 2.442811821673067,
+      "grad_norm": 0.5535102460042848,
+      "learning_rate": 1.0132574522755518e-06,
+      "loss": 0.1592,
+      "step": 7315
+    },
+    {
+      "epoch": 2.44314576723994,
+      "grad_norm": 0.5204726820983054,
+      "learning_rate": 1.0120849145669093e-06,
+      "loss": 0.1534,
+      "step": 7316
+    },
+    {
+      "epoch": 2.4434797128068126,
+      "grad_norm": 0.5244522290512145,
+      "learning_rate": 1.010912979279796e-06,
+      "loss": 0.161,
+      "step": 7317
+    },
+    {
+      "epoch": 2.443813658373685,
+      "grad_norm": 0.4986635964399862,
+      "learning_rate": 1.009741646591248e-06,
+      "loss": 0.1537,
+      "step": 7318
+    },
+    {
+      "epoch": 2.4441476039405576,
+      "grad_norm": 0.5206734432989935,
+      "learning_rate": 1.0085709166782088e-06,
+      "loss": 0.1612,
+      "step": 7319
+    },
+    {
+      "epoch": 2.4444815495074304,
+      "grad_norm": 0.5336710529700994,
+      "learning_rate": 1.0074007897175291e-06,
+      "loss": 0.1531,
+      "step": 7320
+    },
+    {
+      "epoch": 2.444815495074303,
+      "grad_norm": 0.5533500819209353,
+      "learning_rate": 1.0062312658859723e-06,
+      "loss": 0.1611,
+      "step": 7321
+    },
+    {
+      "epoch": 2.4451494406411753,
+      "grad_norm": 0.5658983196079189,
+      "learning_rate": 1.0050623453602075e-06,
+      "loss": 0.1667,
+      "step": 7322
+    },
+    {
+      "epoch": 2.445483386208048,
+      "grad_norm": 0.5287095854597083,
+      "learning_rate": 1.0038940283168136e-06,
+      "loss": 0.1505,
+      "step": 7323
+    },
+    {
+      "epoch": 2.4458173317749208,
+      "grad_norm": 0.5375178265696032,
+      "learning_rate": 1.0027263149322797e-06,
+      "loss": 0.1637,
+      "step": 7324
+    },
+    {
+      "epoch": 2.4461512773417935,
+      "grad_norm": 0.47445833611149235,
+      "learning_rate": 1.001559205383003e-06,
+      "loss": 0.1446,
+      "step": 7325
+    },
+    {
+      "epoch": 2.4464852229086658,
+      "grad_norm": 0.4768203964078314,
+      "learning_rate": 1.000392699845288e-06,
+      "loss": 0.1345,
+      "step": 7326
+    },
+    {
+      "epoch": 2.4468191684755385,
+      "grad_norm": 0.5206013386975504,
+      "learning_rate": 9.992267984953503e-07,
+      "loss": 0.1528,
+      "step": 7327
+    },
+    {
+      "epoch": 2.447153114042411,
+      "grad_norm": 0.5182522042264855,
+      "learning_rate": 9.98061501509311e-07,
+      "loss": 0.1566,
+      "step": 7328
+    },
+    {
+      "epoch": 2.4474870596092835,
+      "grad_norm": 0.505189403897211,
+      "learning_rate": 9.968968090632032e-07,
+      "loss": 0.1522,
+      "step": 7329
+    },
+    {
+      "epoch": 2.447821005176156,
+      "grad_norm": 0.5362073600754579,
+      "learning_rate": 9.957327213329687e-07,
+      "loss": 0.1546,
+      "step": 7330
+    },
+    {
+      "epoch": 2.448154950743029,
+      "grad_norm": 0.5102731648397336,
+      "learning_rate": 9.945692384944544e-07,
+      "loss": 0.1461,
+      "step": 7331
+    },
+    {
+      "epoch": 2.4484888963099016,
+      "grad_norm": 0.5226654494134502,
+      "learning_rate": 9.934063607234202e-07,
+      "loss": 0.1532,
+      "step": 7332
+    },
+    {
+      "epoch": 2.448822841876774,
+      "grad_norm": 0.5230643942856512,
+      "learning_rate": 9.922440881955298e-07,
+      "loss": 0.1523,
+      "step": 7333
+    },
+    {
+      "epoch": 2.4491567874436466,
+      "grad_norm": 0.5655913741188585,
+      "learning_rate": 9.910824210863611e-07,
+      "loss": 0.1639,
+      "step": 7334
+    },
+    {
+      "epoch": 2.4494907330105193,
+      "grad_norm": 0.5386421493301948,
+      "learning_rate": 9.899213595713935e-07,
+      "loss": 0.1628,
+      "step": 7335
+    },
+    {
+      "epoch": 2.449824678577392,
+      "grad_norm": 0.49776939645132456,
+      "learning_rate": 9.887609038260243e-07,
+      "loss": 0.1549,
+      "step": 7336
+    },
+    {
+      "epoch": 2.4501586241442643,
+      "grad_norm": 0.558716304445521,
+      "learning_rate": 9.876010540255504e-07,
+      "loss": 0.1635,
+      "step": 7337
+    },
+    {
+      "epoch": 2.450492569711137,
+      "grad_norm": 0.4917972097451291,
+      "learning_rate": 9.86441810345183e-07,
+      "loss": 0.1527,
+      "step": 7338
+    },
+    {
+      "epoch": 2.4508265152780098,
+      "grad_norm": 0.5377757832010998,
+      "learning_rate": 9.852831729600365e-07,
+      "loss": 0.162,
+      "step": 7339
+    },
+    {
+      "epoch": 2.451160460844882,
+      "grad_norm": 0.4868757826124942,
+      "learning_rate": 9.841251420451398e-07,
+      "loss": 0.1444,
+      "step": 7340
+    },
+    {
+      "epoch": 2.4514944064117548,
+      "grad_norm": 0.5439255876381551,
+      "learning_rate": 9.829677177754231e-07,
+      "loss": 0.1568,
+      "step": 7341
+    },
+    {
+      "epoch": 2.4518283519786275,
+      "grad_norm": 0.5411457197387434,
+      "learning_rate": 9.818109003257348e-07,
+      "loss": 0.1649,
+      "step": 7342
+    },
+    {
+      "epoch": 2.4521622975455,
+      "grad_norm": 0.5373406277835615,
+      "learning_rate": 9.806546898708213e-07,
+      "loss": 0.1587,
+      "step": 7343
+    },
+    {
+      "epoch": 2.452496243112373,
+      "grad_norm": 0.5184857587685223,
+      "learning_rate": 9.794990865853444e-07,
+      "loss": 0.1553,
+      "step": 7344
+    },
+    {
+      "epoch": 2.452830188679245,
+      "grad_norm": 0.4972829968409121,
+      "learning_rate": 9.783440906438686e-07,
+      "loss": 0.1518,
+      "step": 7345
+    },
+    {
+      "epoch": 2.453164134246118,
+      "grad_norm": 0.5263298934429176,
+      "learning_rate": 9.771897022208732e-07,
+      "loss": 0.1605,
+      "step": 7346
+    },
+    {
+      "epoch": 2.4534980798129906,
+      "grad_norm": 0.5039481262297244,
+      "learning_rate": 9.760359214907372e-07,
+      "loss": 0.1489,
+      "step": 7347
+    },
+    {
+      "epoch": 2.453832025379863,
+      "grad_norm": 0.5540379369889635,
+      "learning_rate": 9.74882748627759e-07,
+      "loss": 0.1582,
+      "step": 7348
+    },
+    {
+      "epoch": 2.4541659709467356,
+      "grad_norm": 0.5561596840406983,
+      "learning_rate": 9.737301838061342e-07,
+      "loss": 0.1647,
+      "step": 7349
+    },
+    {
+      "epoch": 2.4544999165136083,
+      "grad_norm": 0.5045374898134211,
+      "learning_rate": 9.725782271999744e-07,
+      "loss": 0.1477,
+      "step": 7350
+    },
+    {
+      "epoch": 2.454833862080481,
+      "grad_norm": 0.49671437127259194,
+      "learning_rate": 9.714268789832937e-07,
+      "loss": 0.1501,
+      "step": 7351
+    },
+    {
+      "epoch": 2.4551678076473533,
+      "grad_norm": 0.5284559165780311,
+      "learning_rate": 9.702761393300176e-07,
+      "loss": 0.1578,
+      "step": 7352
+    },
+    {
+      "epoch": 2.455501753214226,
+      "grad_norm": 0.5117383212099872,
+      "learning_rate": 9.691260084139802e-07,
+      "loss": 0.1496,
+      "step": 7353
+    },
+    {
+      "epoch": 2.4558356987810988,
+      "grad_norm": 0.526036454760328,
+      "learning_rate": 9.679764864089203e-07,
+      "loss": 0.1562,
+      "step": 7354
+    },
+    {
+      "epoch": 2.4561696443479715,
+      "grad_norm": 0.49859254666734404,
+      "learning_rate": 9.668275734884885e-07,
+      "loss": 0.1517,
+      "step": 7355
+    },
+    {
+      "epoch": 2.4565035899148437,
+      "grad_norm": 0.5353355704871284,
+      "learning_rate": 9.656792698262402e-07,
+      "loss": 0.1527,
+      "step": 7356
+    },
+    {
+      "epoch": 2.4568375354817165,
+      "grad_norm": 0.5062378256192873,
+      "learning_rate": 9.645315755956413e-07,
+      "loss": 0.1534,
+      "step": 7357
+    },
+    {
+      "epoch": 2.457171481048589,
+      "grad_norm": 0.46223832301303525,
+      "learning_rate": 9.633844909700618e-07,
+      "loss": 0.1401,
+      "step": 7358
+    },
+    {
+      "epoch": 2.4575054266154615,
+      "grad_norm": 0.5524054906187581,
+      "learning_rate": 9.622380161227873e-07,
+      "loss": 0.161,
+      "step": 7359
+    },
+    {
+      "epoch": 2.457839372182334,
+      "grad_norm": 0.5056246623447084,
+      "learning_rate": 9.61092151227002e-07,
+      "loss": 0.1548,
+      "step": 7360
+    },
+    {
+      "epoch": 2.458173317749207,
+      "grad_norm": 0.5121677065114587,
+      "learning_rate": 9.599468964558051e-07,
+      "loss": 0.1459,
+      "step": 7361
+    },
+    {
+      "epoch": 2.4585072633160796,
+      "grad_norm": 0.5588974512360972,
+      "learning_rate": 9.588022519821983e-07,
+      "loss": 0.161,
+      "step": 7362
+    },
+    {
+      "epoch": 2.4588412088829523,
+      "grad_norm": 0.5641228129018973,
+      "learning_rate": 9.576582179790967e-07,
+      "loss": 0.1607,
+      "step": 7363
+    },
+    {
+      "epoch": 2.4591751544498246,
+      "grad_norm": 0.5565054266004965,
+      "learning_rate": 9.565147946193149e-07,
+      "loss": 0.1578,
+      "step": 7364
+    },
+    {
+      "epoch": 2.4595091000166973,
+      "grad_norm": 0.5954752685742998,
+      "learning_rate": 9.553719820755869e-07,
+      "loss": 0.1652,
+      "step": 7365
+    },
+    {
+      "epoch": 2.45984304558357,
+      "grad_norm": 0.5018003233841967,
+      "learning_rate": 9.542297805205436e-07,
+      "loss": 0.1422,
+      "step": 7366
+    },
+    {
+      "epoch": 2.4601769911504423,
+      "grad_norm": 0.5505488375112121,
+      "learning_rate": 9.530881901267308e-07,
+      "loss": 0.1484,
+      "step": 7367
+    },
+    {
+      "epoch": 2.460510936717315,
+      "grad_norm": 0.5195725678865587,
+      "learning_rate": 9.519472110665967e-07,
+      "loss": 0.1467,
+      "step": 7368
+    },
+    {
+      "epoch": 2.4608448822841877,
+      "grad_norm": 0.5151373816065521,
+      "learning_rate": 9.508068435125012e-07,
+      "loss": 0.1553,
+      "step": 7369
+    },
+    {
+      "epoch": 2.4611788278510605,
+      "grad_norm": 0.5281539463812102,
+      "learning_rate": 9.496670876367076e-07,
+      "loss": 0.1568,
+      "step": 7370
+    },
+    {
+      "epoch": 2.4615127734179327,
+      "grad_norm": 0.5334649152791151,
+      "learning_rate": 9.485279436113942e-07,
+      "loss": 0.1671,
+      "step": 7371
+    },
+    {
+      "epoch": 2.4618467189848054,
+      "grad_norm": 0.541217305277803,
+      "learning_rate": 9.473894116086379e-07,
+      "loss": 0.1582,
+      "step": 7372
+    },
+    {
+      "epoch": 2.462180664551678,
+      "grad_norm": 0.4819910188513477,
+      "learning_rate": 9.462514918004301e-07,
+      "loss": 0.1401,
+      "step": 7373
+    },
+    {
+      "epoch": 2.462514610118551,
+      "grad_norm": 0.5450310802132138,
+      "learning_rate": 9.451141843586647e-07,
+      "loss": 0.1556,
+      "step": 7374
+    },
+    {
+      "epoch": 2.462848555685423,
+      "grad_norm": 0.5413919382336527,
+      "learning_rate": 9.439774894551479e-07,
+      "loss": 0.1603,
+      "step": 7375
+    },
+    {
+      "epoch": 2.463182501252296,
+      "grad_norm": 0.5547122615105633,
+      "learning_rate": 9.428414072615877e-07,
+      "loss": 0.1618,
+      "step": 7376
+    },
+    {
+      "epoch": 2.4635164468191686,
+      "grad_norm": 0.5160881731381368,
+      "learning_rate": 9.417059379496047e-07,
+      "loss": 0.1609,
+      "step": 7377
+    },
+    {
+      "epoch": 2.463850392386041,
+      "grad_norm": 0.5382470923154085,
+      "learning_rate": 9.40571081690726e-07,
+      "loss": 0.1624,
+      "step": 7378
+    },
+    {
+      "epoch": 2.4641843379529136,
+      "grad_norm": 0.5136862931411745,
+      "learning_rate": 9.394368386563823e-07,
+      "loss": 0.1537,
+      "step": 7379
+    },
+    {
+      "epoch": 2.4645182835197863,
+      "grad_norm": 0.5130561945345334,
+      "learning_rate": 9.383032090179173e-07,
+      "loss": 0.1595,
+      "step": 7380
+    },
+    {
+      "epoch": 2.464852229086659,
+      "grad_norm": 0.4979229387087476,
+      "learning_rate": 9.371701929465759e-07,
+      "loss": 0.1538,
+      "step": 7381
+    },
+    {
+      "epoch": 2.4651861746535313,
+      "grad_norm": 0.5288085451267187,
+      "learning_rate": 9.360377906135148e-07,
+      "loss": 0.1653,
+      "step": 7382
+    },
+    {
+      "epoch": 2.465520120220404,
+      "grad_norm": 0.5246843887272186,
+      "learning_rate": 9.349060021897976e-07,
+      "loss": 0.1599,
+      "step": 7383
+    },
+    {
+      "epoch": 2.4658540657872767,
+      "grad_norm": 0.5128075936979838,
+      "learning_rate": 9.337748278463948e-07,
+      "loss": 0.1592,
+      "step": 7384
+    },
+    {
+      "epoch": 2.4661880113541494,
+      "grad_norm": 0.5244124581207874,
+      "learning_rate": 9.326442677541813e-07,
+      "loss": 0.1576,
+      "step": 7385
+    },
+    {
+      "epoch": 2.4665219569210217,
+      "grad_norm": 0.5204631164355119,
+      "learning_rate": 9.31514322083944e-07,
+      "loss": 0.153,
+      "step": 7386
+    },
+    {
+      "epoch": 2.4668559024878944,
+      "grad_norm": 0.5045026609152344,
+      "learning_rate": 9.303849910063717e-07,
+      "loss": 0.1523,
+      "step": 7387
+    },
+    {
+      "epoch": 2.467189848054767,
+      "grad_norm": 0.5230456092218726,
+      "learning_rate": 9.292562746920647e-07,
+      "loss": 0.1592,
+      "step": 7388
+    },
+    {
+      "epoch": 2.4675237936216394,
+      "grad_norm": 0.5378631540624517,
+      "learning_rate": 9.281281733115288e-07,
+      "loss": 0.1546,
+      "step": 7389
+    },
+    {
+      "epoch": 2.467857739188512,
+      "grad_norm": 0.5030886402668243,
+      "learning_rate": 9.270006870351789e-07,
+      "loss": 0.1416,
+      "step": 7390
+    },
+    {
+      "epoch": 2.468191684755385,
+      "grad_norm": 0.5286770365980373,
+      "learning_rate": 9.258738160333314e-07,
+      "loss": 0.1533,
+      "step": 7391
+    },
+    {
+      "epoch": 2.4685256303222576,
+      "grad_norm": 0.5174511810173182,
+      "learning_rate": 9.247475604762168e-07,
+      "loss": 0.1629,
+      "step": 7392
+    },
+    {
+      "epoch": 2.4688595758891303,
+      "grad_norm": 0.5775562479058309,
+      "learning_rate": 9.236219205339647e-07,
+      "loss": 0.1591,
+      "step": 7393
+    },
+    {
+      "epoch": 2.4691935214560026,
+      "grad_norm": 0.5530670682386785,
+      "learning_rate": 9.224968963766223e-07,
+      "loss": 0.1617,
+      "step": 7394
+    },
+    {
+      "epoch": 2.4695274670228753,
+      "grad_norm": 0.5588841493266121,
+      "learning_rate": 9.213724881741337e-07,
+      "loss": 0.1697,
+      "step": 7395
+    },
+    {
+      "epoch": 2.469861412589748,
+      "grad_norm": 0.496013975272932,
+      "learning_rate": 9.202486960963559e-07,
+      "loss": 0.1531,
+      "step": 7396
+    },
+    {
+      "epoch": 2.4701953581566203,
+      "grad_norm": 0.5076710510224289,
+      "learning_rate": 9.191255203130489e-07,
+      "loss": 0.1587,
+      "step": 7397
+    },
+    {
+      "epoch": 2.470529303723493,
+      "grad_norm": 0.5020408243905019,
+      "learning_rate": 9.18002960993884e-07,
+      "loss": 0.1518,
+      "step": 7398
+    },
+    {
+      "epoch": 2.4708632492903657,
+      "grad_norm": 0.5348771472192876,
+      "learning_rate": 9.168810183084348e-07,
+      "loss": 0.1601,
+      "step": 7399
+    },
+    {
+      "epoch": 2.4711971948572384,
+      "grad_norm": 0.5044766853939331,
+      "learning_rate": 9.157596924261847e-07,
+      "loss": 0.1542,
+      "step": 7400
+    },
+    {
+      "epoch": 2.4715311404241107,
+      "grad_norm": 0.5757230051543882,
+      "learning_rate": 9.146389835165248e-07,
+      "loss": 0.1769,
+      "step": 7401
+    },
+    {
+      "epoch": 2.4718650859909834,
+      "grad_norm": 0.5052922924503145,
+      "learning_rate": 9.135188917487487e-07,
+      "loss": 0.1507,
+      "step": 7402
+    },
+    {
+      "epoch": 2.472199031557856,
+      "grad_norm": 0.5365350669677343,
+      "learning_rate": 9.12399417292062e-07,
+      "loss": 0.1488,
+      "step": 7403
+    },
+    {
+      "epoch": 2.472532977124729,
+      "grad_norm": 0.507603981314568,
+      "learning_rate": 9.112805603155716e-07,
+      "loss": 0.1534,
+      "step": 7404
+    },
+    {
+      "epoch": 2.472866922691601,
+      "grad_norm": 0.8258984417878298,
+      "learning_rate": 9.101623209882965e-07,
+      "loss": 0.157,
+      "step": 7405
+    },
+    {
+      "epoch": 2.473200868258474,
+      "grad_norm": 0.5625468162092421,
+      "learning_rate": 9.090446994791585e-07,
+      "loss": 0.1589,
+      "step": 7406
+    },
+    {
+      "epoch": 2.4735348138253466,
+      "grad_norm": 0.5349082547026425,
+      "learning_rate": 9.079276959569899e-07,
+      "loss": 0.1647,
+      "step": 7407
+    },
+    {
+      "epoch": 2.473868759392219,
+      "grad_norm": 0.5270376007769813,
+      "learning_rate": 9.068113105905235e-07,
+      "loss": 0.155,
+      "step": 7408
+    },
+    {
+      "epoch": 2.4742027049590916,
+      "grad_norm": 0.5257667676560155,
+      "learning_rate": 9.056955435484061e-07,
+      "loss": 0.1517,
+      "step": 7409
+    },
+    {
+      "epoch": 2.4745366505259643,
+      "grad_norm": 0.5310131029906042,
+      "learning_rate": 9.045803949991843e-07,
+      "loss": 0.1627,
+      "step": 7410
+    },
+    {
+      "epoch": 2.474870596092837,
+      "grad_norm": 0.5110165371528869,
+      "learning_rate": 9.034658651113154e-07,
+      "loss": 0.1547,
+      "step": 7411
+    },
+    {
+      "epoch": 2.4752045416597097,
+      "grad_norm": 0.5240970004480227,
+      "learning_rate": 9.023519540531633e-07,
+      "loss": 0.1574,
+      "step": 7412
+    },
+    {
+      "epoch": 2.475538487226582,
+      "grad_norm": 0.5146289857699795,
+      "learning_rate": 9.01238661992998e-07,
+      "loss": 0.1519,
+      "step": 7413
+    },
+    {
+      "epoch": 2.4758724327934547,
+      "grad_norm": 0.5125323382835603,
+      "learning_rate": 9.001259890989927e-07,
+      "loss": 0.1452,
+      "step": 7414
+    },
+    {
+      "epoch": 2.4762063783603274,
+      "grad_norm": 0.548528572666765,
+      "learning_rate": 8.990139355392324e-07,
+      "loss": 0.1633,
+      "step": 7415
+    },
+    {
+      "epoch": 2.4765403239271997,
+      "grad_norm": 0.5426017894190176,
+      "learning_rate": 8.979025014817039e-07,
+      "loss": 0.1554,
+      "step": 7416
+    },
+    {
+      "epoch": 2.4768742694940724,
+      "grad_norm": 0.5273871603200672,
+      "learning_rate": 8.967916870943028e-07,
+      "loss": 0.156,
+      "step": 7417
+    },
+    {
+      "epoch": 2.477208215060945,
+      "grad_norm": 0.5015673721907438,
+      "learning_rate": 8.956814925448309e-07,
+      "loss": 0.1551,
+      "step": 7418
+    },
+    {
+      "epoch": 2.477542160627818,
+      "grad_norm": 0.5351497045095686,
+      "learning_rate": 8.945719180009977e-07,
+      "loss": 0.1593,
+      "step": 7419
+    },
+    {
+      "epoch": 2.47787610619469,
+      "grad_norm": 0.5248370885000369,
+      "learning_rate": 8.934629636304149e-07,
+      "loss": 0.1505,
+      "step": 7420
+    },
+    {
+      "epoch": 2.478210051761563,
+      "grad_norm": 0.5879544927092775,
+      "learning_rate": 8.923546296006058e-07,
+      "loss": 0.166,
+      "step": 7421
+    },
+    {
+      "epoch": 2.4785439973284356,
+      "grad_norm": 0.5316868901778447,
+      "learning_rate": 8.912469160789944e-07,
+      "loss": 0.1548,
+      "step": 7422
+    },
+    {
+      "epoch": 2.4788779428953083,
+      "grad_norm": 0.5293767236125144,
+      "learning_rate": 8.901398232329156e-07,
+      "loss": 0.1555,
+      "step": 7423
+    },
+    {
+      "epoch": 2.4792118884621805,
+      "grad_norm": 0.5766018928369575,
+      "learning_rate": 8.890333512296095e-07,
+      "loss": 0.1685,
+      "step": 7424
+    },
+    {
+      "epoch": 2.4795458340290533,
+      "grad_norm": 0.49283680451273953,
+      "learning_rate": 8.879275002362197e-07,
+      "loss": 0.1481,
+      "step": 7425
+    },
+    {
+      "epoch": 2.479879779595926,
+      "grad_norm": 0.7372173936319003,
+      "learning_rate": 8.868222704198004e-07,
+      "loss": 0.1562,
+      "step": 7426
+    },
+    {
+      "epoch": 2.4802137251627983,
+      "grad_norm": 0.5586198661064684,
+      "learning_rate": 8.857176619473068e-07,
+      "loss": 0.1478,
+      "step": 7427
+    },
+    {
+      "epoch": 2.480547670729671,
+      "grad_norm": 0.48785260397998287,
+      "learning_rate": 8.846136749856044e-07,
+      "loss": 0.1443,
+      "step": 7428
+    },
+    {
+      "epoch": 2.4808816162965437,
+      "grad_norm": 0.5223348130739237,
+      "learning_rate": 8.835103097014636e-07,
+      "loss": 0.1563,
+      "step": 7429
+    },
+    {
+      "epoch": 2.4812155618634164,
+      "grad_norm": 0.5289422818248141,
+      "learning_rate": 8.824075662615617e-07,
+      "loss": 0.1539,
+      "step": 7430
+    },
+    {
+      "epoch": 2.4815495074302887,
+      "grad_norm": 0.5366617936145673,
+      "learning_rate": 8.813054448324792e-07,
+      "loss": 0.1564,
+      "step": 7431
+    },
+    {
+      "epoch": 2.4818834529971614,
+      "grad_norm": 0.5091155974176387,
+      "learning_rate": 8.80203945580706e-07,
+      "loss": 0.1421,
+      "step": 7432
+    },
+    {
+      "epoch": 2.482217398564034,
+      "grad_norm": 0.5162217183502404,
+      "learning_rate": 8.791030686726349e-07,
+      "loss": 0.1594,
+      "step": 7433
+    },
+    {
+      "epoch": 2.482551344130907,
+      "grad_norm": 0.53574017718763,
+      "learning_rate": 8.780028142745673e-07,
+      "loss": 0.158,
+      "step": 7434
+    },
+    {
+      "epoch": 2.482885289697779,
+      "grad_norm": 0.5170105505185464,
+      "learning_rate": 8.769031825527097e-07,
+      "loss": 0.1585,
+      "step": 7435
+    },
+    {
+      "epoch": 2.483219235264652,
+      "grad_norm": 0.544046520845691,
+      "learning_rate": 8.758041736731753e-07,
+      "loss": 0.1589,
+      "step": 7436
+    },
+    {
+      "epoch": 2.4835531808315245,
+      "grad_norm": 0.5531553610216963,
+      "learning_rate": 8.747057878019799e-07,
+      "loss": 0.1604,
+      "step": 7437
+    },
+    {
+      "epoch": 2.483887126398397,
+      "grad_norm": 0.4998496915162335,
+      "learning_rate": 8.736080251050505e-07,
+      "loss": 0.1451,
+      "step": 7438
+    },
+    {
+      "epoch": 2.4842210719652695,
+      "grad_norm": 0.5360141755777315,
+      "learning_rate": 8.725108857482145e-07,
+      "loss": 0.1601,
+      "step": 7439
+    },
+    {
+      "epoch": 2.4845550175321423,
+      "grad_norm": 0.5323788988525391,
+      "learning_rate": 8.714143698972083e-07,
+      "loss": 0.155,
+      "step": 7440
+    },
+    {
+      "epoch": 2.484888963099015,
+      "grad_norm": 0.48108964205325405,
+      "learning_rate": 8.703184777176743e-07,
+      "loss": 0.1426,
+      "step": 7441
+    },
+    {
+      "epoch": 2.4852229086658877,
+      "grad_norm": 0.5182627973890735,
+      "learning_rate": 8.692232093751613e-07,
+      "loss": 0.1491,
+      "step": 7442
+    },
+    {
+      "epoch": 2.48555685423276,
+      "grad_norm": 0.48494737578990094,
+      "learning_rate": 8.68128565035119e-07,
+      "loss": 0.1471,
+      "step": 7443
+    },
+    {
+      "epoch": 2.4858907997996327,
+      "grad_norm": 0.49811481612701375,
+      "learning_rate": 8.670345448629097e-07,
+      "loss": 0.1515,
+      "step": 7444
+    },
+    {
+      "epoch": 2.4862247453665054,
+      "grad_norm": 0.5458036809897917,
+      "learning_rate": 8.659411490237951e-07,
+      "loss": 0.1586,
+      "step": 7445
+    },
+    {
+      "epoch": 2.4865586909333777,
+      "grad_norm": 0.5294276700638828,
+      "learning_rate": 8.648483776829469e-07,
+      "loss": 0.1663,
+      "step": 7446
+    },
+    {
+      "epoch": 2.4868926365002504,
+      "grad_norm": 0.491083599496525,
+      "learning_rate": 8.637562310054425e-07,
+      "loss": 0.1483,
+      "step": 7447
+    },
+    {
+      "epoch": 2.487226582067123,
+      "grad_norm": 0.551263065477862,
+      "learning_rate": 8.626647091562612e-07,
+      "loss": 0.1617,
+      "step": 7448
+    },
+    {
+      "epoch": 2.487560527633996,
+      "grad_norm": 0.49702157711243133,
+      "learning_rate": 8.61573812300292e-07,
+      "loss": 0.1504,
+      "step": 7449
+    },
+    {
+      "epoch": 2.487894473200868,
+      "grad_norm": 0.5200659870372368,
+      "learning_rate": 8.604835406023254e-07,
+      "loss": 0.1598,
+      "step": 7450
+    },
+    {
+      "epoch": 2.488228418767741,
+      "grad_norm": 0.49403596255559235,
+      "learning_rate": 8.593938942270613e-07,
+      "loss": 0.1489,
+      "step": 7451
+    },
+    {
+      "epoch": 2.4885623643346135,
+      "grad_norm": 0.5572892670064632,
+      "learning_rate": 8.583048733391036e-07,
+      "loss": 0.1634,
+      "step": 7452
+    },
+    {
+      "epoch": 2.4888963099014862,
+      "grad_norm": 0.524050270984186,
+      "learning_rate": 8.57216478102963e-07,
+      "loss": 0.152,
+      "step": 7453
+    },
+    {
+      "epoch": 2.4892302554683585,
+      "grad_norm": 0.5450655789511571,
+      "learning_rate": 8.561287086830516e-07,
+      "loss": 0.1578,
+      "step": 7454
+    },
+    {
+      "epoch": 2.4895642010352312,
+      "grad_norm": 0.4938587213270162,
+      "learning_rate": 8.550415652436927e-07,
+      "loss": 0.1541,
+      "step": 7455
+    },
+    {
+      "epoch": 2.489898146602104,
+      "grad_norm": 0.5114068685962905,
+      "learning_rate": 8.539550479491093e-07,
+      "loss": 0.151,
+      "step": 7456
+    },
+    {
+      "epoch": 2.4902320921689762,
+      "grad_norm": 0.5449036600871238,
+      "learning_rate": 8.528691569634357e-07,
+      "loss": 0.1687,
+      "step": 7457
+    },
+    {
+      "epoch": 2.490566037735849,
+      "grad_norm": 0.5858119072813122,
+      "learning_rate": 8.517838924507039e-07,
+      "loss": 0.1708,
+      "step": 7458
+    },
+    {
+      "epoch": 2.4908999833027217,
+      "grad_norm": 0.5347934032616631,
+      "learning_rate": 8.50699254574861e-07,
+      "loss": 0.1602,
+      "step": 7459
+    },
+    {
+      "epoch": 2.4912339288695944,
+      "grad_norm": 0.511032385630495,
+      "learning_rate": 8.496152434997518e-07,
+      "loss": 0.1505,
+      "step": 7460
+    },
+    {
+      "epoch": 2.491567874436467,
+      "grad_norm": 0.5590357970061134,
+      "learning_rate": 8.485318593891295e-07,
+      "loss": 0.1422,
+      "step": 7461
+    },
+    {
+      "epoch": 2.4919018200033394,
+      "grad_norm": 0.5920055950996739,
+      "learning_rate": 8.474491024066512e-07,
+      "loss": 0.1667,
+      "step": 7462
+    },
+    {
+      "epoch": 2.492235765570212,
+      "grad_norm": 0.5492390339266932,
+      "learning_rate": 8.463669727158819e-07,
+      "loss": 0.1587,
+      "step": 7463
+    },
+    {
+      "epoch": 2.492569711137085,
+      "grad_norm": 0.526164253031194,
+      "learning_rate": 8.45285470480286e-07,
+      "loss": 0.1534,
+      "step": 7464
+    },
+    {
+      "epoch": 2.492903656703957,
+      "grad_norm": 0.5364159996478518,
+      "learning_rate": 8.442045958632428e-07,
+      "loss": 0.1597,
+      "step": 7465
+    },
+    {
+      "epoch": 2.49323760227083,
+      "grad_norm": 0.5481814028546967,
+      "learning_rate": 8.431243490280267e-07,
+      "loss": 0.1607,
+      "step": 7466
+    },
+    {
+      "epoch": 2.4935715478377025,
+      "grad_norm": 0.5497464097584005,
+      "learning_rate": 8.420447301378249e-07,
+      "loss": 0.1603,
+      "step": 7467
+    },
+    {
+      "epoch": 2.4939054934045752,
+      "grad_norm": 0.49499461566929853,
+      "learning_rate": 8.409657393557236e-07,
+      "loss": 0.1532,
+      "step": 7468
+    },
+    {
+      "epoch": 2.4942394389714475,
+      "grad_norm": 0.5280999383444667,
+      "learning_rate": 8.39887376844718e-07,
+      "loss": 0.1513,
+      "step": 7469
+    },
+    {
+      "epoch": 2.4945733845383202,
+      "grad_norm": 0.541491593990084,
+      "learning_rate": 8.388096427677095e-07,
+      "loss": 0.1575,
+      "step": 7470
+    },
+    {
+      "epoch": 2.494907330105193,
+      "grad_norm": 0.5567512139651408,
+      "learning_rate": 8.377325372874995e-07,
+      "loss": 0.1587,
+      "step": 7471
+    },
+    {
+      "epoch": 2.4952412756720657,
+      "grad_norm": 0.4789075537156718,
+      "learning_rate": 8.366560605668006e-07,
+      "loss": 0.1466,
+      "step": 7472
+    },
+    {
+      "epoch": 2.495575221238938,
+      "grad_norm": 0.5241227372625797,
+      "learning_rate": 8.355802127682238e-07,
+      "loss": 0.1575,
+      "step": 7473
+    },
+    {
+      "epoch": 2.4959091668058107,
+      "grad_norm": 0.5261817341228623,
+      "learning_rate": 8.345049940542904e-07,
+      "loss": 0.1473,
+      "step": 7474
+    },
+    {
+      "epoch": 2.4962431123726834,
+      "grad_norm": 0.5125544170513542,
+      "learning_rate": 8.334304045874248e-07,
+      "loss": 0.1622,
+      "step": 7475
+    },
+    {
+      "epoch": 2.4965770579395556,
+      "grad_norm": 0.4886019125041492,
+      "learning_rate": 8.323564445299575e-07,
+      "loss": 0.1477,
+      "step": 7476
+    },
+    {
+      "epoch": 2.4969110035064284,
+      "grad_norm": 0.46844423797008194,
+      "learning_rate": 8.312831140441207e-07,
+      "loss": 0.1444,
+      "step": 7477
+    },
+    {
+      "epoch": 2.497244949073301,
+      "grad_norm": 0.5183590771956148,
+      "learning_rate": 8.302104132920552e-07,
+      "loss": 0.1565,
+      "step": 7478
+    },
+    {
+      "epoch": 2.497578894640174,
+      "grad_norm": 0.5577300530187361,
+      "learning_rate": 8.291383424358041e-07,
+      "loss": 0.1617,
+      "step": 7479
+    },
+    {
+      "epoch": 2.497912840207046,
+      "grad_norm": 0.5636308968550843,
+      "learning_rate": 8.280669016373172e-07,
+      "loss": 0.1691,
+      "step": 7480
+    },
+    {
+      "epoch": 2.498246785773919,
+      "grad_norm": 0.5084283575774319,
+      "learning_rate": 8.269960910584457e-07,
+      "loss": 0.155,
+      "step": 7481
+    },
+    {
+      "epoch": 2.4985807313407915,
+      "grad_norm": 0.5246190756503197,
+      "learning_rate": 8.259259108609524e-07,
+      "loss": 0.1592,
+      "step": 7482
+    },
+    {
+      "epoch": 2.4989146769076642,
+      "grad_norm": 0.4765757170409379,
+      "learning_rate": 8.248563612064969e-07,
+      "loss": 0.1451,
+      "step": 7483
+    },
+    {
+      "epoch": 2.4992486224745365,
+      "grad_norm": 0.5183633745167002,
+      "learning_rate": 8.237874422566505e-07,
+      "loss": 0.1571,
+      "step": 7484
+    },
+    {
+      "epoch": 2.499582568041409,
+      "grad_norm": 0.49644317659578663,
+      "learning_rate": 8.227191541728829e-07,
+      "loss": 0.1486,
+      "step": 7485
+    },
+    {
+      "epoch": 2.499916513608282,
+      "grad_norm": 0.5302602187154111,
+      "learning_rate": 8.21651497116574e-07,
+      "loss": 0.1544,
+      "step": 7486
+    },
+    {
+      "epoch": 2.500250459175154,
+      "grad_norm": 0.5532769666070599,
+      "learning_rate": 8.205844712490024e-07,
+      "loss": 0.1677,
+      "step": 7487
+    },
+    {
+      "epoch": 2.500584404742027,
+      "grad_norm": 0.555633213185817,
+      "learning_rate": 8.195180767313604e-07,
+      "loss": 0.1645,
+      "step": 7488
+    },
+    {
+      "epoch": 2.5009183503088996,
+      "grad_norm": 0.5607026487442186,
+      "learning_rate": 8.184523137247346e-07,
+      "loss": 0.1576,
+      "step": 7489
+    },
+    {
+      "epoch": 2.5012522958757724,
+      "grad_norm": 0.5288814062908609,
+      "learning_rate": 8.173871823901247e-07,
+      "loss": 0.1486,
+      "step": 7490
+    },
+    {
+      "epoch": 2.501586241442645,
+      "grad_norm": 0.5171926986273441,
+      "learning_rate": 8.16322682888428e-07,
+      "loss": 0.1588,
+      "step": 7491
+    },
+    {
+      "epoch": 2.5019201870095173,
+      "grad_norm": 0.5293127044915098,
+      "learning_rate": 8.15258815380453e-07,
+      "loss": 0.1583,
+      "step": 7492
+    },
+    {
+      "epoch": 2.50225413257639,
+      "grad_norm": 0.5176749228194966,
+      "learning_rate": 8.141955800269058e-07,
+      "loss": 0.1549,
+      "step": 7493
+    },
+    {
+      "epoch": 2.502588078143263,
+      "grad_norm": 0.5159921514636313,
+      "learning_rate": 8.131329769884027e-07,
+      "loss": 0.1512,
+      "step": 7494
+    },
+    {
+      "epoch": 2.502922023710135,
+      "grad_norm": 0.5257063522299663,
+      "learning_rate": 8.120710064254634e-07,
+      "loss": 0.1536,
+      "step": 7495
+    },
+    {
+      "epoch": 2.5032559692770078,
+      "grad_norm": 0.5264810763788741,
+      "learning_rate": 8.110096684985086e-07,
+      "loss": 0.1622,
+      "step": 7496
+    },
+    {
+      "epoch": 2.5035899148438805,
+      "grad_norm": 0.5639565998252986,
+      "learning_rate": 8.099489633678676e-07,
+      "loss": 0.1718,
+      "step": 7497
+    },
+    {
+      "epoch": 2.503923860410753,
+      "grad_norm": 0.5502371897141942,
+      "learning_rate": 8.088888911937726e-07,
+      "loss": 0.1676,
+      "step": 7498
+    },
+    {
+      "epoch": 2.5042578059776255,
+      "grad_norm": 0.6154170180004221,
+      "learning_rate": 8.078294521363584e-07,
+      "loss": 0.1745,
+      "step": 7499
+    },
+    {
+      "epoch": 2.504591751544498,
+      "grad_norm": 0.5371791465358621,
+      "learning_rate": 8.067706463556663e-07,
+      "loss": 0.1613,
+      "step": 7500
+    },
+    {
+      "epoch": 2.504925697111371,
+      "grad_norm": 0.5207840621091431,
+      "learning_rate": 8.057124740116434e-07,
+      "loss": 0.1594,
+      "step": 7501
+    },
+    {
+      "epoch": 2.5052596426782436,
+      "grad_norm": 0.4840124847038789,
+      "learning_rate": 8.046549352641359e-07,
+      "loss": 0.1549,
+      "step": 7502
+    },
+    {
+      "epoch": 2.505593588245116,
+      "grad_norm": 0.4843225011059664,
+      "learning_rate": 8.035980302729008e-07,
+      "loss": 0.1444,
+      "step": 7503
+    },
+    {
+      "epoch": 2.5059275338119886,
+      "grad_norm": 0.5015309666946438,
+      "learning_rate": 8.025417591975926e-07,
+      "loss": 0.1495,
+      "step": 7504
+    },
+    {
+      "epoch": 2.5062614793788613,
+      "grad_norm": 0.5588866802180681,
+      "learning_rate": 8.014861221977749e-07,
+      "loss": 0.1632,
+      "step": 7505
+    },
+    {
+      "epoch": 2.5065954249457336,
+      "grad_norm": 0.5258734952648931,
+      "learning_rate": 8.004311194329145e-07,
+      "loss": 0.1564,
+      "step": 7506
+    },
+    {
+      "epoch": 2.5069293705126063,
+      "grad_norm": 0.5018390452316217,
+      "learning_rate": 7.993767510623834e-07,
+      "loss": 0.1539,
+      "step": 7507
+    },
+    {
+      "epoch": 2.507263316079479,
+      "grad_norm": 0.5238239925463696,
+      "learning_rate": 7.983230172454531e-07,
+      "loss": 0.1519,
+      "step": 7508
+    },
+    {
+      "epoch": 2.5075972616463518,
+      "grad_norm": 0.5573481196677725,
+      "learning_rate": 7.972699181413058e-07,
+      "loss": 0.1591,
+      "step": 7509
+    },
+    {
+      "epoch": 2.5079312072132245,
+      "grad_norm": 0.5166173807653006,
+      "learning_rate": 7.962174539090201e-07,
+      "loss": 0.1494,
+      "step": 7510
+    },
+    {
+      "epoch": 2.5082651527800968,
+      "grad_norm": 0.5055095687381181,
+      "learning_rate": 7.951656247075884e-07,
+      "loss": 0.1519,
+      "step": 7511
+    },
+    {
+      "epoch": 2.5085990983469695,
+      "grad_norm": 0.5560200453792518,
+      "learning_rate": 7.941144306958986e-07,
+      "loss": 0.1676,
+      "step": 7512
+    },
+    {
+      "epoch": 2.508933043913842,
+      "grad_norm": 0.539443573117671,
+      "learning_rate": 7.930638720327477e-07,
+      "loss": 0.1644,
+      "step": 7513
+    },
+    {
+      "epoch": 2.5092669894807145,
+      "grad_norm": 0.5078255433817627,
+      "learning_rate": 7.920139488768325e-07,
+      "loss": 0.1542,
+      "step": 7514
+    },
+    {
+      "epoch": 2.509600935047587,
+      "grad_norm": 0.5588296693905245,
+      "learning_rate": 7.909646613867594e-07,
+      "loss": 0.1689,
+      "step": 7515
+    },
+    {
+      "epoch": 2.50993488061446,
+      "grad_norm": 0.5243310356959708,
+      "learning_rate": 7.899160097210329e-07,
+      "loss": 0.1535,
+      "step": 7516
+    },
+    {
+      "epoch": 2.510268826181332,
+      "grad_norm": 0.4984528436569994,
+      "learning_rate": 7.888679940380644e-07,
+      "loss": 0.1573,
+      "step": 7517
+    },
+    {
+      "epoch": 2.510602771748205,
+      "grad_norm": 0.4913467718422485,
+      "learning_rate": 7.87820614496172e-07,
+      "loss": 0.1558,
+      "step": 7518
+    },
+    {
+      "epoch": 2.5109367173150776,
+      "grad_norm": 0.5518878812557201,
+      "learning_rate": 7.867738712535711e-07,
+      "loss": 0.1622,
+      "step": 7519
+    },
+    {
+      "epoch": 2.5112706628819503,
+      "grad_norm": 0.5117546474410686,
+      "learning_rate": 7.857277644683858e-07,
+      "loss": 0.151,
+      "step": 7520
+    },
+    {
+      "epoch": 2.511604608448823,
+      "grad_norm": 0.5716066786013106,
+      "learning_rate": 7.846822942986449e-07,
+      "loss": 0.1585,
+      "step": 7521
+    },
+    {
+      "epoch": 2.5119385540156953,
+      "grad_norm": 0.526200387575101,
+      "learning_rate": 7.836374609022756e-07,
+      "loss": 0.146,
+      "step": 7522
+    },
+    {
+      "epoch": 2.512272499582568,
+      "grad_norm": 0.5237426581490771,
+      "learning_rate": 7.825932644371137e-07,
+      "loss": 0.1565,
+      "step": 7523
+    },
+    {
+      "epoch": 2.5126064451494408,
+      "grad_norm": 0.5157049326498191,
+      "learning_rate": 7.815497050608989e-07,
+      "loss": 0.1586,
+      "step": 7524
+    },
+    {
+      "epoch": 2.512940390716313,
+      "grad_norm": 0.5075710985330287,
+      "learning_rate": 7.805067829312707e-07,
+      "loss": 0.1526,
+      "step": 7525
+    },
+    {
+      "epoch": 2.5132743362831858,
+      "grad_norm": 0.5214226974333179,
+      "learning_rate": 7.79464498205777e-07,
+      "loss": 0.153,
+      "step": 7526
+    },
+    {
+      "epoch": 2.5136082818500585,
+      "grad_norm": 0.519026372135375,
+      "learning_rate": 7.78422851041865e-07,
+      "loss": 0.1557,
+      "step": 7527
+    },
+    {
+      "epoch": 2.513942227416931,
+      "grad_norm": 0.5682299743488699,
+      "learning_rate": 7.773818415968887e-07,
+      "loss": 0.1679,
+      "step": 7528
+    },
+    {
+      "epoch": 2.514276172983804,
+      "grad_norm": 0.527132465649919,
+      "learning_rate": 7.763414700281053e-07,
+      "loss": 0.1512,
+      "step": 7529
+    },
+    {
+      "epoch": 2.514610118550676,
+      "grad_norm": 0.5662292574706226,
+      "learning_rate": 7.753017364926757e-07,
+      "loss": 0.1618,
+      "step": 7530
+    },
+    {
+      "epoch": 2.514944064117549,
+      "grad_norm": 0.5394910775395201,
+      "learning_rate": 7.742626411476617e-07,
+      "loss": 0.1461,
+      "step": 7531
+    },
+    {
+      "epoch": 2.5152780096844216,
+      "grad_norm": 0.5124017265010042,
+      "learning_rate": 7.732241841500332e-07,
+      "loss": 0.1497,
+      "step": 7532
+    },
+    {
+      "epoch": 2.515611955251294,
+      "grad_norm": 0.5236527981564442,
+      "learning_rate": 7.721863656566597e-07,
+      "loss": 0.1495,
+      "step": 7533
+    },
+    {
+      "epoch": 2.5159459008181666,
+      "grad_norm": 0.45904066998357373,
+      "learning_rate": 7.711491858243164e-07,
+      "loss": 0.1397,
+      "step": 7534
+    },
+    {
+      "epoch": 2.5162798463850393,
+      "grad_norm": 0.5023271030916909,
+      "learning_rate": 7.701126448096813e-07,
+      "loss": 0.1515,
+      "step": 7535
+    },
+    {
+      "epoch": 2.5166137919519116,
+      "grad_norm": 0.5461686273133802,
+      "learning_rate": 7.69076742769338e-07,
+      "loss": 0.1552,
+      "step": 7536
+    },
+    {
+      "epoch": 2.5169477375187843,
+      "grad_norm": 0.4674555571007729,
+      "learning_rate": 7.68041479859769e-07,
+      "loss": 0.1399,
+      "step": 7537
+    },
+    {
+      "epoch": 2.517281683085657,
+      "grad_norm": 0.5515888155967048,
+      "learning_rate": 7.670068562373656e-07,
+      "loss": 0.1637,
+      "step": 7538
+    },
+    {
+      "epoch": 2.5176156286525297,
+      "grad_norm": 0.5957056156113891,
+      "learning_rate": 7.65972872058417e-07,
+      "loss": 0.1634,
+      "step": 7539
+    },
+    {
+      "epoch": 2.5179495742194025,
+      "grad_norm": 0.5036132612130619,
+      "learning_rate": 7.6493952747912e-07,
+      "loss": 0.1535,
+      "step": 7540
+    },
+    {
+      "epoch": 2.5182835197862747,
+      "grad_norm": 0.5190668427338941,
+      "learning_rate": 7.639068226555751e-07,
+      "loss": 0.1519,
+      "step": 7541
+    },
+    {
+      "epoch": 2.5186174653531475,
+      "grad_norm": 0.5037790938432386,
+      "learning_rate": 7.628747577437817e-07,
+      "loss": 0.1524,
+      "step": 7542
+    },
+    {
+      "epoch": 2.51895141092002,
+      "grad_norm": 0.5004546836079737,
+      "learning_rate": 7.618433328996466e-07,
+      "loss": 0.1532,
+      "step": 7543
+    },
+    {
+      "epoch": 2.5192853564868924,
+      "grad_norm": 0.5501512241433032,
+      "learning_rate": 7.608125482789802e-07,
+      "loss": 0.1653,
+      "step": 7544
+    },
+    {
+      "epoch": 2.519619302053765,
+      "grad_norm": 0.6007238389340253,
+      "learning_rate": 7.597824040374918e-07,
+      "loss": 0.1758,
+      "step": 7545
+    },
+    {
+      "epoch": 2.519953247620638,
+      "grad_norm": 0.49556874191589473,
+      "learning_rate": 7.587529003307981e-07,
+      "loss": 0.1545,
+      "step": 7546
+    },
+    {
+      "epoch": 2.5202871931875106,
+      "grad_norm": 0.5251067061399489,
+      "learning_rate": 7.57724037314419e-07,
+      "loss": 0.1582,
+      "step": 7547
+    },
+    {
+      "epoch": 2.520621138754383,
+      "grad_norm": 0.5381560130179013,
+      "learning_rate": 7.566958151437743e-07,
+      "loss": 0.159,
+      "step": 7548
+    },
+    {
+      "epoch": 2.5209550843212556,
+      "grad_norm": 0.4986522998380296,
+      "learning_rate": 7.556682339741911e-07,
+      "loss": 0.1498,
+      "step": 7549
+    },
+    {
+      "epoch": 2.5212890298881283,
+      "grad_norm": 0.4996807633077003,
+      "learning_rate": 7.546412939608955e-07,
+      "loss": 0.1471,
+      "step": 7550
+    },
+    {
+      "epoch": 2.521622975455001,
+      "grad_norm": 0.6054073039657178,
+      "learning_rate": 7.5361499525902e-07,
+      "loss": 0.1613,
+      "step": 7551
+    },
+    {
+      "epoch": 2.5219569210218733,
+      "grad_norm": 0.48971374377963084,
+      "learning_rate": 7.525893380235988e-07,
+      "loss": 0.1441,
+      "step": 7552
+    },
+    {
+      "epoch": 2.522290866588746,
+      "grad_norm": 0.5391665062112588,
+      "learning_rate": 7.515643224095709e-07,
+      "loss": 0.1617,
+      "step": 7553
+    },
+    {
+      "epoch": 2.5226248121556187,
+      "grad_norm": 0.5135011269872816,
+      "learning_rate": 7.505399485717746e-07,
+      "loss": 0.1564,
+      "step": 7554
+    },
+    {
+      "epoch": 2.522958757722491,
+      "grad_norm": 0.5283825677531684,
+      "learning_rate": 7.495162166649561e-07,
+      "loss": 0.1627,
+      "step": 7555
+    },
+    {
+      "epoch": 2.5232927032893637,
+      "grad_norm": 0.5178565756705336,
+      "learning_rate": 7.484931268437595e-07,
+      "loss": 0.1548,
+      "step": 7556
+    },
+    {
+      "epoch": 2.5236266488562364,
+      "grad_norm": 0.5152821898488568,
+      "learning_rate": 7.474706792627362e-07,
+      "loss": 0.1573,
+      "step": 7557
+    },
+    {
+      "epoch": 2.523960594423109,
+      "grad_norm": 0.5314921955835342,
+      "learning_rate": 7.464488740763387e-07,
+      "loss": 0.152,
+      "step": 7558
+    },
+    {
+      "epoch": 2.524294539989982,
+      "grad_norm": 0.4911118017182647,
+      "learning_rate": 7.454277114389241e-07,
+      "loss": 0.1472,
+      "step": 7559
+    },
+    {
+      "epoch": 2.524628485556854,
+      "grad_norm": 0.5249373913173525,
+      "learning_rate": 7.444071915047479e-07,
+      "loss": 0.1576,
+      "step": 7560
+    },
+    {
+      "epoch": 2.524962431123727,
+      "grad_norm": 0.5337994119643109,
+      "learning_rate": 7.433873144279751e-07,
+      "loss": 0.1643,
+      "step": 7561
+    },
+    {
+      "epoch": 2.5252963766905996,
+      "grad_norm": 0.6119106329771593,
+      "learning_rate": 7.42368080362667e-07,
+      "loss": 0.1687,
+      "step": 7562
+    },
+    {
+      "epoch": 2.525630322257472,
+      "grad_norm": 0.5577030632990558,
+      "learning_rate": 7.413494894627926e-07,
+      "loss": 0.1777,
+      "step": 7563
+    },
+    {
+      "epoch": 2.5259642678243446,
+      "grad_norm": 0.5203702420065588,
+      "learning_rate": 7.403315418822215e-07,
+      "loss": 0.1485,
+      "step": 7564
+    },
+    {
+      "epoch": 2.5262982133912173,
+      "grad_norm": 0.5038351465089509,
+      "learning_rate": 7.393142377747287e-07,
+      "loss": 0.1475,
+      "step": 7565
+    },
+    {
+      "epoch": 2.5266321589580896,
+      "grad_norm": 0.48326180757043785,
+      "learning_rate": 7.382975772939866e-07,
+      "loss": 0.1471,
+      "step": 7566
+    },
+    {
+      "epoch": 2.5269661045249623,
+      "grad_norm": 0.5366216732331346,
+      "learning_rate": 7.372815605935763e-07,
+      "loss": 0.1617,
+      "step": 7567
+    },
+    {
+      "epoch": 2.527300050091835,
+      "grad_norm": 0.5061743583902273,
+      "learning_rate": 7.362661878269772e-07,
+      "loss": 0.1571,
+      "step": 7568
+    },
+    {
+      "epoch": 2.5276339956587077,
+      "grad_norm": 0.5307042051406543,
+      "learning_rate": 7.352514591475746e-07,
+      "loss": 0.1586,
+      "step": 7569
+    },
+    {
+      "epoch": 2.5279679412255804,
+      "grad_norm": 0.680745479156831,
+      "learning_rate": 7.342373747086557e-07,
+      "loss": 0.1666,
+      "step": 7570
+    },
+    {
+      "epoch": 2.5283018867924527,
+      "grad_norm": 0.5242025885825541,
+      "learning_rate": 7.332239346634079e-07,
+      "loss": 0.1573,
+      "step": 7571
+    },
+    {
+      "epoch": 2.5286358323593254,
+      "grad_norm": 0.5393298239285648,
+      "learning_rate": 7.322111391649261e-07,
+      "loss": 0.1593,
+      "step": 7572
+    },
+    {
+      "epoch": 2.528969777926198,
+      "grad_norm": 0.5185798461075484,
+      "learning_rate": 7.311989883662018e-07,
+      "loss": 0.1588,
+      "step": 7573
+    },
+    {
+      "epoch": 2.5293037234930704,
+      "grad_norm": 0.592503504928077,
+      "learning_rate": 7.301874824201349e-07,
+      "loss": 0.1737,
+      "step": 7574
+    },
+    {
+      "epoch": 2.529637669059943,
+      "grad_norm": 0.5289070259578227,
+      "learning_rate": 7.29176621479522e-07,
+      "loss": 0.1541,
+      "step": 7575
+    },
+    {
+      "epoch": 2.529971614626816,
+      "grad_norm": 0.5399286735957974,
+      "learning_rate": 7.2816640569707e-07,
+      "loss": 0.154,
+      "step": 7576
+    },
+    {
+      "epoch": 2.5303055601936886,
+      "grad_norm": 0.5254361535936164,
+      "learning_rate": 7.271568352253804e-07,
+      "loss": 0.1564,
+      "step": 7577
+    },
+    {
+      "epoch": 2.5306395057605613,
+      "grad_norm": 0.555206381458067,
+      "learning_rate": 7.261479102169627e-07,
+      "loss": 0.1725,
+      "step": 7578
+    },
+    {
+      "epoch": 2.5309734513274336,
+      "grad_norm": 0.5290307988418901,
+      "learning_rate": 7.251396308242259e-07,
+      "loss": 0.1484,
+      "step": 7579
+    },
+    {
+      "epoch": 2.5313073968943063,
+      "grad_norm": 0.504515982522606,
+      "learning_rate": 7.241319971994831e-07,
+      "loss": 0.1496,
+      "step": 7580
+    },
+    {
+      "epoch": 2.531641342461179,
+      "grad_norm": 0.49559680913923954,
+      "learning_rate": 7.231250094949472e-07,
+      "loss": 0.1461,
+      "step": 7581
+    },
+    {
+      "epoch": 2.5319752880280513,
+      "grad_norm": 0.5978243995818697,
+      "learning_rate": 7.221186678627389e-07,
+      "loss": 0.1729,
+      "step": 7582
+    },
+    {
+      "epoch": 2.532309233594924,
+      "grad_norm": 0.5606973816163814,
+      "learning_rate": 7.211129724548754e-07,
+      "loss": 0.1581,
+      "step": 7583
+    },
+    {
+      "epoch": 2.5326431791617967,
+      "grad_norm": 0.5306020480428629,
+      "learning_rate": 7.201079234232805e-07,
+      "loss": 0.1624,
+      "step": 7584
+    },
+    {
+      "epoch": 2.532977124728669,
+      "grad_norm": 0.5223762718058418,
+      "learning_rate": 7.191035209197772e-07,
+      "loss": 0.1497,
+      "step": 7585
+    },
+    {
+      "epoch": 2.5333110702955417,
+      "grad_norm": 0.540345518499979,
+      "learning_rate": 7.180997650960936e-07,
+      "loss": 0.16,
+      "step": 7586
+    },
+    {
+      "epoch": 2.5336450158624144,
+      "grad_norm": 0.5450407713397211,
+      "learning_rate": 7.170966561038561e-07,
+      "loss": 0.1657,
+      "step": 7587
+    },
+    {
+      "epoch": 2.533978961429287,
+      "grad_norm": 0.5391559776735416,
+      "learning_rate": 7.160941940946009e-07,
+      "loss": 0.1577,
+      "step": 7588
+    },
+    {
+      "epoch": 2.53431290699616,
+      "grad_norm": 0.5230788079516275,
+      "learning_rate": 7.150923792197579e-07,
+      "loss": 0.1556,
+      "step": 7589
+    },
+    {
+      "epoch": 2.534646852563032,
+      "grad_norm": 0.5216974254081449,
+      "learning_rate": 7.140912116306648e-07,
+      "loss": 0.1486,
+      "step": 7590
+    },
+    {
+      "epoch": 2.534980798129905,
+      "grad_norm": 0.6321980599276655,
+      "learning_rate": 7.130906914785585e-07,
+      "loss": 0.1547,
+      "step": 7591
+    },
+    {
+      "epoch": 2.5353147436967776,
+      "grad_norm": 0.506483027802621,
+      "learning_rate": 7.120908189145798e-07,
+      "loss": 0.1447,
+      "step": 7592
+    },
+    {
+      "epoch": 2.53564868926365,
+      "grad_norm": 0.5169580171339891,
+      "learning_rate": 7.110915940897722e-07,
+      "loss": 0.1485,
+      "step": 7593
+    },
+    {
+      "epoch": 2.5359826348305226,
+      "grad_norm": 0.5278085164861606,
+      "learning_rate": 7.100930171550785e-07,
+      "loss": 0.1552,
+      "step": 7594
+    },
+    {
+      "epoch": 2.5363165803973953,
+      "grad_norm": 0.5350544885579419,
+      "learning_rate": 7.090950882613479e-07,
+      "loss": 0.1638,
+      "step": 7595
+    },
+    {
+      "epoch": 2.536650525964268,
+      "grad_norm": 0.5060006641486688,
+      "learning_rate": 7.08097807559327e-07,
+      "loss": 0.148,
+      "step": 7596
+    },
+    {
+      "epoch": 2.5369844715311403,
+      "grad_norm": 0.5396958297533546,
+      "learning_rate": 7.071011751996687e-07,
+      "loss": 0.1581,
+      "step": 7597
+    },
+    {
+      "epoch": 2.537318417098013,
+      "grad_norm": 0.556091790600878,
+      "learning_rate": 7.061051913329231e-07,
+      "loss": 0.1697,
+      "step": 7598
+    },
+    {
+      "epoch": 2.5376523626648857,
+      "grad_norm": 0.5181420886400087,
+      "learning_rate": 7.051098561095493e-07,
+      "loss": 0.1489,
+      "step": 7599
+    },
+    {
+      "epoch": 2.5379863082317584,
+      "grad_norm": 0.4768470665798708,
+      "learning_rate": 7.041151696799014e-07,
+      "loss": 0.1363,
+      "step": 7600
+    },
+    {
+      "epoch": 2.5383202537986307,
+      "grad_norm": 0.5196161421818665,
+      "learning_rate": 7.031211321942405e-07,
+      "loss": 0.1533,
+      "step": 7601
+    },
+    {
+      "epoch": 2.5386541993655034,
+      "grad_norm": 0.5232659742245588,
+      "learning_rate": 7.021277438027258e-07,
+      "loss": 0.1568,
+      "step": 7602
+    },
+    {
+      "epoch": 2.538988144932376,
+      "grad_norm": 0.5176128865084071,
+      "learning_rate": 7.011350046554227e-07,
+      "loss": 0.1542,
+      "step": 7603
+    },
+    {
+      "epoch": 2.5393220904992484,
+      "grad_norm": 0.5651500840876793,
+      "learning_rate": 7.001429149022915e-07,
+      "loss": 0.1647,
+      "step": 7604
+    },
+    {
+      "epoch": 2.539656036066121,
+      "grad_norm": 0.5098263760222316,
+      "learning_rate": 6.991514746932048e-07,
+      "loss": 0.1498,
+      "step": 7605
+    },
+    {
+      "epoch": 2.539989981632994,
+      "grad_norm": 0.536465839742826,
+      "learning_rate": 6.981606841779281e-07,
+      "loss": 0.1533,
+      "step": 7606
+    },
+    {
+      "epoch": 2.5403239271998665,
+      "grad_norm": 0.5226597424328394,
+      "learning_rate": 6.971705435061333e-07,
+      "loss": 0.1516,
+      "step": 7607
+    },
+    {
+      "epoch": 2.5406578727667393,
+      "grad_norm": 0.5804541135430479,
+      "learning_rate": 6.96181052827391e-07,
+      "loss": 0.165,
+      "step": 7608
+    },
+    {
+      "epoch": 2.5409918183336115,
+      "grad_norm": 0.5237353158113098,
+      "learning_rate": 6.951922122911775e-07,
+      "loss": 0.1538,
+      "step": 7609
+    },
+    {
+      "epoch": 2.5413257639004843,
+      "grad_norm": 0.5080706830627284,
+      "learning_rate": 6.942040220468654e-07,
+      "loss": 0.148,
+      "step": 7610
+    },
+    {
+      "epoch": 2.541659709467357,
+      "grad_norm": 0.5140049188783807,
+      "learning_rate": 6.932164822437371e-07,
+      "loss": 0.1557,
+      "step": 7611
+    },
+    {
+      "epoch": 2.5419936550342292,
+      "grad_norm": 0.5930589485845684,
+      "learning_rate": 6.922295930309691e-07,
+      "loss": 0.165,
+      "step": 7612
+    },
+    {
+      "epoch": 2.542327600601102,
+      "grad_norm": 0.549299614539444,
+      "learning_rate": 6.912433545576446e-07,
+      "loss": 0.1579,
+      "step": 7613
+    },
+    {
+      "epoch": 2.5426615461679747,
+      "grad_norm": 0.562070790263285,
+      "learning_rate": 6.90257766972744e-07,
+      "loss": 0.1699,
+      "step": 7614
+    },
+    {
+      "epoch": 2.542995491734847,
+      "grad_norm": 0.5227209625996607,
+      "learning_rate": 6.892728304251544e-07,
+      "loss": 0.1612,
+      "step": 7615
+    },
+    {
+      "epoch": 2.5433294373017197,
+      "grad_norm": 0.5487714357697593,
+      "learning_rate": 6.8828854506366e-07,
+      "loss": 0.16,
+      "step": 7616
+    },
+    {
+      "epoch": 2.5436633828685924,
+      "grad_norm": 0.5139354728097603,
+      "learning_rate": 6.873049110369495e-07,
+      "loss": 0.1473,
+      "step": 7617
+    },
+    {
+      "epoch": 2.543997328435465,
+      "grad_norm": 0.495487042816524,
+      "learning_rate": 6.863219284936135e-07,
+      "loss": 0.143,
+      "step": 7618
+    },
+    {
+      "epoch": 2.544331274002338,
+      "grad_norm": 0.5240554996595517,
+      "learning_rate": 6.853395975821414e-07,
+      "loss": 0.158,
+      "step": 7619
+    },
+    {
+      "epoch": 2.54466521956921,
+      "grad_norm": 0.5291080480147916,
+      "learning_rate": 6.843579184509275e-07,
+      "loss": 0.1543,
+      "step": 7620
+    },
+    {
+      "epoch": 2.544999165136083,
+      "grad_norm": 0.49305356830405517,
+      "learning_rate": 6.833768912482636e-07,
+      "loss": 0.1466,
+      "step": 7621
+    },
+    {
+      "epoch": 2.5453331107029555,
+      "grad_norm": 0.4901391626064939,
+      "learning_rate": 6.823965161223472e-07,
+      "loss": 0.142,
+      "step": 7622
+    },
+    {
+      "epoch": 2.545667056269828,
+      "grad_norm": 0.49462502258324764,
+      "learning_rate": 6.814167932212751e-07,
+      "loss": 0.1467,
+      "step": 7623
+    },
+    {
+      "epoch": 2.5460010018367005,
+      "grad_norm": 0.5141470034471365,
+      "learning_rate": 6.804377226930469e-07,
+      "loss": 0.1573,
+      "step": 7624
+    },
+    {
+      "epoch": 2.5463349474035732,
+      "grad_norm": 0.5089120455293014,
+      "learning_rate": 6.794593046855613e-07,
+      "loss": 0.1534,
+      "step": 7625
+    },
+    {
+      "epoch": 2.546668892970446,
+      "grad_norm": 0.5090069224737189,
+      "learning_rate": 6.784815393466215e-07,
+      "loss": 0.1551,
+      "step": 7626
+    },
+    {
+      "epoch": 2.5470028385373187,
+      "grad_norm": 0.518767697913172,
+      "learning_rate": 6.775044268239278e-07,
+      "loss": 0.1463,
+      "step": 7627
+    },
+    {
+      "epoch": 2.547336784104191,
+      "grad_norm": 0.5376965894711552,
+      "learning_rate": 6.765279672650865e-07,
+      "loss": 0.1537,
+      "step": 7628
+    },
+    {
+      "epoch": 2.5476707296710637,
+      "grad_norm": 0.5257888120052993,
+      "learning_rate": 6.75552160817603e-07,
+      "loss": 0.1505,
+      "step": 7629
+    },
+    {
+      "epoch": 2.5480046752379364,
+      "grad_norm": 0.5316600959696877,
+      "learning_rate": 6.745770076288854e-07,
+      "loss": 0.1562,
+      "step": 7630
+    },
+    {
+      "epoch": 2.5483386208048087,
+      "grad_norm": 0.5346487077712377,
+      "learning_rate": 6.736025078462399e-07,
+      "loss": 0.1618,
+      "step": 7631
+    },
+    {
+      "epoch": 2.5486725663716814,
+      "grad_norm": 0.5060699607051514,
+      "learning_rate": 6.726286616168781e-07,
+      "loss": 0.1487,
+      "step": 7632
+    },
+    {
+      "epoch": 2.549006511938554,
+      "grad_norm": 0.5685768568375997,
+      "learning_rate": 6.716554690879085e-07,
+      "loss": 0.1582,
+      "step": 7633
+    },
+    {
+      "epoch": 2.5493404575054264,
+      "grad_norm": 0.5704678810184574,
+      "learning_rate": 6.706829304063467e-07,
+      "loss": 0.1697,
+      "step": 7634
+    },
+    {
+      "epoch": 2.549674403072299,
+      "grad_norm": 0.5273554677422916,
+      "learning_rate": 6.697110457191031e-07,
+      "loss": 0.1525,
+      "step": 7635
+    },
+    {
+      "epoch": 2.550008348639172,
+      "grad_norm": 0.5291739725258179,
+      "learning_rate": 6.687398151729951e-07,
+      "loss": 0.1556,
+      "step": 7636
+    },
+    {
+      "epoch": 2.5503422942060445,
+      "grad_norm": 0.49722507845623964,
+      "learning_rate": 6.677692389147355e-07,
+      "loss": 0.1444,
+      "step": 7637
+    },
+    {
+      "epoch": 2.5506762397729172,
+      "grad_norm": 0.5218911341868374,
+      "learning_rate": 6.667993170909437e-07,
+      "loss": 0.1534,
+      "step": 7638
+    },
+    {
+      "epoch": 2.5510101853397895,
+      "grad_norm": 0.4754035667843285,
+      "learning_rate": 6.658300498481363e-07,
+      "loss": 0.143,
+      "step": 7639
+    },
+    {
+      "epoch": 2.5513441309066622,
+      "grad_norm": 0.5646948692153895,
+      "learning_rate": 6.648614373327328e-07,
+      "loss": 0.1604,
+      "step": 7640
+    },
+    {
+      "epoch": 2.551678076473535,
+      "grad_norm": 0.5619083057889515,
+      "learning_rate": 6.638934796910545e-07,
+      "loss": 0.1614,
+      "step": 7641
+    },
+    {
+      "epoch": 2.5520120220404072,
+      "grad_norm": 0.5161418977249806,
+      "learning_rate": 6.629261770693213e-07,
+      "loss": 0.1586,
+      "step": 7642
+    },
+    {
+      "epoch": 2.55234596760728,
+      "grad_norm": 0.5378163686695885,
+      "learning_rate": 6.619595296136577e-07,
+      "loss": 0.1624,
+      "step": 7643
+    },
+    {
+      "epoch": 2.5526799131741527,
+      "grad_norm": 0.4863809933072381,
+      "learning_rate": 6.609935374700849e-07,
+      "loss": 0.1483,
+      "step": 7644
+    },
+    {
+      "epoch": 2.5530138587410254,
+      "grad_norm": 0.5311657608751162,
+      "learning_rate": 6.600282007845277e-07,
+      "loss": 0.158,
+      "step": 7645
+    },
+    {
+      "epoch": 2.5533478043078977,
+      "grad_norm": 0.5304450691241134,
+      "learning_rate": 6.590635197028128e-07,
+      "loss": 0.16,
+      "step": 7646
+    },
+    {
+      "epoch": 2.5536817498747704,
+      "grad_norm": 0.5094024551256224,
+      "learning_rate": 6.580994943706675e-07,
+      "loss": 0.161,
+      "step": 7647
+    },
+    {
+      "epoch": 2.554015695441643,
+      "grad_norm": 0.4882401458282333,
+      "learning_rate": 6.571361249337161e-07,
+      "loss": 0.1455,
+      "step": 7648
+    },
+    {
+      "epoch": 2.554349641008516,
+      "grad_norm": 0.5183521869165364,
+      "learning_rate": 6.561734115374901e-07,
+      "loss": 0.1589,
+      "step": 7649
+    },
+    {
+      "epoch": 2.554683586575388,
+      "grad_norm": 0.4919516392113919,
+      "learning_rate": 6.552113543274158e-07,
+      "loss": 0.1466,
+      "step": 7650
+    },
+    {
+      "epoch": 2.555017532142261,
+      "grad_norm": 0.4819998758087338,
+      "learning_rate": 6.54249953448825e-07,
+      "loss": 0.1441,
+      "step": 7651
+    },
+    {
+      "epoch": 2.5553514777091335,
+      "grad_norm": 0.5147802225320831,
+      "learning_rate": 6.532892090469484e-07,
+      "loss": 0.1563,
+      "step": 7652
+    },
+    {
+      "epoch": 2.555685423276006,
+      "grad_norm": 0.5320623190525292,
+      "learning_rate": 6.52329121266918e-07,
+      "loss": 0.1685,
+      "step": 7653
+    },
+    {
+      "epoch": 2.5560193688428785,
+      "grad_norm": 0.5253504225252258,
+      "learning_rate": 6.513696902537653e-07,
+      "loss": 0.166,
+      "step": 7654
+    },
+    {
+      "epoch": 2.556353314409751,
+      "grad_norm": 0.5296212624005909,
+      "learning_rate": 6.504109161524257e-07,
+      "loss": 0.1573,
+      "step": 7655
+    },
+    {
+      "epoch": 2.556687259976624,
+      "grad_norm": 0.5441043227921218,
+      "learning_rate": 6.494527991077304e-07,
+      "loss": 0.162,
+      "step": 7656
+    },
+    {
+      "epoch": 2.5570212055434967,
+      "grad_norm": 0.5589909412419124,
+      "learning_rate": 6.484953392644161e-07,
+      "loss": 0.1505,
+      "step": 7657
+    },
+    {
+      "epoch": 2.557355151110369,
+      "grad_norm": 0.5117170772974657,
+      "learning_rate": 6.475385367671183e-07,
+      "loss": 0.1514,
+      "step": 7658
+    },
+    {
+      "epoch": 2.5576890966772416,
+      "grad_norm": 0.534672576883815,
+      "learning_rate": 6.465823917603742e-07,
+      "loss": 0.1478,
+      "step": 7659
+    },
+    {
+      "epoch": 2.5580230422441144,
+      "grad_norm": 0.5313783504428939,
+      "learning_rate": 6.456269043886182e-07,
+      "loss": 0.1628,
+      "step": 7660
+    },
+    {
+      "epoch": 2.5583569878109866,
+      "grad_norm": 0.5160524410564582,
+      "learning_rate": 6.446720747961904e-07,
+      "loss": 0.1531,
+      "step": 7661
+    },
+    {
+      "epoch": 2.5586909333778594,
+      "grad_norm": 0.47096063309538416,
+      "learning_rate": 6.437179031273272e-07,
+      "loss": 0.1446,
+      "step": 7662
+    },
+    {
+      "epoch": 2.559024878944732,
+      "grad_norm": 0.537685199712028,
+      "learning_rate": 6.427643895261687e-07,
+      "loss": 0.1578,
+      "step": 7663
+    },
+    {
+      "epoch": 2.5593588245116043,
+      "grad_norm": 0.5315214663957065,
+      "learning_rate": 6.418115341367543e-07,
+      "loss": 0.1524,
+      "step": 7664
+    },
+    {
+      "epoch": 2.559692770078477,
+      "grad_norm": 0.5027992814477966,
+      "learning_rate": 6.408593371030231e-07,
+      "loss": 0.1503,
+      "step": 7665
+    },
+    {
+      "epoch": 2.56002671564535,
+      "grad_norm": 0.5622015188810098,
+      "learning_rate": 6.399077985688168e-07,
+      "loss": 0.1647,
+      "step": 7666
+    },
+    {
+      "epoch": 2.5603606612122225,
+      "grad_norm": 0.5444389360370473,
+      "learning_rate": 6.389569186778754e-07,
+      "loss": 0.1541,
+      "step": 7667
+    },
+    {
+      "epoch": 2.560694606779095,
+      "grad_norm": 0.4900284290776257,
+      "learning_rate": 6.38006697573842e-07,
+      "loss": 0.1419,
+      "step": 7668
+    },
+    {
+      "epoch": 2.5610285523459675,
+      "grad_norm": 0.5383726943088658,
+      "learning_rate": 6.370571354002553e-07,
+      "loss": 0.1616,
+      "step": 7669
+    },
+    {
+      "epoch": 2.56136249791284,
+      "grad_norm": 0.5003686273859252,
+      "learning_rate": 6.361082323005624e-07,
+      "loss": 0.1542,
+      "step": 7670
+    },
+    {
+      "epoch": 2.561696443479713,
+      "grad_norm": 0.5788182480274164,
+      "learning_rate": 6.351599884181037e-07,
+      "loss": 0.1705,
+      "step": 7671
+    },
+    {
+      "epoch": 2.562030389046585,
+      "grad_norm": 0.5398324783213839,
+      "learning_rate": 6.342124038961234e-07,
+      "loss": 0.1549,
+      "step": 7672
+    },
+    {
+      "epoch": 2.562364334613458,
+      "grad_norm": 0.5253027875595168,
+      "learning_rate": 6.332654788777642e-07,
+      "loss": 0.1559,
+      "step": 7673
+    },
+    {
+      "epoch": 2.5626982801803306,
+      "grad_norm": 0.5641699534412207,
+      "learning_rate": 6.323192135060713e-07,
+      "loss": 0.1604,
+      "step": 7674
+    },
+    {
+      "epoch": 2.5630322257472034,
+      "grad_norm": 0.5536198139751572,
+      "learning_rate": 6.31373607923989e-07,
+      "loss": 0.1575,
+      "step": 7675
+    },
+    {
+      "epoch": 2.563366171314076,
+      "grad_norm": 0.5608823102074348,
+      "learning_rate": 6.304286622743627e-07,
+      "loss": 0.1548,
+      "step": 7676
+    },
+    {
+      "epoch": 2.5637001168809483,
+      "grad_norm": 0.5147008886535758,
+      "learning_rate": 6.294843766999364e-07,
+      "loss": 0.1644,
+      "step": 7677
+    },
+    {
+      "epoch": 2.564034062447821,
+      "grad_norm": 0.5051122394513354,
+      "learning_rate": 6.285407513433572e-07,
+      "loss": 0.1504,
+      "step": 7678
+    },
+    {
+      "epoch": 2.564368008014694,
+      "grad_norm": 0.5089827237941057,
+      "learning_rate": 6.275977863471683e-07,
+      "loss": 0.1541,
+      "step": 7679
+    },
+    {
+      "epoch": 2.564701953581566,
+      "grad_norm": 0.5833479789208602,
+      "learning_rate": 6.266554818538173e-07,
+      "loss": 0.1733,
+      "step": 7680
+    },
+    {
+      "epoch": 2.5650358991484388,
+      "grad_norm": 0.49383382112180885,
+      "learning_rate": 6.257138380056505e-07,
+      "loss": 0.1437,
+      "step": 7681
+    },
+    {
+      "epoch": 2.5653698447153115,
+      "grad_norm": 0.5396466133422341,
+      "learning_rate": 6.24772854944915e-07,
+      "loss": 0.164,
+      "step": 7682
+    },
+    {
+      "epoch": 2.5657037902821838,
+      "grad_norm": 0.5253514663001958,
+      "learning_rate": 6.238325328137552e-07,
+      "loss": 0.1444,
+      "step": 7683
+    },
+    {
+      "epoch": 2.5660377358490565,
+      "grad_norm": 0.5067795635086644,
+      "learning_rate": 6.228928717542205e-07,
+      "loss": 0.1549,
+      "step": 7684
+    },
+    {
+      "epoch": 2.566371681415929,
+      "grad_norm": 0.5002550750718825,
+      "learning_rate": 6.219538719082546e-07,
+      "loss": 0.1482,
+      "step": 7685
+    },
+    {
+      "epoch": 2.566705626982802,
+      "grad_norm": 0.5179834571022827,
+      "learning_rate": 6.210155334177064e-07,
+      "loss": 0.1456,
+      "step": 7686
+    },
+    {
+      "epoch": 2.5670395725496746,
+      "grad_norm": 0.5463570828984406,
+      "learning_rate": 6.200778564243237e-07,
+      "loss": 0.1557,
+      "step": 7687
+    },
+    {
+      "epoch": 2.567373518116547,
+      "grad_norm": 0.533683576059173,
+      "learning_rate": 6.19140841069752e-07,
+      "loss": 0.1511,
+      "step": 7688
+    },
+    {
+      "epoch": 2.5677074636834196,
+      "grad_norm": 0.5512649791188792,
+      "learning_rate": 6.1820448749554e-07,
+      "loss": 0.1568,
+      "step": 7689
+    },
+    {
+      "epoch": 2.5680414092502923,
+      "grad_norm": 0.5157111237491274,
+      "learning_rate": 6.172687958431328e-07,
+      "loss": 0.1499,
+      "step": 7690
+    },
+    {
+      "epoch": 2.5683753548171646,
+      "grad_norm": 0.5054894343421307,
+      "learning_rate": 6.163337662538793e-07,
+      "loss": 0.1648,
+      "step": 7691
+    },
+    {
+      "epoch": 2.5687093003840373,
+      "grad_norm": 0.5738044330278376,
+      "learning_rate": 6.153993988690266e-07,
+      "loss": 0.1696,
+      "step": 7692
+    },
+    {
+      "epoch": 2.56904324595091,
+      "grad_norm": 0.5139856319427727,
+      "learning_rate": 6.144656938297227e-07,
+      "loss": 0.1616,
+      "step": 7693
+    },
+    {
+      "epoch": 2.5693771915177828,
+      "grad_norm": 0.529624079244088,
+      "learning_rate": 6.135326512770124e-07,
+      "loss": 0.1554,
+      "step": 7694
+    },
+    {
+      "epoch": 2.569711137084655,
+      "grad_norm": 0.532186839181486,
+      "learning_rate": 6.126002713518453e-07,
+      "loss": 0.1634,
+      "step": 7695
+    },
+    {
+      "epoch": 2.5700450826515278,
+      "grad_norm": 0.4761868956330876,
+      "learning_rate": 6.116685541950663e-07,
+      "loss": 0.1485,
+      "step": 7696
+    },
+    {
+      "epoch": 2.5703790282184005,
+      "grad_norm": 0.5616544286220363,
+      "learning_rate": 6.107374999474236e-07,
+      "loss": 0.1708,
+      "step": 7697
+    },
+    {
+      "epoch": 2.570712973785273,
+      "grad_norm": 0.5155203837062691,
+      "learning_rate": 6.098071087495621e-07,
+      "loss": 0.1484,
+      "step": 7698
+    },
+    {
+      "epoch": 2.5710469193521455,
+      "grad_norm": 0.5225906206355545,
+      "learning_rate": 6.088773807420312e-07,
+      "loss": 0.1591,
+      "step": 7699
+    },
+    {
+      "epoch": 2.571380864919018,
+      "grad_norm": 0.5466633655649673,
+      "learning_rate": 6.07948316065275e-07,
+      "loss": 0.1612,
+      "step": 7700
+    },
+    {
+      "epoch": 2.571714810485891,
+      "grad_norm": 0.5276109996287601,
+      "learning_rate": 6.070199148596411e-07,
+      "loss": 0.1543,
+      "step": 7701
+    },
+    {
+      "epoch": 2.572048756052763,
+      "grad_norm": 0.5516155736454269,
+      "learning_rate": 6.060921772653738e-07,
+      "loss": 0.1618,
+      "step": 7702
+    },
+    {
+      "epoch": 2.572382701619636,
+      "grad_norm": 0.4830812549050556,
+      "learning_rate": 6.051651034226208e-07,
+      "loss": 0.1473,
+      "step": 7703
+    },
+    {
+      "epoch": 2.5727166471865086,
+      "grad_norm": 0.5248643245059543,
+      "learning_rate": 6.042386934714245e-07,
+      "loss": 0.156,
+      "step": 7704
+    },
+    {
+      "epoch": 2.5730505927533813,
+      "grad_norm": 0.5189440442245319,
+      "learning_rate": 6.03312947551734e-07,
+      "loss": 0.153,
+      "step": 7705
+    },
+    {
+      "epoch": 2.573384538320254,
+      "grad_norm": 0.5311742748839656,
+      "learning_rate": 6.02387865803391e-07,
+      "loss": 0.1588,
+      "step": 7706
+    },
+    {
+      "epoch": 2.5737184838871263,
+      "grad_norm": 0.5306508792963632,
+      "learning_rate": 6.014634483661419e-07,
+      "loss": 0.1582,
+      "step": 7707
+    },
+    {
+      "epoch": 2.574052429453999,
+      "grad_norm": 0.5187201844876844,
+      "learning_rate": 6.005396953796294e-07,
+      "loss": 0.1506,
+      "step": 7708
+    },
+    {
+      "epoch": 2.5743863750208718,
+      "grad_norm": 0.50422011790657,
+      "learning_rate": 5.996166069833976e-07,
+      "loss": 0.1513,
+      "step": 7709
+    },
+    {
+      "epoch": 2.574720320587744,
+      "grad_norm": 0.5131733480453597,
+      "learning_rate": 5.986941833168913e-07,
+      "loss": 0.1477,
+      "step": 7710
+    },
+    {
+      "epoch": 2.5750542661546167,
+      "grad_norm": 0.5346311131575376,
+      "learning_rate": 5.97772424519451e-07,
+      "loss": 0.1571,
+      "step": 7711
+    },
+    {
+      "epoch": 2.5753882117214895,
+      "grad_norm": 0.5096595724529045,
+      "learning_rate": 5.96851330730322e-07,
+      "loss": 0.1472,
+      "step": 7712
+    },
+    {
+      "epoch": 2.5757221572883617,
+      "grad_norm": 0.4972298467703904,
+      "learning_rate": 5.959309020886433e-07,
+      "loss": 0.1452,
+      "step": 7713
+    },
+    {
+      "epoch": 2.5760561028552345,
+      "grad_norm": 0.5252331142466029,
+      "learning_rate": 5.950111387334584e-07,
+      "loss": 0.1486,
+      "step": 7714
+    },
+    {
+      "epoch": 2.576390048422107,
+      "grad_norm": 0.5088947881046305,
+      "learning_rate": 5.940920408037081e-07,
+      "loss": 0.1479,
+      "step": 7715
+    },
+    {
+      "epoch": 2.57672399398898,
+      "grad_norm": 0.5705717088562303,
+      "learning_rate": 5.93173608438234e-07,
+      "loss": 0.1641,
+      "step": 7716
+    },
+    {
+      "epoch": 2.5770579395558526,
+      "grad_norm": 0.5117709453123926,
+      "learning_rate": 5.92255841775774e-07,
+      "loss": 0.1552,
+      "step": 7717
+    },
+    {
+      "epoch": 2.577391885122725,
+      "grad_norm": 0.5272255582904503,
+      "learning_rate": 5.913387409549693e-07,
+      "loss": 0.1486,
+      "step": 7718
+    },
+    {
+      "epoch": 2.5777258306895976,
+      "grad_norm": 0.5500948968492612,
+      "learning_rate": 5.904223061143577e-07,
+      "loss": 0.1614,
+      "step": 7719
+    },
+    {
+      "epoch": 2.5780597762564703,
+      "grad_norm": 0.5422871702657105,
+      "learning_rate": 5.895065373923781e-07,
+      "loss": 0.1612,
+      "step": 7720
+    },
+    {
+      "epoch": 2.5783937218233426,
+      "grad_norm": 0.5311152125105255,
+      "learning_rate": 5.885914349273664e-07,
+      "loss": 0.1588,
+      "step": 7721
+    },
+    {
+      "epoch": 2.5787276673902153,
+      "grad_norm": 0.5100562305085944,
+      "learning_rate": 5.876769988575631e-07,
+      "loss": 0.1479,
+      "step": 7722
+    },
+    {
+      "epoch": 2.579061612957088,
+      "grad_norm": 0.5539085439821296,
+      "learning_rate": 5.867632293211011e-07,
+      "loss": 0.1611,
+      "step": 7723
+    },
+    {
+      "epoch": 2.5793955585239607,
+      "grad_norm": 0.5398868308395143,
+      "learning_rate": 5.85850126456019e-07,
+      "loss": 0.1503,
+      "step": 7724
+    },
+    {
+      "epoch": 2.5797295040908335,
+      "grad_norm": 0.5099924237947704,
+      "learning_rate": 5.84937690400249e-07,
+      "loss": 0.1465,
+      "step": 7725
+    },
+    {
+      "epoch": 2.5800634496577057,
+      "grad_norm": 0.5346486269242904,
+      "learning_rate": 5.840259212916277e-07,
+      "loss": 0.1515,
+      "step": 7726
+    },
+    {
+      "epoch": 2.5803973952245784,
+      "grad_norm": 0.5473017465597664,
+      "learning_rate": 5.831148192678853e-07,
+      "loss": 0.1598,
+      "step": 7727
+    },
+    {
+      "epoch": 2.580731340791451,
+      "grad_norm": 0.5494409248977583,
+      "learning_rate": 5.822043844666586e-07,
+      "loss": 0.1605,
+      "step": 7728
+    },
+    {
+      "epoch": 2.5810652863583234,
+      "grad_norm": 0.5465152627591858,
+      "learning_rate": 5.812946170254763e-07,
+      "loss": 0.1607,
+      "step": 7729
+    },
+    {
+      "epoch": 2.581399231925196,
+      "grad_norm": 0.547006910687151,
+      "learning_rate": 5.803855170817718e-07,
+      "loss": 0.1538,
+      "step": 7730
+    },
+    {
+      "epoch": 2.581733177492069,
+      "grad_norm": 0.5075310278681255,
+      "learning_rate": 5.794770847728736e-07,
+      "loss": 0.147,
+      "step": 7731
+    },
+    {
+      "epoch": 2.582067123058941,
+      "grad_norm": 0.5263962669768227,
+      "learning_rate": 5.785693202360121e-07,
+      "loss": 0.155,
+      "step": 7732
+    },
+    {
+      "epoch": 2.582401068625814,
+      "grad_norm": 0.5384871321201414,
+      "learning_rate": 5.776622236083146e-07,
+      "loss": 0.1469,
+      "step": 7733
+    },
+    {
+      "epoch": 2.5827350141926866,
+      "grad_norm": 0.5167209940597584,
+      "learning_rate": 5.767557950268099e-07,
+      "loss": 0.1554,
+      "step": 7734
+    },
+    {
+      "epoch": 2.5830689597595593,
+      "grad_norm": 0.5339870635390116,
+      "learning_rate": 5.758500346284252e-07,
+      "loss": 0.1633,
+      "step": 7735
+    },
+    {
+      "epoch": 2.583402905326432,
+      "grad_norm": 0.5639288536519077,
+      "learning_rate": 5.749449425499843e-07,
+      "loss": 0.1646,
+      "step": 7736
+    },
+    {
+      "epoch": 2.5837368508933043,
+      "grad_norm": 0.5703969992909854,
+      "learning_rate": 5.740405189282134e-07,
+      "loss": 0.1692,
+      "step": 7737
+    },
+    {
+      "epoch": 2.584070796460177,
+      "grad_norm": 0.49620284954011423,
+      "learning_rate": 5.73136763899737e-07,
+      "loss": 0.1574,
+      "step": 7738
+    },
+    {
+      "epoch": 2.5844047420270497,
+      "grad_norm": 0.5336886804139199,
+      "learning_rate": 5.722336776010756e-07,
+      "loss": 0.158,
+      "step": 7739
+    },
+    {
+      "epoch": 2.584738687593922,
+      "grad_norm": 0.564930073801695,
+      "learning_rate": 5.713312601686533e-07,
+      "loss": 0.1654,
+      "step": 7740
+    },
+    {
+      "epoch": 2.5850726331607947,
+      "grad_norm": 0.5116232221341999,
+      "learning_rate": 5.704295117387904e-07,
+      "loss": 0.1563,
+      "step": 7741
+    },
+    {
+      "epoch": 2.5854065787276674,
+      "grad_norm": 0.5427584437481574,
+      "learning_rate": 5.695284324477052e-07,
+      "loss": 0.1608,
+      "step": 7742
+    },
+    {
+      "epoch": 2.58574052429454,
+      "grad_norm": 0.49116718894623634,
+      "learning_rate": 5.686280224315189e-07,
+      "loss": 0.1455,
+      "step": 7743
+    },
+    {
+      "epoch": 2.5860744698614124,
+      "grad_norm": 0.4711470768494421,
+      "learning_rate": 5.677282818262464e-07,
+      "loss": 0.1355,
+      "step": 7744
+    },
+    {
+      "epoch": 2.586408415428285,
+      "grad_norm": 0.5313007777846921,
+      "learning_rate": 5.668292107678048e-07,
+      "loss": 0.1594,
+      "step": 7745
+    },
+    {
+      "epoch": 2.586742360995158,
+      "grad_norm": 0.5555662849152072,
+      "learning_rate": 5.659308093920101e-07,
+      "loss": 0.1595,
+      "step": 7746
+    },
+    {
+      "epoch": 2.5870763065620306,
+      "grad_norm": 0.5467008793166378,
+      "learning_rate": 5.650330778345776e-07,
+      "loss": 0.1571,
+      "step": 7747
+    },
+    {
+      "epoch": 2.587410252128903,
+      "grad_norm": 0.5230866932851201,
+      "learning_rate": 5.641360162311171e-07,
+      "loss": 0.1468,
+      "step": 7748
+    },
+    {
+      "epoch": 2.5877441976957756,
+      "grad_norm": 0.49357033180205706,
+      "learning_rate": 5.632396247171429e-07,
+      "loss": 0.1513,
+      "step": 7749
+    },
+    {
+      "epoch": 2.5880781432626483,
+      "grad_norm": 0.5286758416649487,
+      "learning_rate": 5.623439034280625e-07,
+      "loss": 0.1546,
+      "step": 7750
+    },
+    {
+      "epoch": 2.5884120888295206,
+      "grad_norm": 0.5572992513940127,
+      "learning_rate": 5.614488524991896e-07,
+      "loss": 0.1607,
+      "step": 7751
+    },
+    {
+      "epoch": 2.5887460343963933,
+      "grad_norm": 0.5053585297967675,
+      "learning_rate": 5.605544720657286e-07,
+      "loss": 0.1557,
+      "step": 7752
+    },
+    {
+      "epoch": 2.589079979963266,
+      "grad_norm": 0.5642451185748008,
+      "learning_rate": 5.596607622627887e-07,
+      "loss": 0.1607,
+      "step": 7753
+    },
+    {
+      "epoch": 2.5894139255301387,
+      "grad_norm": 0.5401260030717812,
+      "learning_rate": 5.587677232253725e-07,
+      "loss": 0.1525,
+      "step": 7754
+    },
+    {
+      "epoch": 2.5897478710970114,
+      "grad_norm": 0.4954060068895129,
+      "learning_rate": 5.57875355088387e-07,
+      "loss": 0.1487,
+      "step": 7755
+    },
+    {
+      "epoch": 2.5900818166638837,
+      "grad_norm": 0.5493464971995057,
+      "learning_rate": 5.569836579866316e-07,
+      "loss": 0.1593,
+      "step": 7756
+    },
+    {
+      "epoch": 2.5904157622307564,
+      "grad_norm": 0.5442184853873584,
+      "learning_rate": 5.560926320548105e-07,
+      "loss": 0.1657,
+      "step": 7757
+    },
+    {
+      "epoch": 2.590749707797629,
+      "grad_norm": 0.5006345373798716,
+      "learning_rate": 5.552022774275228e-07,
+      "loss": 0.153,
+      "step": 7758
+    },
+    {
+      "epoch": 2.5910836533645014,
+      "grad_norm": 0.5469704888693632,
+      "learning_rate": 5.543125942392664e-07,
+      "loss": 0.1609,
+      "step": 7759
+    },
+    {
+      "epoch": 2.591417598931374,
+      "grad_norm": 0.494422781019865,
+      "learning_rate": 5.534235826244389e-07,
+      "loss": 0.1493,
+      "step": 7760
+    },
+    {
+      "epoch": 2.591751544498247,
+      "grad_norm": 0.5474099861552523,
+      "learning_rate": 5.525352427173369e-07,
+      "loss": 0.1608,
+      "step": 7761
+    },
+    {
+      "epoch": 2.592085490065119,
+      "grad_norm": 0.5147734733031684,
+      "learning_rate": 5.516475746521527e-07,
+      "loss": 0.1536,
+      "step": 7762
+    },
+    {
+      "epoch": 2.592419435631992,
+      "grad_norm": 0.5449219405525347,
+      "learning_rate": 5.507605785629794e-07,
+      "loss": 0.1661,
+      "step": 7763
+    },
+    {
+      "epoch": 2.5927533811988646,
+      "grad_norm": 0.510727938826208,
+      "learning_rate": 5.498742545838104e-07,
+      "loss": 0.1602,
+      "step": 7764
+    },
+    {
+      "epoch": 2.5930873267657373,
+      "grad_norm": 0.49424743101460045,
+      "learning_rate": 5.48988602848533e-07,
+      "loss": 0.1438,
+      "step": 7765
+    },
+    {
+      "epoch": 2.59342127233261,
+      "grad_norm": 0.5161401437755033,
+      "learning_rate": 5.481036234909365e-07,
+      "loss": 0.1544,
+      "step": 7766
+    },
+    {
+      "epoch": 2.5937552178994823,
+      "grad_norm": 0.5239447561809009,
+      "learning_rate": 5.472193166447065e-07,
+      "loss": 0.1617,
+      "step": 7767
+    },
+    {
+      "epoch": 2.594089163466355,
+      "grad_norm": 0.48449270455326193,
+      "learning_rate": 5.463356824434285e-07,
+      "loss": 0.1475,
+      "step": 7768
+    },
+    {
+      "epoch": 2.5944231090332277,
+      "grad_norm": 0.5139041345004897,
+      "learning_rate": 5.454527210205857e-07,
+      "loss": 0.1523,
+      "step": 7769
+    },
+    {
+      "epoch": 2.5947570546001,
+      "grad_norm": 0.5735900578841117,
+      "learning_rate": 5.445704325095613e-07,
+      "loss": 0.1689,
+      "step": 7770
+    },
+    {
+      "epoch": 2.5950910001669727,
+      "grad_norm": 0.5461313985515455,
+      "learning_rate": 5.436888170436327e-07,
+      "loss": 0.1595,
+      "step": 7771
+    },
+    {
+      "epoch": 2.5954249457338454,
+      "grad_norm": 0.5188701878410461,
+      "learning_rate": 5.428078747559806e-07,
+      "loss": 0.1524,
+      "step": 7772
+    },
+    {
+      "epoch": 2.595758891300718,
+      "grad_norm": 0.5380374188365196,
+      "learning_rate": 5.419276057796802e-07,
+      "loss": 0.1522,
+      "step": 7773
+    },
+    {
+      "epoch": 2.596092836867591,
+      "grad_norm": 0.5483299645095472,
+      "learning_rate": 5.410480102477067e-07,
+      "loss": 0.1706,
+      "step": 7774
+    },
+    {
+      "epoch": 2.596426782434463,
+      "grad_norm": 0.5307610952040726,
+      "learning_rate": 5.401690882929333e-07,
+      "loss": 0.1538,
+      "step": 7775
+    },
+    {
+      "epoch": 2.596760728001336,
+      "grad_norm": 0.51063905487827,
+      "learning_rate": 5.392908400481334e-07,
+      "loss": 0.1476,
+      "step": 7776
+    },
+    {
+      "epoch": 2.5970946735682086,
+      "grad_norm": 0.4993184207547867,
+      "learning_rate": 5.384132656459745e-07,
+      "loss": 0.1525,
+      "step": 7777
+    },
+    {
+      "epoch": 2.597428619135081,
+      "grad_norm": 0.5401423593539602,
+      "learning_rate": 5.375363652190257e-07,
+      "loss": 0.1602,
+      "step": 7778
+    },
+    {
+      "epoch": 2.5977625647019535,
+      "grad_norm": 0.5034998004381569,
+      "learning_rate": 5.366601388997522e-07,
+      "loss": 0.1438,
+      "step": 7779
+    },
+    {
+      "epoch": 2.5980965102688263,
+      "grad_norm": 0.48936768353800086,
+      "learning_rate": 5.357845868205191e-07,
+      "loss": 0.1464,
+      "step": 7780
+    },
+    {
+      "epoch": 2.5984304558356985,
+      "grad_norm": 0.5169519189655872,
+      "learning_rate": 5.34909709113589e-07,
+      "loss": 0.1577,
+      "step": 7781
+    },
+    {
+      "epoch": 2.5987644014025713,
+      "grad_norm": 0.5433344403290058,
+      "learning_rate": 5.340355059111213e-07,
+      "loss": 0.1594,
+      "step": 7782
+    },
+    {
+      "epoch": 2.599098346969444,
+      "grad_norm": 0.5137287693888043,
+      "learning_rate": 5.331619773451757e-07,
+      "loss": 0.1555,
+      "step": 7783
+    },
+    {
+      "epoch": 2.5994322925363167,
+      "grad_norm": 0.543531749495905,
+      "learning_rate": 5.32289123547709e-07,
+      "loss": 0.1582,
+      "step": 7784
+    },
+    {
+      "epoch": 2.5997662381031894,
+      "grad_norm": 0.5571790592000054,
+      "learning_rate": 5.314169446505757e-07,
+      "loss": 0.1595,
+      "step": 7785
+    },
+    {
+      "epoch": 2.6001001836700617,
+      "grad_norm": 0.5359827903002512,
+      "learning_rate": 5.305454407855282e-07,
+      "loss": 0.1579,
+      "step": 7786
+    },
+    {
+      "epoch": 2.6004341292369344,
+      "grad_norm": 0.5327339424957118,
+      "learning_rate": 5.296746120842189e-07,
+      "loss": 0.1504,
+      "step": 7787
+    },
+    {
+      "epoch": 2.600768074803807,
+      "grad_norm": 0.532171748237851,
+      "learning_rate": 5.288044586781955e-07,
+      "loss": 0.161,
+      "step": 7788
+    },
+    {
+      "epoch": 2.6011020203706794,
+      "grad_norm": 0.5941419346349204,
+      "learning_rate": 5.279349806989054e-07,
+      "loss": 0.1709,
+      "step": 7789
+    },
+    {
+      "epoch": 2.601435965937552,
+      "grad_norm": 0.5568723344691379,
+      "learning_rate": 5.270661782776931e-07,
+      "loss": 0.1656,
+      "step": 7790
+    },
+    {
+      "epoch": 2.601769911504425,
+      "grad_norm": 0.5382829971872203,
+      "learning_rate": 5.26198051545801e-07,
+      "loss": 0.1664,
+      "step": 7791
+    },
+    {
+      "epoch": 2.6021038570712975,
+      "grad_norm": 0.4687505363311479,
+      "learning_rate": 5.253306006343706e-07,
+      "loss": 0.1445,
+      "step": 7792
+    },
+    {
+      "epoch": 2.60243780263817,
+      "grad_norm": 0.5365972383720856,
+      "learning_rate": 5.244638256744422e-07,
+      "loss": 0.1563,
+      "step": 7793
+    },
+    {
+      "epoch": 2.6027717482050425,
+      "grad_norm": 0.531629913868646,
+      "learning_rate": 5.235977267969489e-07,
+      "loss": 0.1545,
+      "step": 7794
+    },
+    {
+      "epoch": 2.6031056937719153,
+      "grad_norm": 0.5512052735448711,
+      "learning_rate": 5.227323041327281e-07,
+      "loss": 0.1651,
+      "step": 7795
+    },
+    {
+      "epoch": 2.603439639338788,
+      "grad_norm": 0.5192438240687594,
+      "learning_rate": 5.218675578125099e-07,
+      "loss": 0.1593,
+      "step": 7796
+    },
+    {
+      "epoch": 2.6037735849056602,
+      "grad_norm": 0.5341457063593223,
+      "learning_rate": 5.210034879669257e-07,
+      "loss": 0.1606,
+      "step": 7797
+    },
+    {
+      "epoch": 2.604107530472533,
+      "grad_norm": 0.4849767604435831,
+      "learning_rate": 5.201400947265029e-07,
+      "loss": 0.1402,
+      "step": 7798
+    },
+    {
+      "epoch": 2.6044414760394057,
+      "grad_norm": 0.5178804361452914,
+      "learning_rate": 5.192773782216681e-07,
+      "loss": 0.1586,
+      "step": 7799
+    },
+    {
+      "epoch": 2.604775421606278,
+      "grad_norm": 0.5509222801070119,
+      "learning_rate": 5.184153385827434e-07,
+      "loss": 0.1608,
+      "step": 7800
+    },
+    {
+      "epoch": 2.6051093671731507,
+      "grad_norm": 0.5055391926707585,
+      "learning_rate": 5.175539759399518e-07,
+      "loss": 0.149,
+      "step": 7801
+    },
+    {
+      "epoch": 2.6054433127400234,
+      "grad_norm": 0.4990687566409661,
+      "learning_rate": 5.166932904234101e-07,
+      "loss": 0.1407,
+      "step": 7802
+    },
+    {
+      "epoch": 2.605777258306896,
+      "grad_norm": 0.5445711995101518,
+      "learning_rate": 5.158332821631362e-07,
+      "loss": 0.1567,
+      "step": 7803
+    },
+    {
+      "epoch": 2.606111203873769,
+      "grad_norm": 0.5676814520874788,
+      "learning_rate": 5.149739512890445e-07,
+      "loss": 0.1597,
+      "step": 7804
+    },
+    {
+      "epoch": 2.606445149440641,
+      "grad_norm": 0.49071635784943063,
+      "learning_rate": 5.141152979309477e-07,
+      "loss": 0.1433,
+      "step": 7805
+    },
+    {
+      "epoch": 2.606779095007514,
+      "grad_norm": 0.5593982162095931,
+      "learning_rate": 5.132573222185539e-07,
+      "loss": 0.1624,
+      "step": 7806
+    },
+    {
+      "epoch": 2.6071130405743865,
+      "grad_norm": 0.5026472702629329,
+      "learning_rate": 5.124000242814725e-07,
+      "loss": 0.1545,
+      "step": 7807
+    },
+    {
+      "epoch": 2.607446986141259,
+      "grad_norm": 0.5215743662653198,
+      "learning_rate": 5.115434042492057e-07,
+      "loss": 0.1497,
+      "step": 7808
+    },
+    {
+      "epoch": 2.6077809317081315,
+      "grad_norm": 0.5329900153400169,
+      "learning_rate": 5.106874622511576e-07,
+      "loss": 0.1571,
+      "step": 7809
+    },
+    {
+      "epoch": 2.6081148772750042,
+      "grad_norm": 0.5110805817757201,
+      "learning_rate": 5.098321984166293e-07,
+      "loss": 0.1449,
+      "step": 7810
+    },
+    {
+      "epoch": 2.6084488228418765,
+      "grad_norm": 0.5610884253526799,
+      "learning_rate": 5.089776128748169e-07,
+      "loss": 0.1589,
+      "step": 7811
+    },
+    {
+      "epoch": 2.6087827684087492,
+      "grad_norm": 0.5378244387007248,
+      "learning_rate": 5.081237057548166e-07,
+      "loss": 0.157,
+      "step": 7812
+    },
+    {
+      "epoch": 2.609116713975622,
+      "grad_norm": 0.5931449426137503,
+      "learning_rate": 5.072704771856201e-07,
+      "loss": 0.1687,
+      "step": 7813
+    },
+    {
+      "epoch": 2.6094506595424947,
+      "grad_norm": 0.4805946751761879,
+      "learning_rate": 5.06417927296119e-07,
+      "loss": 0.1457,
+      "step": 7814
+    },
+    {
+      "epoch": 2.6097846051093674,
+      "grad_norm": 0.4980998435747012,
+      "learning_rate": 5.055660562150983e-07,
+      "loss": 0.1444,
+      "step": 7815
+    },
+    {
+      "epoch": 2.6101185506762397,
+      "grad_norm": 0.531313199842911,
+      "learning_rate": 5.047148640712468e-07,
+      "loss": 0.1558,
+      "step": 7816
+    },
+    {
+      "epoch": 2.6104524962431124,
+      "grad_norm": 0.5085288422637632,
+      "learning_rate": 5.038643509931446e-07,
+      "loss": 0.156,
+      "step": 7817
+    },
+    {
+      "epoch": 2.610786441809985,
+      "grad_norm": 0.5819241309614197,
+      "learning_rate": 5.030145171092732e-07,
+      "loss": 0.1652,
+      "step": 7818
+    },
+    {
+      "epoch": 2.6111203873768574,
+      "grad_norm": 0.5511100906751708,
+      "learning_rate": 5.021653625480089e-07,
+      "loss": 0.1678,
+      "step": 7819
+    },
+    {
+      "epoch": 2.61145433294373,
+      "grad_norm": 0.5176029698496152,
+      "learning_rate": 5.013168874376273e-07,
+      "loss": 0.1526,
+      "step": 7820
+    },
+    {
+      "epoch": 2.611788278510603,
+      "grad_norm": 0.6173691909315729,
+      "learning_rate": 5.004690919062983e-07,
+      "loss": 0.1704,
+      "step": 7821
+    },
+    {
+      "epoch": 2.6121222240774755,
+      "grad_norm": 0.539288615553018,
+      "learning_rate": 4.996219760820947e-07,
+      "loss": 0.1529,
+      "step": 7822
+    },
+    {
+      "epoch": 2.6124561696443482,
+      "grad_norm": 0.5481896336994918,
+      "learning_rate": 4.987755400929817e-07,
+      "loss": 0.1596,
+      "step": 7823
+    },
+    {
+      "epoch": 2.6127901152112205,
+      "grad_norm": 0.5191460081896047,
+      "learning_rate": 4.97929784066824e-07,
+      "loss": 0.151,
+      "step": 7824
+    },
+    {
+      "epoch": 2.6131240607780932,
+      "grad_norm": 0.5264355185799506,
+      "learning_rate": 4.970847081313818e-07,
+      "loss": 0.1561,
+      "step": 7825
+    },
+    {
+      "epoch": 2.613458006344966,
+      "grad_norm": 0.5277097383950515,
+      "learning_rate": 4.962403124143156e-07,
+      "loss": 0.1513,
+      "step": 7826
+    },
+    {
+      "epoch": 2.613791951911838,
+      "grad_norm": 0.5003241442742133,
+      "learning_rate": 4.953965970431779e-07,
+      "loss": 0.153,
+      "step": 7827
+    },
+    {
+      "epoch": 2.614125897478711,
+      "grad_norm": 0.5230613526678042,
+      "learning_rate": 4.945535621454268e-07,
+      "loss": 0.1479,
+      "step": 7828
+    },
+    {
+      "epoch": 2.6144598430455837,
+      "grad_norm": 0.5437893134594357,
+      "learning_rate": 4.937112078484086e-07,
+      "loss": 0.1573,
+      "step": 7829
+    },
+    {
+      "epoch": 2.614793788612456,
+      "grad_norm": 0.5504021915641389,
+      "learning_rate": 4.928695342793733e-07,
+      "loss": 0.1626,
+      "step": 7830
+    },
+    {
+      "epoch": 2.6151277341793286,
+      "grad_norm": 0.5153465880324457,
+      "learning_rate": 4.92028541565464e-07,
+      "loss": 0.1548,
+      "step": 7831
+    },
+    {
+      "epoch": 2.6154616797462014,
+      "grad_norm": 0.5340166597075059,
+      "learning_rate": 4.911882298337228e-07,
+      "loss": 0.1628,
+      "step": 7832
+    },
+    {
+      "epoch": 2.615795625313074,
+      "grad_norm": 0.5363461554034484,
+      "learning_rate": 4.903485992110901e-07,
+      "loss": 0.1536,
+      "step": 7833
+    },
+    {
+      "epoch": 2.616129570879947,
+      "grad_norm": 0.5255258415056727,
+      "learning_rate": 4.895096498243995e-07,
+      "loss": 0.1517,
+      "step": 7834
+    },
+    {
+      "epoch": 2.616463516446819,
+      "grad_norm": 0.5926188747420524,
+      "learning_rate": 4.886713818003874e-07,
+      "loss": 0.1681,
+      "step": 7835
+    },
+    {
+      "epoch": 2.616797462013692,
+      "grad_norm": 0.5227398551663379,
+      "learning_rate": 4.878337952656809e-07,
+      "loss": 0.1522,
+      "step": 7836
+    },
+    {
+      "epoch": 2.6171314075805645,
+      "grad_norm": 0.5447007227711369,
+      "learning_rate": 4.869968903468092e-07,
+      "loss": 0.1609,
+      "step": 7837
+    },
+    {
+      "epoch": 2.617465353147437,
+      "grad_norm": 0.5237554693670284,
+      "learning_rate": 4.861606671701946e-07,
+      "loss": 0.1559,
+      "step": 7838
+    },
+    {
+      "epoch": 2.6177992987143095,
+      "grad_norm": 0.5797396060428281,
+      "learning_rate": 4.853251258621621e-07,
+      "loss": 0.1723,
+      "step": 7839
+    },
+    {
+      "epoch": 2.618133244281182,
+      "grad_norm": 0.5201124138243138,
+      "learning_rate": 4.844902665489265e-07,
+      "loss": 0.159,
+      "step": 7840
+    },
+    {
+      "epoch": 2.618467189848055,
+      "grad_norm": 0.514676268580547,
+      "learning_rate": 4.836560893566056e-07,
+      "loss": 0.1527,
+      "step": 7841
+    },
+    {
+      "epoch": 2.618801135414927,
+      "grad_norm": 0.5207632098157826,
+      "learning_rate": 4.828225944112097e-07,
+      "loss": 0.1545,
+      "step": 7842
+    },
+    {
+      "epoch": 2.6191350809818,
+      "grad_norm": 0.5262119712435508,
+      "learning_rate": 4.819897818386499e-07,
+      "loss": 0.1511,
+      "step": 7843
+    },
+    {
+      "epoch": 2.6194690265486726,
+      "grad_norm": 0.5143227879462103,
+      "learning_rate": 4.811576517647299e-07,
+      "loss": 0.148,
+      "step": 7844
+    },
+    {
+      "epoch": 2.6198029721155454,
+      "grad_norm": 0.5622213956904916,
+      "learning_rate": 4.803262043151557e-07,
+      "loss": 0.1617,
+      "step": 7845
+    },
+    {
+      "epoch": 2.6201369176824176,
+      "grad_norm": 0.5978754119797444,
+      "learning_rate": 4.794954396155249e-07,
+      "loss": 0.1688,
+      "step": 7846
+    },
+    {
+      "epoch": 2.6204708632492903,
+      "grad_norm": 0.5177420641465259,
+      "learning_rate": 4.786653577913364e-07,
+      "loss": 0.1503,
+      "step": 7847
+    },
+    {
+      "epoch": 2.620804808816163,
+      "grad_norm": 0.5276437691064124,
+      "learning_rate": 4.77835958967981e-07,
+      "loss": 0.1479,
+      "step": 7848
+    },
+    {
+      "epoch": 2.6211387543830353,
+      "grad_norm": 0.5363122643509928,
+      "learning_rate": 4.770072432707523e-07,
+      "loss": 0.1486,
+      "step": 7849
+    },
+    {
+      "epoch": 2.621472699949908,
+      "grad_norm": 0.5652686462454588,
+      "learning_rate": 4.761792108248342e-07,
+      "loss": 0.1667,
+      "step": 7850
+    },
+    {
+      "epoch": 2.6218066455167808,
+      "grad_norm": 0.5305584330166254,
+      "learning_rate": 4.753518617553138e-07,
+      "loss": 0.1591,
+      "step": 7851
+    },
+    {
+      "epoch": 2.6221405910836535,
+      "grad_norm": 0.5253770844239304,
+      "learning_rate": 4.745251961871705e-07,
+      "loss": 0.156,
+      "step": 7852
+    },
+    {
+      "epoch": 2.622474536650526,
+      "grad_norm": 0.5468790587793163,
+      "learning_rate": 4.736992142452823e-07,
+      "loss": 0.1598,
+      "step": 7853
+    },
+    {
+      "epoch": 2.6228084822173985,
+      "grad_norm": 0.5425759314571648,
+      "learning_rate": 4.728739160544227e-07,
+      "loss": 0.153,
+      "step": 7854
+    },
+    {
+      "epoch": 2.623142427784271,
+      "grad_norm": 0.5271469526144371,
+      "learning_rate": 4.720493017392641e-07,
+      "loss": 0.152,
+      "step": 7855
+    },
+    {
+      "epoch": 2.623476373351144,
+      "grad_norm": 0.49647538460886176,
+      "learning_rate": 4.712253714243725e-07,
+      "loss": 0.1552,
+      "step": 7856
+    },
+    {
+      "epoch": 2.623810318918016,
+      "grad_norm": 0.5114529568024085,
+      "learning_rate": 4.7040212523421335e-07,
+      "loss": 0.1582,
+      "step": 7857
+    },
+    {
+      "epoch": 2.624144264484889,
+      "grad_norm": 0.5079040388836952,
+      "learning_rate": 4.695795632931477e-07,
+      "loss": 0.1553,
+      "step": 7858
+    },
+    {
+      "epoch": 2.6244782100517616,
+      "grad_norm": 0.48829885332335254,
+      "learning_rate": 4.687576857254328e-07,
+      "loss": 0.1446,
+      "step": 7859
+    },
+    {
+      "epoch": 2.624812155618634,
+      "grad_norm": 0.4728253401593144,
+      "learning_rate": 4.679364926552238e-07,
+      "loss": 0.1451,
+      "step": 7860
+    },
+    {
+      "epoch": 2.6251461011855066,
+      "grad_norm": 0.49875372399465623,
+      "learning_rate": 4.671159842065698e-07,
+      "loss": 0.1521,
+      "step": 7861
+    },
+    {
+      "epoch": 2.6254800467523793,
+      "grad_norm": 0.5076996510828747,
+      "learning_rate": 4.662961605034194e-07,
+      "loss": 0.1442,
+      "step": 7862
+    },
+    {
+      "epoch": 2.625813992319252,
+      "grad_norm": 0.5122407403110252,
+      "learning_rate": 4.654770216696169e-07,
+      "loss": 0.15,
+      "step": 7863
+    },
+    {
+      "epoch": 2.6261479378861248,
+      "grad_norm": 0.5321463370987862,
+      "learning_rate": 4.646585678289034e-07,
+      "loss": 0.1531,
+      "step": 7864
+    },
+    {
+      "epoch": 2.626481883452997,
+      "grad_norm": 0.5579442129363771,
+      "learning_rate": 4.6384079910491376e-07,
+      "loss": 0.1634,
+      "step": 7865
+    },
+    {
+      "epoch": 2.6268158290198698,
+      "grad_norm": 0.5191857872148999,
+      "learning_rate": 4.630237156211842e-07,
+      "loss": 0.1521,
+      "step": 7866
+    },
+    {
+      "epoch": 2.6271497745867425,
+      "grad_norm": 0.537176066982126,
+      "learning_rate": 4.6220731750114267e-07,
+      "loss": 0.1581,
+      "step": 7867
+    },
+    {
+      "epoch": 2.6274837201536148,
+      "grad_norm": 0.5506531503517323,
+      "learning_rate": 4.6139160486811663e-07,
+      "loss": 0.1521,
+      "step": 7868
+    },
+    {
+      "epoch": 2.6278176657204875,
+      "grad_norm": 0.48213991422421537,
+      "learning_rate": 4.605765778453292e-07,
+      "loss": 0.1415,
+      "step": 7869
+    },
+    {
+      "epoch": 2.62815161128736,
+      "grad_norm": 0.6021241570026848,
+      "learning_rate": 4.597622365559007e-07,
+      "loss": 0.1767,
+      "step": 7870
+    },
+    {
+      "epoch": 2.628485556854233,
+      "grad_norm": 0.5383826080314704,
+      "learning_rate": 4.5894858112284445e-07,
+      "loss": 0.1555,
+      "step": 7871
+    },
+    {
+      "epoch": 2.6288195024211056,
+      "grad_norm": 0.5133156672319394,
+      "learning_rate": 4.581356116690755e-07,
+      "loss": 0.1429,
+      "step": 7872
+    },
+    {
+      "epoch": 2.629153447987978,
+      "grad_norm": 0.45845178214461585,
+      "learning_rate": 4.573233283173989e-07,
+      "loss": 0.1351,
+      "step": 7873
+    },
+    {
+      "epoch": 2.6294873935548506,
+      "grad_norm": 0.5028748544138391,
+      "learning_rate": 4.5651173119052427e-07,
+      "loss": 0.1568,
+      "step": 7874
+    },
+    {
+      "epoch": 2.6298213391217233,
+      "grad_norm": 0.523847639748306,
+      "learning_rate": 4.5570082041104915e-07,
+      "loss": 0.1592,
+      "step": 7875
+    },
+    {
+      "epoch": 2.6301552846885956,
+      "grad_norm": 0.5533083847144648,
+      "learning_rate": 4.5489059610147323e-07,
+      "loss": 0.1642,
+      "step": 7876
+    },
+    {
+      "epoch": 2.6304892302554683,
+      "grad_norm": 0.5741688540753057,
+      "learning_rate": 4.5408105838418924e-07,
+      "loss": 0.1677,
+      "step": 7877
+    },
+    {
+      "epoch": 2.630823175822341,
+      "grad_norm": 0.5274978341726365,
+      "learning_rate": 4.5327220738148823e-07,
+      "loss": 0.1568,
+      "step": 7878
+    },
+    {
+      "epoch": 2.6311571213892133,
+      "grad_norm": 0.49812528970220843,
+      "learning_rate": 4.524640432155558e-07,
+      "loss": 0.1485,
+      "step": 7879
+    },
+    {
+      "epoch": 2.631491066956086,
+      "grad_norm": 0.5285848686406707,
+      "learning_rate": 4.516565660084754e-07,
+      "loss": 0.1573,
+      "step": 7880
+    },
+    {
+      "epoch": 2.6318250125229588,
+      "grad_norm": 0.544151028133756,
+      "learning_rate": 4.5084977588222613e-07,
+      "loss": 0.1612,
+      "step": 7881
+    },
+    {
+      "epoch": 2.6321589580898315,
+      "grad_norm": 0.5246947358977803,
+      "learning_rate": 4.500436729586821e-07,
+      "loss": 0.1584,
+      "step": 7882
+    },
+    {
+      "epoch": 2.632492903656704,
+      "grad_norm": 0.5428300093923333,
+      "learning_rate": 4.4923825735961604e-07,
+      "loss": 0.165,
+      "step": 7883
+    },
+    {
+      "epoch": 2.6328268492235765,
+      "grad_norm": 0.4968719145895019,
+      "learning_rate": 4.484335292066938e-07,
+      "loss": 0.1451,
+      "step": 7884
+    },
+    {
+      "epoch": 2.633160794790449,
+      "grad_norm": 0.5269566759015472,
+      "learning_rate": 4.476294886214799e-07,
+      "loss": 0.1479,
+      "step": 7885
+    },
+    {
+      "epoch": 2.633494740357322,
+      "grad_norm": 0.5511501631755724,
+      "learning_rate": 4.468261357254339e-07,
+      "loss": 0.1517,
+      "step": 7886
+    },
+    {
+      "epoch": 2.633828685924194,
+      "grad_norm": 0.5047376296992027,
+      "learning_rate": 4.46023470639913e-07,
+      "loss": 0.1522,
+      "step": 7887
+    },
+    {
+      "epoch": 2.634162631491067,
+      "grad_norm": 0.5883433464805333,
+      "learning_rate": 4.452214934861676e-07,
+      "loss": 0.165,
+      "step": 7888
+    },
+    {
+      "epoch": 2.6344965770579396,
+      "grad_norm": 0.5735857232549988,
+      "learning_rate": 4.4442020438534737e-07,
+      "loss": 0.1584,
+      "step": 7889
+    },
+    {
+      "epoch": 2.6348305226248123,
+      "grad_norm": 0.5085681511536392,
+      "learning_rate": 4.436196034584944e-07,
+      "loss": 0.1553,
+      "step": 7890
+    },
+    {
+      "epoch": 2.6351644681916846,
+      "grad_norm": 0.5134750451469148,
+      "learning_rate": 4.4281969082654976e-07,
+      "loss": 0.1521,
+      "step": 7891
+    },
+    {
+      "epoch": 2.6354984137585573,
+      "grad_norm": 0.5501752039523184,
+      "learning_rate": 4.4202046661035e-07,
+      "loss": 0.1586,
+      "step": 7892
+    },
+    {
+      "epoch": 2.63583235932543,
+      "grad_norm": 0.4935318484140184,
+      "learning_rate": 4.4122193093062815e-07,
+      "loss": 0.1499,
+      "step": 7893
+    },
+    {
+      "epoch": 2.6361663048923027,
+      "grad_norm": 0.5381149489809288,
+      "learning_rate": 4.4042408390801097e-07,
+      "loss": 0.1524,
+      "step": 7894
+    },
+    {
+      "epoch": 2.636500250459175,
+      "grad_norm": 0.48618271758439613,
+      "learning_rate": 4.3962692566302366e-07,
+      "loss": 0.1459,
+      "step": 7895
+    },
+    {
+      "epoch": 2.6368341960260477,
+      "grad_norm": 0.5542917107201565,
+      "learning_rate": 4.38830456316085e-07,
+      "loss": 0.1621,
+      "step": 7896
+    },
+    {
+      "epoch": 2.6371681415929205,
+      "grad_norm": 0.5934052591504712,
+      "learning_rate": 4.38034675987512e-07,
+      "loss": 0.1693,
+      "step": 7897
+    },
+    {
+      "epoch": 2.6375020871597927,
+      "grad_norm": 0.6050578302721633,
+      "learning_rate": 4.372395847975164e-07,
+      "loss": 0.1661,
+      "step": 7898
+    },
+    {
+      "epoch": 2.6378360327266654,
+      "grad_norm": 0.5416419907657133,
+      "learning_rate": 4.364451828662075e-07,
+      "loss": 0.1546,
+      "step": 7899
+    },
+    {
+      "epoch": 2.638169978293538,
+      "grad_norm": 0.5494613936282062,
+      "learning_rate": 4.356514703135867e-07,
+      "loss": 0.1629,
+      "step": 7900
+    },
+    {
+      "epoch": 2.638503923860411,
+      "grad_norm": 0.5203439914354697,
+      "learning_rate": 4.348584472595557e-07,
+      "loss": 0.1557,
+      "step": 7901
+    },
+    {
+      "epoch": 2.6388378694272836,
+      "grad_norm": 0.5468240192643921,
+      "learning_rate": 4.3406611382390826e-07,
+      "loss": 0.1554,
+      "step": 7902
+    },
+    {
+      "epoch": 2.639171814994156,
+      "grad_norm": 0.5255586767892553,
+      "learning_rate": 4.3327447012633695e-07,
+      "loss": 0.1536,
+      "step": 7903
+    },
+    {
+      "epoch": 2.6395057605610286,
+      "grad_norm": 0.5141342995815591,
+      "learning_rate": 4.324835162864283e-07,
+      "loss": 0.1467,
+      "step": 7904
+    },
+    {
+      "epoch": 2.6398397061279013,
+      "grad_norm": 0.5304691027531064,
+      "learning_rate": 4.31693252423665e-07,
+      "loss": 0.1543,
+      "step": 7905
+    },
+    {
+      "epoch": 2.6401736516947736,
+      "grad_norm": 0.5023980897506727,
+      "learning_rate": 4.3090367865742666e-07,
+      "loss": 0.1553,
+      "step": 7906
+    },
+    {
+      "epoch": 2.6405075972616463,
+      "grad_norm": 0.5085208473006091,
+      "learning_rate": 4.3011479510698615e-07,
+      "loss": 0.1499,
+      "step": 7907
+    },
+    {
+      "epoch": 2.640841542828519,
+      "grad_norm": 0.5327667092975095,
+      "learning_rate": 4.293266018915149e-07,
+      "loss": 0.1558,
+      "step": 7908
+    },
+    {
+      "epoch": 2.6411754883953913,
+      "grad_norm": 0.5607038079211306,
+      "learning_rate": 4.2853909913007807e-07,
+      "loss": 0.1694,
+      "step": 7909
+    },
+    {
+      "epoch": 2.641509433962264,
+      "grad_norm": 0.5499665080788384,
+      "learning_rate": 4.277522869416384e-07,
+      "loss": 0.1645,
+      "step": 7910
+    },
+    {
+      "epoch": 2.6418433795291367,
+      "grad_norm": 0.5290539398893629,
+      "learning_rate": 4.269661654450513e-07,
+      "loss": 0.1529,
+      "step": 7911
+    },
+    {
+      "epoch": 2.6421773250960094,
+      "grad_norm": 0.49108126097882077,
+      "learning_rate": 4.261807347590713e-07,
+      "loss": 0.1364,
+      "step": 7912
+    },
+    {
+      "epoch": 2.642511270662882,
+      "grad_norm": 0.5284309578302578,
+      "learning_rate": 4.253959950023456e-07,
+      "loss": 0.1559,
+      "step": 7913
+    },
+    {
+      "epoch": 2.6428452162297544,
+      "grad_norm": 0.5074179173504525,
+      "learning_rate": 4.246119462934195e-07,
+      "loss": 0.1559,
+      "step": 7914
+    },
+    {
+      "epoch": 2.643179161796627,
+      "grad_norm": 0.5065896426675138,
+      "learning_rate": 4.238285887507315e-07,
+      "loss": 0.155,
+      "step": 7915
+    },
+    {
+      "epoch": 2.6435131073635,
+      "grad_norm": 0.5039285388619513,
+      "learning_rate": 4.230459224926198e-07,
+      "loss": 0.1531,
+      "step": 7916
+    },
+    {
+      "epoch": 2.643847052930372,
+      "grad_norm": 0.5283799716103906,
+      "learning_rate": 4.222639476373119e-07,
+      "loss": 0.1448,
+      "step": 7917
+    },
+    {
+      "epoch": 2.644180998497245,
+      "grad_norm": 0.5626566515484439,
+      "learning_rate": 4.2148266430293627e-07,
+      "loss": 0.154,
+      "step": 7918
+    },
+    {
+      "epoch": 2.6445149440641176,
+      "grad_norm": 0.5156261937650805,
+      "learning_rate": 4.207020726075145e-07,
+      "loss": 0.1585,
+      "step": 7919
+    },
+    {
+      "epoch": 2.6448488896309903,
+      "grad_norm": 0.5440110225549866,
+      "learning_rate": 4.199221726689634e-07,
+      "loss": 0.1593,
+      "step": 7920
+    },
+    {
+      "epoch": 2.645182835197863,
+      "grad_norm": 0.570266964313994,
+      "learning_rate": 4.191429646050971e-07,
+      "loss": 0.1671,
+      "step": 7921
+    },
+    {
+      "epoch": 2.6455167807647353,
+      "grad_norm": 0.5221227396321205,
+      "learning_rate": 4.1836444853362465e-07,
+      "loss": 0.1506,
+      "step": 7922
+    },
+    {
+      "epoch": 2.645850726331608,
+      "grad_norm": 0.497219251017224,
+      "learning_rate": 4.1758662457214884e-07,
+      "loss": 0.1373,
+      "step": 7923
+    },
+    {
+      "epoch": 2.6461846718984807,
+      "grad_norm": 0.541261652338555,
+      "learning_rate": 4.1680949283816996e-07,
+      "loss": 0.1576,
+      "step": 7924
+    },
+    {
+      "epoch": 2.646518617465353,
+      "grad_norm": 0.6069496137295698,
+      "learning_rate": 4.160330534490814e-07,
+      "loss": 0.158,
+      "step": 7925
+    },
+    {
+      "epoch": 2.6468525630322257,
+      "grad_norm": 0.5806445366255634,
+      "learning_rate": 4.152573065221749e-07,
+      "loss": 0.1691,
+      "step": 7926
+    },
+    {
+      "epoch": 2.6471865085990984,
+      "grad_norm": 0.5463692447613053,
+      "learning_rate": 4.1448225217463724e-07,
+      "loss": 0.1595,
+      "step": 7927
+    },
+    {
+      "epoch": 2.6475204541659707,
+      "grad_norm": 0.493295458582906,
+      "learning_rate": 4.1370789052354644e-07,
+      "loss": 0.1481,
+      "step": 7928
+    },
+    {
+      "epoch": 2.6478543997328434,
+      "grad_norm": 0.5120650298647899,
+      "learning_rate": 4.129342216858817e-07,
+      "loss": 0.1541,
+      "step": 7929
+    },
+    {
+      "epoch": 2.648188345299716,
+      "grad_norm": 0.5501003904091634,
+      "learning_rate": 4.1216124577851293e-07,
+      "loss": 0.1646,
+      "step": 7930
+    },
+    {
+      "epoch": 2.648522290866589,
+      "grad_norm": 0.5994137158090266,
+      "learning_rate": 4.113889629182083e-07,
+      "loss": 0.1743,
+      "step": 7931
+    },
+    {
+      "epoch": 2.6488562364334616,
+      "grad_norm": 0.5321524927305665,
+      "learning_rate": 4.106173732216295e-07,
+      "loss": 0.1548,
+      "step": 7932
+    },
+    {
+      "epoch": 2.649190182000334,
+      "grad_norm": 0.5392827869317124,
+      "learning_rate": 4.0984647680533564e-07,
+      "loss": 0.1551,
+      "step": 7933
+    },
+    {
+      "epoch": 2.6495241275672066,
+      "grad_norm": 0.5489927227974719,
+      "learning_rate": 4.090762737857784e-07,
+      "loss": 0.1408,
+      "step": 7934
+    },
+    {
+      "epoch": 2.6498580731340793,
+      "grad_norm": 0.5787246971378018,
+      "learning_rate": 4.0830676427930646e-07,
+      "loss": 0.1632,
+      "step": 7935
+    },
+    {
+      "epoch": 2.6501920187009516,
+      "grad_norm": 0.5134979626899223,
+      "learning_rate": 4.0753794840216296e-07,
+      "loss": 0.1568,
+      "step": 7936
+    },
+    {
+      "epoch": 2.6505259642678243,
+      "grad_norm": 0.5604056768059363,
+      "learning_rate": 4.067698262704878e-07,
+      "loss": 0.1671,
+      "step": 7937
+    },
+    {
+      "epoch": 2.650859909834697,
+      "grad_norm": 0.5144901739013071,
+      "learning_rate": 4.0600239800031136e-07,
+      "loss": 0.1545,
+      "step": 7938
+    },
+    {
+      "epoch": 2.6511938554015697,
+      "grad_norm": 0.5203460567839596,
+      "learning_rate": 4.0523566370756774e-07,
+      "loss": 0.1491,
+      "step": 7939
+    },
+    {
+      "epoch": 2.651527800968442,
+      "grad_norm": 0.5067411197472471,
+      "learning_rate": 4.044696235080775e-07,
+      "loss": 0.1451,
+      "step": 7940
+    },
+    {
+      "epoch": 2.6518617465353147,
+      "grad_norm": 0.5252416651013114,
+      "learning_rate": 4.037042775175626e-07,
+      "loss": 0.1578,
+      "step": 7941
+    },
+    {
+      "epoch": 2.6521956921021874,
+      "grad_norm": 0.6042433021264649,
+      "learning_rate": 4.0293962585163493e-07,
+      "loss": 0.1583,
+      "step": 7942
+    },
+    {
+      "epoch": 2.65252963766906,
+      "grad_norm": 0.49441861117276115,
+      "learning_rate": 4.02175668625806e-07,
+      "loss": 0.1454,
+      "step": 7943
+    },
+    {
+      "epoch": 2.6528635832359324,
+      "grad_norm": 0.5240774832604242,
+      "learning_rate": 4.014124059554786e-07,
+      "loss": 0.1564,
+      "step": 7944
+    },
+    {
+      "epoch": 2.653197528802805,
+      "grad_norm": 0.5320277327365568,
+      "learning_rate": 4.006498379559559e-07,
+      "loss": 0.1569,
+      "step": 7945
+    },
+    {
+      "epoch": 2.653531474369678,
+      "grad_norm": 0.530128976732692,
+      "learning_rate": 3.9988796474242977e-07,
+      "loss": 0.149,
+      "step": 7946
+    },
+    {
+      "epoch": 2.65386541993655,
+      "grad_norm": 0.5389535705555505,
+      "learning_rate": 3.9912678642999134e-07,
+      "loss": 0.1621,
+      "step": 7947
+    },
+    {
+      "epoch": 2.654199365503423,
+      "grad_norm": 0.511151976632125,
+      "learning_rate": 3.983663031336249e-07,
+      "loss": 0.1574,
+      "step": 7948
+    },
+    {
+      "epoch": 2.6545333110702956,
+      "grad_norm": 0.5122387147490539,
+      "learning_rate": 3.976065149682112e-07,
+      "loss": 0.1439,
+      "step": 7949
+    },
+    {
+      "epoch": 2.6548672566371683,
+      "grad_norm": 0.5142207485921048,
+      "learning_rate": 3.968474220485252e-07,
+      "loss": 0.1555,
+      "step": 7950
+    },
+    {
+      "epoch": 2.655201202204041,
+      "grad_norm": 0.6583635200461101,
+      "learning_rate": 3.960890244892362e-07,
+      "loss": 0.1672,
+      "step": 7951
+    },
+    {
+      "epoch": 2.6555351477709133,
+      "grad_norm": 0.4891978157653677,
+      "learning_rate": 3.953313224049099e-07,
+      "loss": 0.151,
+      "step": 7952
+    },
+    {
+      "epoch": 2.655869093337786,
+      "grad_norm": 0.535158030555081,
+      "learning_rate": 3.945743159100046e-07,
+      "loss": 0.1524,
+      "step": 7953
+    },
+    {
+      "epoch": 2.6562030389046587,
+      "grad_norm": 0.5370561320841315,
+      "learning_rate": 3.938180051188756e-07,
+      "loss": 0.1574,
+      "step": 7954
+    },
+    {
+      "epoch": 2.656536984471531,
+      "grad_norm": 0.5358457597336919,
+      "learning_rate": 3.930623901457736e-07,
+      "loss": 0.1575,
+      "step": 7955
+    },
+    {
+      "epoch": 2.6568709300384037,
+      "grad_norm": 0.55410806790667,
+      "learning_rate": 3.92307471104843e-07,
+      "loss": 0.1572,
+      "step": 7956
+    },
+    {
+      "epoch": 2.6572048756052764,
+      "grad_norm": 0.5379106010763365,
+      "learning_rate": 3.915532481101225e-07,
+      "loss": 0.1611,
+      "step": 7957
+    },
+    {
+      "epoch": 2.6575388211721487,
+      "grad_norm": 0.5380414994383523,
+      "learning_rate": 3.9079972127554657e-07,
+      "loss": 0.1574,
+      "step": 7958
+    },
+    {
+      "epoch": 2.6578727667390214,
+      "grad_norm": 0.5306946489832371,
+      "learning_rate": 3.9004689071494406e-07,
+      "loss": 0.1552,
+      "step": 7959
+    },
+    {
+      "epoch": 2.658206712305894,
+      "grad_norm": 0.5083058729379549,
+      "learning_rate": 3.8929475654203963e-07,
+      "loss": 0.1509,
+      "step": 7960
+    },
+    {
+      "epoch": 2.658540657872767,
+      "grad_norm": 0.5221341981384049,
+      "learning_rate": 3.8854331887045016e-07,
+      "loss": 0.1501,
+      "step": 7961
+    },
+    {
+      "epoch": 2.6588746034396396,
+      "grad_norm": 0.5406700255249568,
+      "learning_rate": 3.877925778136921e-07,
+      "loss": 0.1588,
+      "step": 7962
+    },
+    {
+      "epoch": 2.659208549006512,
+      "grad_norm": 0.5407476407616856,
+      "learning_rate": 3.870425334851713e-07,
+      "loss": 0.158,
+      "step": 7963
+    },
+    {
+      "epoch": 2.6595424945733845,
+      "grad_norm": 0.5643950493663836,
+      "learning_rate": 3.8629318599819224e-07,
+      "loss": 0.1672,
+      "step": 7964
+    },
+    {
+      "epoch": 2.6598764401402573,
+      "grad_norm": 0.4808623015760249,
+      "learning_rate": 3.855445354659515e-07,
+      "loss": 0.1474,
+      "step": 7965
+    },
+    {
+      "epoch": 2.6602103857071295,
+      "grad_norm": 0.554313310672822,
+      "learning_rate": 3.847965820015426e-07,
+      "loss": 0.1649,
+      "step": 7966
+    },
+    {
+      "epoch": 2.6605443312740022,
+      "grad_norm": 0.49048685231778866,
+      "learning_rate": 3.8404932571795115e-07,
+      "loss": 0.1489,
+      "step": 7967
+    },
+    {
+      "epoch": 2.660878276840875,
+      "grad_norm": 0.5770239422265581,
+      "learning_rate": 3.833027667280614e-07,
+      "loss": 0.1627,
+      "step": 7968
+    },
+    {
+      "epoch": 2.6612122224077477,
+      "grad_norm": 0.5259826441398753,
+      "learning_rate": 3.825569051446476e-07,
+      "loss": 0.1535,
+      "step": 7969
+    },
+    {
+      "epoch": 2.6615461679746204,
+      "grad_norm": 0.5236874169599129,
+      "learning_rate": 3.8181174108038286e-07,
+      "loss": 0.164,
+      "step": 7970
+    },
+    {
+      "epoch": 2.6618801135414927,
+      "grad_norm": 0.5653557919006781,
+      "learning_rate": 3.810672746478317e-07,
+      "loss": 0.1617,
+      "step": 7971
+    },
+    {
+      "epoch": 2.6622140591083654,
+      "grad_norm": 0.5668011201139045,
+      "learning_rate": 3.803235059594551e-07,
+      "loss": 0.1694,
+      "step": 7972
+    },
+    {
+      "epoch": 2.662548004675238,
+      "grad_norm": 0.5805953775265659,
+      "learning_rate": 3.795804351276072e-07,
+      "loss": 0.1654,
+      "step": 7973
+    },
+    {
+      "epoch": 2.6628819502421104,
+      "grad_norm": 0.5111928242314996,
+      "learning_rate": 3.788380622645382e-07,
+      "loss": 0.1611,
+      "step": 7974
+    },
+    {
+      "epoch": 2.663215895808983,
+      "grad_norm": 0.5324550619095361,
+      "learning_rate": 3.780963874823934e-07,
+      "loss": 0.152,
+      "step": 7975
+    },
+    {
+      "epoch": 2.663549841375856,
+      "grad_norm": 0.533264878533902,
+      "learning_rate": 3.773554108932093e-07,
+      "loss": 0.1625,
+      "step": 7976
+    },
+    {
+      "epoch": 2.663883786942728,
+      "grad_norm": 0.5447286073196415,
+      "learning_rate": 3.7661513260892067e-07,
+      "loss": 0.1549,
+      "step": 7977
+    },
+    {
+      "epoch": 2.664217732509601,
+      "grad_norm": 0.5342876864651773,
+      "learning_rate": 3.7587555274135544e-07,
+      "loss": 0.1578,
+      "step": 7978
+    },
+    {
+      "epoch": 2.6645516780764735,
+      "grad_norm": 0.5094628488311667,
+      "learning_rate": 3.751366714022342e-07,
+      "loss": 0.1491,
+      "step": 7979
+    },
+    {
+      "epoch": 2.6648856236433462,
+      "grad_norm": 0.5346577689522607,
+      "learning_rate": 3.7439848870317487e-07,
+      "loss": 0.1602,
+      "step": 7980
+    },
+    {
+      "epoch": 2.665219569210219,
+      "grad_norm": 0.5777180909450799,
+      "learning_rate": 3.7366100475568935e-07,
+      "loss": 0.1643,
+      "step": 7981
+    },
+    {
+      "epoch": 2.6655535147770912,
+      "grad_norm": 0.552682877062851,
+      "learning_rate": 3.7292421967118185e-07,
+      "loss": 0.1615,
+      "step": 7982
+    },
+    {
+      "epoch": 2.665887460343964,
+      "grad_norm": 0.5213827805993274,
+      "learning_rate": 3.72188133560954e-07,
+      "loss": 0.154,
+      "step": 7983
+    },
+    {
+      "epoch": 2.6662214059108367,
+      "grad_norm": 0.5552413572866788,
+      "learning_rate": 3.7145274653619776e-07,
+      "loss": 0.158,
+      "step": 7984
+    },
+    {
+      "epoch": 2.666555351477709,
+      "grad_norm": 0.4755073461749714,
+      "learning_rate": 3.7071805870800395e-07,
+      "loss": 0.145,
+      "step": 7985
+    },
+    {
+      "epoch": 2.6668892970445817,
+      "grad_norm": 0.5444612344044898,
+      "learning_rate": 3.6998407018735525e-07,
+      "loss": 0.1573,
+      "step": 7986
+    },
+    {
+      "epoch": 2.6672232426114544,
+      "grad_norm": 0.5403035052984416,
+      "learning_rate": 3.6925078108513033e-07,
+      "loss": 0.1538,
+      "step": 7987
+    },
+    {
+      "epoch": 2.667557188178327,
+      "grad_norm": 0.4817666520309885,
+      "learning_rate": 3.6851819151209947e-07,
+      "loss": 0.1466,
+      "step": 7988
+    },
+    {
+      "epoch": 2.6678911337451994,
+      "grad_norm": 0.5338765353416609,
+      "learning_rate": 3.677863015789307e-07,
+      "loss": 0.1503,
+      "step": 7989
+    },
+    {
+      "epoch": 2.668225079312072,
+      "grad_norm": 0.4705920965256273,
+      "learning_rate": 3.6705511139618177e-07,
+      "loss": 0.1401,
+      "step": 7990
+    },
+    {
+      "epoch": 2.668559024878945,
+      "grad_norm": 0.5343266008666374,
+      "learning_rate": 3.66324621074311e-07,
+      "loss": 0.1522,
+      "step": 7991
+    },
+    {
+      "epoch": 2.6688929704458175,
+      "grad_norm": 0.5223355051647398,
+      "learning_rate": 3.6559483072366506e-07,
+      "loss": 0.1549,
+      "step": 7992
+    },
+    {
+      "epoch": 2.66922691601269,
+      "grad_norm": 0.5378562856207899,
+      "learning_rate": 3.6486574045448973e-07,
+      "loss": 0.1561,
+      "step": 7993
+    },
+    {
+      "epoch": 2.6695608615795625,
+      "grad_norm": 0.5167077818281667,
+      "learning_rate": 3.6413735037691966e-07,
+      "loss": 0.1433,
+      "step": 7994
+    },
+    {
+      "epoch": 2.6698948071464352,
+      "grad_norm": 0.5341312147095918,
+      "learning_rate": 3.634096606009896e-07,
+      "loss": 0.1602,
+      "step": 7995
+    },
+    {
+      "epoch": 2.6702287527133075,
+      "grad_norm": 0.5225832292402557,
+      "learning_rate": 3.626826712366233e-07,
+      "loss": 0.1424,
+      "step": 7996
+    },
+    {
+      "epoch": 2.6705626982801802,
+      "grad_norm": 0.5189043060999742,
+      "learning_rate": 3.6195638239364225e-07,
+      "loss": 0.1518,
+      "step": 7997
+    },
+    {
+      "epoch": 2.670896643847053,
+      "grad_norm": 0.5300387936707925,
+      "learning_rate": 3.612307941817622e-07,
+      "loss": 0.1572,
+      "step": 7998
+    },
+    {
+      "epoch": 2.6712305894139257,
+      "grad_norm": 0.45974788723278426,
+      "learning_rate": 3.605059067105887e-07,
+      "loss": 0.1372,
+      "step": 7999
+    },
+    {
+      "epoch": 2.6715645349807984,
+      "grad_norm": 0.5306840434506943,
+      "learning_rate": 3.59781720089627e-07,
+      "loss": 0.1524,
+      "step": 8000
+    },
+    {
+      "epoch": 2.6718984805476707,
+      "grad_norm": 0.5618883169905055,
+      "learning_rate": 3.5905823442827393e-07,
+      "loss": 0.1654,
+      "step": 8001
+    },
+    {
+      "epoch": 2.6722324261145434,
+      "grad_norm": 0.5110715136379484,
+      "learning_rate": 3.583354498358188e-07,
+      "loss": 0.1548,
+      "step": 8002
+    },
+    {
+      "epoch": 2.672566371681416,
+      "grad_norm": 0.5165177851710339,
+      "learning_rate": 3.576133664214476e-07,
+      "loss": 0.1534,
+      "step": 8003
+    },
+    {
+      "epoch": 2.6729003172482884,
+      "grad_norm": 0.5427360752665066,
+      "learning_rate": 3.568919842942409e-07,
+      "loss": 0.1529,
+      "step": 8004
+    },
+    {
+      "epoch": 2.673234262815161,
+      "grad_norm": 0.5339557782418821,
+      "learning_rate": 3.5617130356316977e-07,
+      "loss": 0.1593,
+      "step": 8005
+    },
+    {
+      "epoch": 2.673568208382034,
+      "grad_norm": 0.556828539859666,
+      "learning_rate": 3.554513243371038e-07,
+      "loss": 0.1574,
+      "step": 8006
+    },
+    {
+      "epoch": 2.673902153948906,
+      "grad_norm": 0.538619853386981,
+      "learning_rate": 3.5473204672480224e-07,
+      "loss": 0.1667,
+      "step": 8007
+    },
+    {
+      "epoch": 2.674236099515779,
+      "grad_norm": 0.5434513800104122,
+      "learning_rate": 3.5401347083492077e-07,
+      "loss": 0.1479,
+      "step": 8008
+    },
+    {
+      "epoch": 2.6745700450826515,
+      "grad_norm": 0.5290721889738227,
+      "learning_rate": 3.532955967760093e-07,
+      "loss": 0.1585,
+      "step": 8009
+    },
+    {
+      "epoch": 2.674903990649524,
+      "grad_norm": 0.49434759453802185,
+      "learning_rate": 3.5257842465651226e-07,
+      "loss": 0.1473,
+      "step": 8010
+    },
+    {
+      "epoch": 2.675237936216397,
+      "grad_norm": 0.5113588034851804,
+      "learning_rate": 3.5186195458476515e-07,
+      "loss": 0.1493,
+      "step": 8011
+    },
+    {
+      "epoch": 2.675571881783269,
+      "grad_norm": 0.5688152683995327,
+      "learning_rate": 3.5114618666900023e-07,
+      "loss": 0.1628,
+      "step": 8012
+    },
+    {
+      "epoch": 2.675905827350142,
+      "grad_norm": 0.5039811736577918,
+      "learning_rate": 3.5043112101734166e-07,
+      "loss": 0.151,
+      "step": 8013
+    },
+    {
+      "epoch": 2.6762397729170146,
+      "grad_norm": 0.4828551437518733,
+      "learning_rate": 3.4971675773780913e-07,
+      "loss": 0.1432,
+      "step": 8014
+    },
+    {
+      "epoch": 2.676573718483887,
+      "grad_norm": 0.5549856306713297,
+      "learning_rate": 3.490030969383157e-07,
+      "loss": 0.1592,
+      "step": 8015
+    },
+    {
+      "epoch": 2.6769076640507596,
+      "grad_norm": 0.5379246934070531,
+      "learning_rate": 3.482901387266685e-07,
+      "loss": 0.1618,
+      "step": 8016
+    },
+    {
+      "epoch": 2.6772416096176324,
+      "grad_norm": 0.5072517308727558,
+      "learning_rate": 3.475778832105681e-07,
+      "loss": 0.1423,
+      "step": 8017
+    },
+    {
+      "epoch": 2.677575555184505,
+      "grad_norm": 0.4998020934814151,
+      "learning_rate": 3.468663304976089e-07,
+      "loss": 0.1552,
+      "step": 8018
+    },
+    {
+      "epoch": 2.677909500751378,
+      "grad_norm": 0.5309140304675917,
+      "learning_rate": 3.4615548069527883e-07,
+      "loss": 0.1557,
+      "step": 8019
+    },
+    {
+      "epoch": 2.67824344631825,
+      "grad_norm": 0.5489735365073403,
+      "learning_rate": 3.4544533391096093e-07,
+      "loss": 0.154,
+      "step": 8020
+    },
+    {
+      "epoch": 2.678577391885123,
+      "grad_norm": 0.6022169835721196,
+      "learning_rate": 3.4473589025193155e-07,
+      "loss": 0.1569,
+      "step": 8021
+    },
+    {
+      "epoch": 2.6789113374519955,
+      "grad_norm": 0.5545056781397845,
+      "learning_rate": 3.440271498253589e-07,
+      "loss": 0.1612,
+      "step": 8022
+    },
+    {
+      "epoch": 2.6792452830188678,
+      "grad_norm": 0.5777137091933007,
+      "learning_rate": 3.433191127383079e-07,
+      "loss": 0.1696,
+      "step": 8023
+    },
+    {
+      "epoch": 2.6795792285857405,
+      "grad_norm": 0.496180545244153,
+      "learning_rate": 3.4261177909773624e-07,
+      "loss": 0.1493,
+      "step": 8024
+    },
+    {
+      "epoch": 2.679913174152613,
+      "grad_norm": 0.5524161786976409,
+      "learning_rate": 3.419051490104935e-07,
+      "loss": 0.1599,
+      "step": 8025
+    },
+    {
+      "epoch": 2.6802471197194855,
+      "grad_norm": 0.5548262424062345,
+      "learning_rate": 3.4119922258332496e-07,
+      "loss": 0.1652,
+      "step": 8026
+    },
+    {
+      "epoch": 2.680581065286358,
+      "grad_norm": 0.5153735852895094,
+      "learning_rate": 3.4049399992287067e-07,
+      "loss": 0.1486,
+      "step": 8027
+    },
+    {
+      "epoch": 2.680915010853231,
+      "grad_norm": 0.5127645749070047,
+      "learning_rate": 3.3978948113566056e-07,
+      "loss": 0.1538,
+      "step": 8028
+    },
+    {
+      "epoch": 2.6812489564201036,
+      "grad_norm": 0.5397099844954941,
+      "learning_rate": 3.390856663281228e-07,
+      "loss": 0.1595,
+      "step": 8029
+    },
+    {
+      "epoch": 2.6815829019869764,
+      "grad_norm": 0.5237062122977899,
+      "learning_rate": 3.3838255560657453e-07,
+      "loss": 0.1513,
+      "step": 8030
+    },
+    {
+      "epoch": 2.6819168475538486,
+      "grad_norm": 0.5395864823693816,
+      "learning_rate": 3.3768014907722966e-07,
+      "loss": 0.1548,
+      "step": 8031
+    },
+    {
+      "epoch": 2.6822507931207213,
+      "grad_norm": 0.4867347579548099,
+      "learning_rate": 3.369784468461956e-07,
+      "loss": 0.1428,
+      "step": 8032
+    },
+    {
+      "epoch": 2.682584738687594,
+      "grad_norm": 0.5527869275755876,
+      "learning_rate": 3.3627744901947313e-07,
+      "loss": 0.1527,
+      "step": 8033
+    },
+    {
+      "epoch": 2.6829186842544663,
+      "grad_norm": 0.5023108032433418,
+      "learning_rate": 3.3557715570295523e-07,
+      "loss": 0.1514,
+      "step": 8034
+    },
+    {
+      "epoch": 2.683252629821339,
+      "grad_norm": 0.5360412608959365,
+      "learning_rate": 3.3487756700243014e-07,
+      "loss": 0.1604,
+      "step": 8035
+    },
+    {
+      "epoch": 2.6835865753882118,
+      "grad_norm": 0.5424139704297755,
+      "learning_rate": 3.341786830235777e-07,
+      "loss": 0.1544,
+      "step": 8036
+    },
+    {
+      "epoch": 2.6839205209550845,
+      "grad_norm": 0.5048209881726411,
+      "learning_rate": 3.334805038719735e-07,
+      "loss": 0.1545,
+      "step": 8037
+    },
+    {
+      "epoch": 2.6842544665219568,
+      "grad_norm": 0.5486406304826595,
+      "learning_rate": 3.3278302965308593e-07,
+      "loss": 0.1686,
+      "step": 8038
+    },
+    {
+      "epoch": 2.6845884120888295,
+      "grad_norm": 0.5300204256810848,
+      "learning_rate": 3.3208626047227687e-07,
+      "loss": 0.1601,
+      "step": 8039
+    },
+    {
+      "epoch": 2.684922357655702,
+      "grad_norm": 0.5323499926681304,
+      "learning_rate": 3.313901964348004e-07,
+      "loss": 0.1554,
+      "step": 8040
+    },
+    {
+      "epoch": 2.685256303222575,
+      "grad_norm": 0.5355346384229268,
+      "learning_rate": 3.306948376458069e-07,
+      "loss": 0.163,
+      "step": 8041
+    },
+    {
+      "epoch": 2.685590248789447,
+      "grad_norm": 0.5435803333076292,
+      "learning_rate": 3.3000018421033675e-07,
+      "loss": 0.1511,
+      "step": 8042
+    },
+    {
+      "epoch": 2.68592419435632,
+      "grad_norm": 0.5284470262839093,
+      "learning_rate": 3.29306236233326e-07,
+      "loss": 0.1517,
+      "step": 8043
+    },
+    {
+      "epoch": 2.6862581399231926,
+      "grad_norm": 0.5375116338923487,
+      "learning_rate": 3.286129938196048e-07,
+      "loss": 0.149,
+      "step": 8044
+    },
+    {
+      "epoch": 2.686592085490065,
+      "grad_norm": 0.5196788029953022,
+      "learning_rate": 3.279204570738936e-07,
+      "loss": 0.1494,
+      "step": 8045
+    },
+    {
+      "epoch": 2.6869260310569376,
+      "grad_norm": 0.5989969976700711,
+      "learning_rate": 3.272286261008095e-07,
+      "loss": 0.1594,
+      "step": 8046
+    },
+    {
+      "epoch": 2.6872599766238103,
+      "grad_norm": 0.533802596049426,
+      "learning_rate": 3.2653750100486213e-07,
+      "loss": 0.1606,
+      "step": 8047
+    },
+    {
+      "epoch": 2.687593922190683,
+      "grad_norm": 0.5985362027338466,
+      "learning_rate": 3.25847081890453e-07,
+      "loss": 0.1645,
+      "step": 8048
+    },
+    {
+      "epoch": 2.6879278677575558,
+      "grad_norm": 0.5123641039133764,
+      "learning_rate": 3.251573688618781e-07,
+      "loss": 0.1541,
+      "step": 8049
+    },
+    {
+      "epoch": 2.688261813324428,
+      "grad_norm": 0.5633459684436906,
+      "learning_rate": 3.2446836202332854e-07,
+      "loss": 0.1641,
+      "step": 8050
+    },
+    {
+      "epoch": 2.6885957588913008,
+      "grad_norm": 0.5383588555865818,
+      "learning_rate": 3.237800614788844e-07,
+      "loss": 0.1596,
+      "step": 8051
+    },
+    {
+      "epoch": 2.6889297044581735,
+      "grad_norm": 0.5287494546807463,
+      "learning_rate": 3.230924673325231e-07,
+      "loss": 0.1483,
+      "step": 8052
+    },
+    {
+      "epoch": 2.6892636500250457,
+      "grad_norm": 0.5349375464782349,
+      "learning_rate": 3.2240557968811315e-07,
+      "loss": 0.1511,
+      "step": 8053
+    },
+    {
+      "epoch": 2.6895975955919185,
+      "grad_norm": 0.5826177066382261,
+      "learning_rate": 3.217193986494177e-07,
+      "loss": 0.1605,
+      "step": 8054
+    },
+    {
+      "epoch": 2.689931541158791,
+      "grad_norm": 0.5289988608050177,
+      "learning_rate": 3.2103392432009105e-07,
+      "loss": 0.1425,
+      "step": 8055
+    },
+    {
+      "epoch": 2.6902654867256635,
+      "grad_norm": 0.5460451950316173,
+      "learning_rate": 3.203491568036843e-07,
+      "loss": 0.1566,
+      "step": 8056
+    },
+    {
+      "epoch": 2.690599432292536,
+      "grad_norm": 0.47868697237246965,
+      "learning_rate": 3.196650962036374e-07,
+      "loss": 0.148,
+      "step": 8057
+    },
+    {
+      "epoch": 2.690933377859409,
+      "grad_norm": 0.5647617020216217,
+      "learning_rate": 3.189817426232883e-07,
+      "loss": 0.1633,
+      "step": 8058
+    },
+    {
+      "epoch": 2.6912673234262816,
+      "grad_norm": 0.5517400532178436,
+      "learning_rate": 3.182990961658633e-07,
+      "loss": 0.1611,
+      "step": 8059
+    },
+    {
+      "epoch": 2.6916012689931543,
+      "grad_norm": 0.5601119667361114,
+      "learning_rate": 3.1761715693448546e-07,
+      "loss": 0.1586,
+      "step": 8060
+    },
+    {
+      "epoch": 2.6919352145600266,
+      "grad_norm": 0.5071398052612933,
+      "learning_rate": 3.1693592503216795e-07,
+      "loss": 0.1504,
+      "step": 8061
+    },
+    {
+      "epoch": 2.6922691601268993,
+      "grad_norm": 0.5063705615598098,
+      "learning_rate": 3.162554005618218e-07,
+      "loss": 0.1574,
+      "step": 8062
+    },
+    {
+      "epoch": 2.692603105693772,
+      "grad_norm": 0.5263427681179844,
+      "learning_rate": 3.155755836262464e-07,
+      "loss": 0.1428,
+      "step": 8063
+    },
+    {
+      "epoch": 2.6929370512606443,
+      "grad_norm": 0.5440253785171091,
+      "learning_rate": 3.148964743281363e-07,
+      "loss": 0.1609,
+      "step": 8064
+    },
+    {
+      "epoch": 2.693270996827517,
+      "grad_norm": 0.5213216657932812,
+      "learning_rate": 3.1421807277007885e-07,
+      "loss": 0.1525,
+      "step": 8065
+    },
+    {
+      "epoch": 2.6936049423943897,
+      "grad_norm": 0.5149269451276092,
+      "learning_rate": 3.1354037905455547e-07,
+      "loss": 0.1546,
+      "step": 8066
+    },
+    {
+      "epoch": 2.6939388879612625,
+      "grad_norm": 0.5446427194292313,
+      "learning_rate": 3.1286339328393755e-07,
+      "loss": 0.166,
+      "step": 8067
+    },
+    {
+      "epoch": 2.694272833528135,
+      "grad_norm": 0.5545769562726003,
+      "learning_rate": 3.1218711556049494e-07,
+      "loss": 0.1616,
+      "step": 8068
+    },
+    {
+      "epoch": 2.6946067790950075,
+      "grad_norm": 0.5233064501341601,
+      "learning_rate": 3.115115459863849e-07,
+      "loss": 0.1564,
+      "step": 8069
+    },
+    {
+      "epoch": 2.69494072466188,
+      "grad_norm": 0.5760191622579325,
+      "learning_rate": 3.108366846636618e-07,
+      "loss": 0.162,
+      "step": 8070
+    },
+    {
+      "epoch": 2.695274670228753,
+      "grad_norm": 0.52977035763498,
+      "learning_rate": 3.101625316942697e-07,
+      "loss": 0.1517,
+      "step": 8071
+    },
+    {
+      "epoch": 2.695608615795625,
+      "grad_norm": 0.5592586244479276,
+      "learning_rate": 3.094890871800488e-07,
+      "loss": 0.1644,
+      "step": 8072
+    },
+    {
+      "epoch": 2.695942561362498,
+      "grad_norm": 0.5035640831811913,
+      "learning_rate": 3.0881635122273047e-07,
+      "loss": 0.1459,
+      "step": 8073
+    },
+    {
+      "epoch": 2.6962765069293706,
+      "grad_norm": 0.5410337877985646,
+      "learning_rate": 3.0814432392393847e-07,
+      "loss": 0.1599,
+      "step": 8074
+    },
+    {
+      "epoch": 2.696610452496243,
+      "grad_norm": 0.5273773774689604,
+      "learning_rate": 3.074730053851921e-07,
+      "loss": 0.1553,
+      "step": 8075
+    },
+    {
+      "epoch": 2.6969443980631156,
+      "grad_norm": 0.48100970564002676,
+      "learning_rate": 3.068023957078997e-07,
+      "loss": 0.1479,
+      "step": 8076
+    },
+    {
+      "epoch": 2.6972783436299883,
+      "grad_norm": 0.49067868878960497,
+      "learning_rate": 3.061324949933675e-07,
+      "loss": 0.1342,
+      "step": 8077
+    },
+    {
+      "epoch": 2.697612289196861,
+      "grad_norm": 0.5583056416030368,
+      "learning_rate": 3.054633033427884e-07,
+      "loss": 0.1661,
+      "step": 8078
+    },
+    {
+      "epoch": 2.6979462347637337,
+      "grad_norm": 0.5215917195872181,
+      "learning_rate": 3.0479482085725545e-07,
+      "loss": 0.1538,
+      "step": 8079
+    },
+    {
+      "epoch": 2.698280180330606,
+      "grad_norm": 0.5729067492916314,
+      "learning_rate": 3.0412704763774836e-07,
+      "loss": 0.1695,
+      "step": 8080
+    },
+    {
+      "epoch": 2.6986141258974787,
+      "grad_norm": 0.5332376143310963,
+      "learning_rate": 3.034599837851432e-07,
+      "loss": 0.1607,
+      "step": 8081
+    },
+    {
+      "epoch": 2.6989480714643515,
+      "grad_norm": 0.4942390417812226,
+      "learning_rate": 3.027936294002071e-07,
+      "loss": 0.1476,
+      "step": 8082
+    },
+    {
+      "epoch": 2.6992820170312237,
+      "grad_norm": 0.5414771162644846,
+      "learning_rate": 3.021279845836017e-07,
+      "loss": 0.1559,
+      "step": 8083
+    },
+    {
+      "epoch": 2.6996159625980964,
+      "grad_norm": 0.5527997876111395,
+      "learning_rate": 3.0146304943587833e-07,
+      "loss": 0.1595,
+      "step": 8084
+    },
+    {
+      "epoch": 2.699949908164969,
+      "grad_norm": 0.5062233776814364,
+      "learning_rate": 3.007988240574866e-07,
+      "loss": 0.1522,
+      "step": 8085
+    },
+    {
+      "epoch": 2.7002838537318414,
+      "grad_norm": 0.5204497543953683,
+      "learning_rate": 3.0013530854876296e-07,
+      "loss": 0.1471,
+      "step": 8086
+    },
+    {
+      "epoch": 2.700617799298714,
+      "grad_norm": 0.5209478565356521,
+      "learning_rate": 2.9947250300994046e-07,
+      "loss": 0.1586,
+      "step": 8087
+    },
+    {
+      "epoch": 2.700951744865587,
+      "grad_norm": 0.5312789973530363,
+      "learning_rate": 2.98810407541143e-07,
+      "loss": 0.1588,
+      "step": 8088
+    },
+    {
+      "epoch": 2.7012856904324596,
+      "grad_norm": 0.5075938953896761,
+      "learning_rate": 2.9814902224238886e-07,
+      "loss": 0.1484,
+      "step": 8089
+    },
+    {
+      "epoch": 2.7016196359993323,
+      "grad_norm": 0.5160275474594337,
+      "learning_rate": 2.974883472135859e-07,
+      "loss": 0.147,
+      "step": 8090
+    },
+    {
+      "epoch": 2.7019535815662046,
+      "grad_norm": 0.5232544001518299,
+      "learning_rate": 2.968283825545398e-07,
+      "loss": 0.1496,
+      "step": 8091
+    },
+    {
+      "epoch": 2.7022875271330773,
+      "grad_norm": 0.5034655279070774,
+      "learning_rate": 2.961691283649437e-07,
+      "loss": 0.1468,
+      "step": 8092
+    },
+    {
+      "epoch": 2.70262147269995,
+      "grad_norm": 0.5510510157293431,
+      "learning_rate": 2.955105847443873e-07,
+      "loss": 0.1611,
+      "step": 8093
+    },
+    {
+      "epoch": 2.7029554182668223,
+      "grad_norm": 0.5315473818061955,
+      "learning_rate": 2.9485275179235e-07,
+      "loss": 0.15,
+      "step": 8094
+    },
+    {
+      "epoch": 2.703289363833695,
+      "grad_norm": 0.536253149180482,
+      "learning_rate": 2.9419562960820656e-07,
+      "loss": 0.1589,
+      "step": 8095
+    },
+    {
+      "epoch": 2.7036233094005677,
+      "grad_norm": 0.5306591757905975,
+      "learning_rate": 2.9353921829122167e-07,
+      "loss": 0.153,
+      "step": 8096
+    },
+    {
+      "epoch": 2.7039572549674404,
+      "grad_norm": 0.5436819952810656,
+      "learning_rate": 2.928835179405548e-07,
+      "loss": 0.1599,
+      "step": 8097
+    },
+    {
+      "epoch": 2.704291200534313,
+      "grad_norm": 0.5402673068926906,
+      "learning_rate": 2.922285286552579e-07,
+      "loss": 0.1529,
+      "step": 8098
+    },
+    {
+      "epoch": 2.7046251461011854,
+      "grad_norm": 0.5004065904993527,
+      "learning_rate": 2.915742505342728e-07,
+      "loss": 0.1487,
+      "step": 8099
+    },
+    {
+      "epoch": 2.704959091668058,
+      "grad_norm": 0.5229279567002475,
+      "learning_rate": 2.9092068367643776e-07,
+      "loss": 0.1545,
+      "step": 8100
+    },
+    {
+      "epoch": 2.705293037234931,
+      "grad_norm": 0.48729652498503284,
+      "learning_rate": 2.902678281804805e-07,
+      "loss": 0.1414,
+      "step": 8101
+    },
+    {
+      "epoch": 2.705626982801803,
+      "grad_norm": 0.4747083845884613,
+      "learning_rate": 2.896156841450232e-07,
+      "loss": 0.1441,
+      "step": 8102
+    },
+    {
+      "epoch": 2.705960928368676,
+      "grad_norm": 0.5342365800558092,
+      "learning_rate": 2.8896425166857976e-07,
+      "loss": 0.1543,
+      "step": 8103
+    },
+    {
+      "epoch": 2.7062948739355486,
+      "grad_norm": 0.5292850700939392,
+      "learning_rate": 2.8831353084955717e-07,
+      "loss": 0.1537,
+      "step": 8104
+    },
+    {
+      "epoch": 2.706628819502421,
+      "grad_norm": 0.5500522283182325,
+      "learning_rate": 2.8766352178625387e-07,
+      "loss": 0.1616,
+      "step": 8105
+    },
+    {
+      "epoch": 2.7069627650692936,
+      "grad_norm": 0.5134535972496134,
+      "learning_rate": 2.87014224576862e-07,
+      "loss": 0.1523,
+      "step": 8106
+    },
+    {
+      "epoch": 2.7072967106361663,
+      "grad_norm": 0.5471782056045605,
+      "learning_rate": 2.863656393194636e-07,
+      "loss": 0.1475,
+      "step": 8107
+    },
+    {
+      "epoch": 2.707630656203039,
+      "grad_norm": 0.5110760153803138,
+      "learning_rate": 2.8571776611203804e-07,
+      "loss": 0.1513,
+      "step": 8108
+    },
+    {
+      "epoch": 2.7079646017699117,
+      "grad_norm": 0.5385910497219654,
+      "learning_rate": 2.850706050524521e-07,
+      "loss": 0.1604,
+      "step": 8109
+    },
+    {
+      "epoch": 2.708298547336784,
+      "grad_norm": 0.5334396636556735,
+      "learning_rate": 2.844241562384686e-07,
+      "loss": 0.1556,
+      "step": 8110
+    },
+    {
+      "epoch": 2.7086324929036567,
+      "grad_norm": 0.5204610290097634,
+      "learning_rate": 2.8377841976773955e-07,
+      "loss": 0.1543,
+      "step": 8111
+    },
+    {
+      "epoch": 2.7089664384705294,
+      "grad_norm": 0.5496955299983243,
+      "learning_rate": 2.83133395737813e-07,
+      "loss": 0.1698,
+      "step": 8112
+    },
+    {
+      "epoch": 2.7093003840374017,
+      "grad_norm": 0.5536709752852906,
+      "learning_rate": 2.824890842461242e-07,
+      "loss": 0.1603,
+      "step": 8113
+    },
+    {
+      "epoch": 2.7096343296042744,
+      "grad_norm": 0.5010653680915442,
+      "learning_rate": 2.818454853900082e-07,
+      "loss": 0.1461,
+      "step": 8114
+    },
+    {
+      "epoch": 2.709968275171147,
+      "grad_norm": 0.5267900896237233,
+      "learning_rate": 2.8120259926668505e-07,
+      "loss": 0.152,
+      "step": 8115
+    },
+    {
+      "epoch": 2.71030222073802,
+      "grad_norm": 0.5478437369126266,
+      "learning_rate": 2.8056042597327196e-07,
+      "loss": 0.1687,
+      "step": 8116
+    },
+    {
+      "epoch": 2.7106361663048926,
+      "grad_norm": 0.5359815228385162,
+      "learning_rate": 2.799189656067758e-07,
+      "loss": 0.156,
+      "step": 8117
+    },
+    {
+      "epoch": 2.710970111871765,
+      "grad_norm": 0.5151592514466387,
+      "learning_rate": 2.792782182640974e-07,
+      "loss": 0.153,
+      "step": 8118
+    },
+    {
+      "epoch": 2.7113040574386376,
+      "grad_norm": 0.5344553467307052,
+      "learning_rate": 2.7863818404202823e-07,
+      "loss": 0.16,
+      "step": 8119
+    },
+    {
+      "epoch": 2.7116380030055103,
+      "grad_norm": 0.5700785819621733,
+      "learning_rate": 2.7799886303725376e-07,
+      "loss": 0.1635,
+      "step": 8120
+    },
+    {
+      "epoch": 2.7119719485723826,
+      "grad_norm": 0.5333711626924268,
+      "learning_rate": 2.7736025534635115e-07,
+      "loss": 0.1471,
+      "step": 8121
+    },
+    {
+      "epoch": 2.7123058941392553,
+      "grad_norm": 0.5161369376262971,
+      "learning_rate": 2.767223610657888e-07,
+      "loss": 0.1532,
+      "step": 8122
+    },
+    {
+      "epoch": 2.712639839706128,
+      "grad_norm": 0.579164591351255,
+      "learning_rate": 2.7608518029192897e-07,
+      "loss": 0.1704,
+      "step": 8123
+    },
+    {
+      "epoch": 2.7129737852730003,
+      "grad_norm": 0.556780216361864,
+      "learning_rate": 2.7544871312102485e-07,
+      "loss": 0.1614,
+      "step": 8124
+    },
+    {
+      "epoch": 2.713307730839873,
+      "grad_norm": 0.5134496409120298,
+      "learning_rate": 2.7481295964922216e-07,
+      "loss": 0.1524,
+      "step": 8125
+    },
+    {
+      "epoch": 2.7136416764067457,
+      "grad_norm": 0.5205512225709891,
+      "learning_rate": 2.7417791997255916e-07,
+      "loss": 0.1511,
+      "step": 8126
+    },
+    {
+      "epoch": 2.7139756219736184,
+      "grad_norm": 0.5524009827531768,
+      "learning_rate": 2.735435941869663e-07,
+      "loss": 0.1591,
+      "step": 8127
+    },
+    {
+      "epoch": 2.714309567540491,
+      "grad_norm": 0.5320081308305175,
+      "learning_rate": 2.7290998238826584e-07,
+      "loss": 0.1518,
+      "step": 8128
+    },
+    {
+      "epoch": 2.7146435131073634,
+      "grad_norm": 0.5274646254287043,
+      "learning_rate": 2.7227708467217227e-07,
+      "loss": 0.1586,
+      "step": 8129
+    },
+    {
+      "epoch": 2.714977458674236,
+      "grad_norm": 0.5314673716570912,
+      "learning_rate": 2.71644901134292e-07,
+      "loss": 0.1594,
+      "step": 8130
+    },
+    {
+      "epoch": 2.715311404241109,
+      "grad_norm": 0.5680112090581357,
+      "learning_rate": 2.7101343187012354e-07,
+      "loss": 0.166,
+      "step": 8131
+    },
+    {
+      "epoch": 2.715645349807981,
+      "grad_norm": 0.5302166166756883,
+      "learning_rate": 2.7038267697505894e-07,
+      "loss": 0.1601,
+      "step": 8132
+    },
+    {
+      "epoch": 2.715979295374854,
+      "grad_norm": 0.5334984611340821,
+      "learning_rate": 2.697526365443803e-07,
+      "loss": 0.1538,
+      "step": 8133
+    },
+    {
+      "epoch": 2.7163132409417265,
+      "grad_norm": 0.533924310055932,
+      "learning_rate": 2.691233106732627e-07,
+      "loss": 0.1499,
+      "step": 8134
+    },
+    {
+      "epoch": 2.716647186508599,
+      "grad_norm": 0.5165454426854755,
+      "learning_rate": 2.684946994567733e-07,
+      "loss": 0.1507,
+      "step": 8135
+    },
+    {
+      "epoch": 2.7169811320754715,
+      "grad_norm": 0.5319683743964586,
+      "learning_rate": 2.678668029898712e-07,
+      "loss": 0.1534,
+      "step": 8136
+    },
+    {
+      "epoch": 2.7173150776423443,
+      "grad_norm": 0.49055927054446136,
+      "learning_rate": 2.672396213674072e-07,
+      "loss": 0.1473,
+      "step": 8137
+    },
+    {
+      "epoch": 2.717649023209217,
+      "grad_norm": 0.5013407224064176,
+      "learning_rate": 2.66613154684125e-07,
+      "loss": 0.1515,
+      "step": 8138
+    },
+    {
+      "epoch": 2.7179829687760897,
+      "grad_norm": 0.557314050415258,
+      "learning_rate": 2.659874030346604e-07,
+      "loss": 0.16,
+      "step": 8139
+    },
+    {
+      "epoch": 2.718316914342962,
+      "grad_norm": 0.5188704409179276,
+      "learning_rate": 2.653623665135391e-07,
+      "loss": 0.1485,
+      "step": 8140
+    },
+    {
+      "epoch": 2.7186508599098347,
+      "grad_norm": 0.4733489482475054,
+      "learning_rate": 2.6473804521518097e-07,
+      "loss": 0.1369,
+      "step": 8141
+    },
+    {
+      "epoch": 2.7189848054767074,
+      "grad_norm": 0.5158640174213729,
+      "learning_rate": 2.641144392338968e-07,
+      "loss": 0.1615,
+      "step": 8142
+    },
+    {
+      "epoch": 2.7193187510435797,
+      "grad_norm": 0.5175936291666983,
+      "learning_rate": 2.6349154866389e-07,
+      "loss": 0.1487,
+      "step": 8143
+    },
+    {
+      "epoch": 2.7196526966104524,
+      "grad_norm": 0.4851398068455481,
+      "learning_rate": 2.6286937359925545e-07,
+      "loss": 0.1484,
+      "step": 8144
+    },
+    {
+      "epoch": 2.719986642177325,
+      "grad_norm": 0.4612050712346406,
+      "learning_rate": 2.622479141339801e-07,
+      "loss": 0.1426,
+      "step": 8145
+    },
+    {
+      "epoch": 2.720320587744198,
+      "grad_norm": 0.5103788486703124,
+      "learning_rate": 2.6162717036194274e-07,
+      "loss": 0.1545,
+      "step": 8146
+    },
+    {
+      "epoch": 2.7206545333110705,
+      "grad_norm": 0.5508718255938416,
+      "learning_rate": 2.610071423769128e-07,
+      "loss": 0.1552,
+      "step": 8147
+    },
+    {
+      "epoch": 2.720988478877943,
+      "grad_norm": 0.5482206957546422,
+      "learning_rate": 2.603878302725543e-07,
+      "loss": 0.162,
+      "step": 8148
+    },
+    {
+      "epoch": 2.7213224244448155,
+      "grad_norm": 0.5354816153222969,
+      "learning_rate": 2.5976923414242126e-07,
+      "loss": 0.1591,
+      "step": 8149
+    },
+    {
+      "epoch": 2.7216563700116883,
+      "grad_norm": 0.508383617507333,
+      "learning_rate": 2.5915135407996005e-07,
+      "loss": 0.1486,
+      "step": 8150
+    },
+    {
+      "epoch": 2.7219903155785605,
+      "grad_norm": 0.4846830693165029,
+      "learning_rate": 2.585341901785082e-07,
+      "loss": 0.1463,
+      "step": 8151
+    },
+    {
+      "epoch": 2.7223242611454332,
+      "grad_norm": 0.5589614124312149,
+      "learning_rate": 2.579177425312962e-07,
+      "loss": 0.1631,
+      "step": 8152
+    },
+    {
+      "epoch": 2.722658206712306,
+      "grad_norm": 0.5085255581901547,
+      "learning_rate": 2.5730201123144503e-07,
+      "loss": 0.156,
+      "step": 8153
+    },
+    {
+      "epoch": 2.7229921522791782,
+      "grad_norm": 0.5678891288995872,
+      "learning_rate": 2.566869963719681e-07,
+      "loss": 0.1583,
+      "step": 8154
+    },
+    {
+      "epoch": 2.723326097846051,
+      "grad_norm": 0.5160552742462742,
+      "learning_rate": 2.5607269804577174e-07,
+      "loss": 0.1487,
+      "step": 8155
+    },
+    {
+      "epoch": 2.7236600434129237,
+      "grad_norm": 0.5679998884899234,
+      "learning_rate": 2.5545911634565266e-07,
+      "loss": 0.1675,
+      "step": 8156
+    },
+    {
+      "epoch": 2.7239939889797964,
+      "grad_norm": 0.49259962902633603,
+      "learning_rate": 2.5484625136429854e-07,
+      "loss": 0.1528,
+      "step": 8157
+    },
+    {
+      "epoch": 2.724327934546669,
+      "grad_norm": 0.5321498066466733,
+      "learning_rate": 2.5423410319429075e-07,
+      "loss": 0.1574,
+      "step": 8158
+    },
+    {
+      "epoch": 2.7246618801135414,
+      "grad_norm": 0.5491822405598497,
+      "learning_rate": 2.5362267192810095e-07,
+      "loss": 0.1565,
+      "step": 8159
+    },
+    {
+      "epoch": 2.724995825680414,
+      "grad_norm": 0.5484143142827499,
+      "learning_rate": 2.530119576580936e-07,
+      "loss": 0.1648,
+      "step": 8160
+    },
+    {
+      "epoch": 2.725329771247287,
+      "grad_norm": 0.5123547918351172,
+      "learning_rate": 2.5240196047652377e-07,
+      "loss": 0.1561,
+      "step": 8161
+    },
+    {
+      "epoch": 2.725663716814159,
+      "grad_norm": 0.5079554577300277,
+      "learning_rate": 2.5179268047553937e-07,
+      "loss": 0.1473,
+      "step": 8162
+    },
+    {
+      "epoch": 2.725997662381032,
+      "grad_norm": 0.5048876817838505,
+      "learning_rate": 2.5118411774717857e-07,
+      "loss": 0.1455,
+      "step": 8163
+    },
+    {
+      "epoch": 2.7263316079479045,
+      "grad_norm": 0.5285246244585439,
+      "learning_rate": 2.5057627238337324e-07,
+      "loss": 0.1548,
+      "step": 8164
+    },
+    {
+      "epoch": 2.7266655535147772,
+      "grad_norm": 0.5301222062922429,
+      "learning_rate": 2.4996914447594334e-07,
+      "loss": 0.1638,
+      "step": 8165
+    },
+    {
+      "epoch": 2.72699949908165,
+      "grad_norm": 0.5486195673192618,
+      "learning_rate": 2.493627341166044e-07,
+      "loss": 0.1557,
+      "step": 8166
+    },
+    {
+      "epoch": 2.7273334446485222,
+      "grad_norm": 0.5147618754850339,
+      "learning_rate": 2.48757041396962e-07,
+      "loss": 0.1524,
+      "step": 8167
+    },
+    {
+      "epoch": 2.727667390215395,
+      "grad_norm": 0.47570026933007525,
+      "learning_rate": 2.481520664085113e-07,
+      "loss": 0.1414,
+      "step": 8168
+    },
+    {
+      "epoch": 2.7280013357822677,
+      "grad_norm": 0.5695233117084313,
+      "learning_rate": 2.4754780924264366e-07,
+      "loss": 0.1671,
+      "step": 8169
+    },
+    {
+      "epoch": 2.72833528134914,
+      "grad_norm": 0.5044921765852178,
+      "learning_rate": 2.4694426999063657e-07,
+      "loss": 0.1566,
+      "step": 8170
+    },
+    {
+      "epoch": 2.7286692269160127,
+      "grad_norm": 0.5059445019609731,
+      "learning_rate": 2.463414487436633e-07,
+      "loss": 0.151,
+      "step": 8171
+    },
+    {
+      "epoch": 2.7290031724828854,
+      "grad_norm": 0.5047425599919322,
+      "learning_rate": 2.4573934559278646e-07,
+      "loss": 0.1518,
+      "step": 8172
+    },
+    {
+      "epoch": 2.7293371180497576,
+      "grad_norm": 0.5073492122785407,
+      "learning_rate": 2.4513796062896166e-07,
+      "loss": 0.1472,
+      "step": 8173
+    },
+    {
+      "epoch": 2.7296710636166304,
+      "grad_norm": 0.5427657012511299,
+      "learning_rate": 2.4453729394303404e-07,
+      "loss": 0.1567,
+      "step": 8174
+    },
+    {
+      "epoch": 2.730005009183503,
+      "grad_norm": 0.5360771115077887,
+      "learning_rate": 2.439373456257427e-07,
+      "loss": 0.1562,
+      "step": 8175
+    },
+    {
+      "epoch": 2.730338954750376,
+      "grad_norm": 0.5371756975065656,
+      "learning_rate": 2.433381157677156e-07,
+      "loss": 0.1519,
+      "step": 8176
+    },
+    {
+      "epoch": 2.7306729003172485,
+      "grad_norm": 0.5073709026677747,
+      "learning_rate": 2.427396044594743e-07,
+      "loss": 0.1501,
+      "step": 8177
+    },
+    {
+      "epoch": 2.731006845884121,
+      "grad_norm": 0.6418391887464554,
+      "learning_rate": 2.421418117914298e-07,
+      "loss": 0.1538,
+      "step": 8178
+    },
+    {
+      "epoch": 2.7313407914509935,
+      "grad_norm": 0.5285420352825599,
+      "learning_rate": 2.415447378538871e-07,
+      "loss": 0.1597,
+      "step": 8179
+    },
+    {
+      "epoch": 2.7316747370178662,
+      "grad_norm": 0.594718310704362,
+      "learning_rate": 2.409483827370407e-07,
+      "loss": 0.1739,
+      "step": 8180
+    },
+    {
+      "epoch": 2.7320086825847385,
+      "grad_norm": 0.4977407152777746,
+      "learning_rate": 2.4035274653097797e-07,
+      "loss": 0.1474,
+      "step": 8181
+    },
+    {
+      "epoch": 2.732342628151611,
+      "grad_norm": 0.5529928738083278,
+      "learning_rate": 2.3975782932567473e-07,
+      "loss": 0.1595,
+      "step": 8182
+    },
+    {
+      "epoch": 2.732676573718484,
+      "grad_norm": 0.5064571077259249,
+      "learning_rate": 2.391636312110024e-07,
+      "loss": 0.146,
+      "step": 8183
+    },
+    {
+      "epoch": 2.733010519285356,
+      "grad_norm": 0.5539490252056876,
+      "learning_rate": 2.385701522767192e-07,
+      "loss": 0.1657,
+      "step": 8184
+    },
+    {
+      "epoch": 2.733344464852229,
+      "grad_norm": 0.5679411279580505,
+      "learning_rate": 2.3797739261247955e-07,
+      "loss": 0.1631,
+      "step": 8185
+    },
+    {
+      "epoch": 2.7336784104191016,
+      "grad_norm": 0.5833955890630944,
+      "learning_rate": 2.3738535230782568e-07,
+      "loss": 0.1643,
+      "step": 8186
+    },
+    {
+      "epoch": 2.7340123559859744,
+      "grad_norm": 0.5280431470599531,
+      "learning_rate": 2.3679403145219214e-07,
+      "loss": 0.1478,
+      "step": 8187
+    },
+    {
+      "epoch": 2.734346301552847,
+      "grad_norm": 0.5736288252761815,
+      "learning_rate": 2.362034301349053e-07,
+      "loss": 0.1657,
+      "step": 8188
+    },
+    {
+      "epoch": 2.7346802471197194,
+      "grad_norm": 0.5530244579608455,
+      "learning_rate": 2.3561354844518157e-07,
+      "loss": 0.1642,
+      "step": 8189
+    },
+    {
+      "epoch": 2.735014192686592,
+      "grad_norm": 0.5518902439512378,
+      "learning_rate": 2.3502438647213132e-07,
+      "loss": 0.1647,
+      "step": 8190
+    },
+    {
+      "epoch": 2.735348138253465,
+      "grad_norm": 0.5247716871272328,
+      "learning_rate": 2.3443594430475224e-07,
+      "loss": 0.1542,
+      "step": 8191
+    },
+    {
+      "epoch": 2.735682083820337,
+      "grad_norm": 0.5177961555436752,
+      "learning_rate": 2.3384822203193714e-07,
+      "loss": 0.1528,
+      "step": 8192
+    },
+    {
+      "epoch": 2.73601602938721,
+      "grad_norm": 0.5447187428407865,
+      "learning_rate": 2.332612197424672e-07,
+      "loss": 0.159,
+      "step": 8193
+    },
+    {
+      "epoch": 2.7363499749540825,
+      "grad_norm": 0.5400715115866794,
+      "learning_rate": 2.32674937525017e-07,
+      "loss": 0.1518,
+      "step": 8194
+    },
+    {
+      "epoch": 2.736683920520955,
+      "grad_norm": 0.518061402756023,
+      "learning_rate": 2.3208937546815026e-07,
+      "loss": 0.1563,
+      "step": 8195
+    },
+    {
+      "epoch": 2.737017866087828,
+      "grad_norm": 0.5073899376170143,
+      "learning_rate": 2.3150453366032445e-07,
+      "loss": 0.1533,
+      "step": 8196
+    },
+    {
+      "epoch": 2.7373518116547,
+      "grad_norm": 0.5308916049545908,
+      "learning_rate": 2.309204121898856e-07,
+      "loss": 0.1586,
+      "step": 8197
+    },
+    {
+      "epoch": 2.737685757221573,
+      "grad_norm": 0.5483262825685362,
+      "learning_rate": 2.3033701114507313e-07,
+      "loss": 0.1642,
+      "step": 8198
+    },
+    {
+      "epoch": 2.7380197027884456,
+      "grad_norm": 0.5920765856051325,
+      "learning_rate": 2.2975433061401541e-07,
+      "loss": 0.1734,
+      "step": 8199
+    },
+    {
+      "epoch": 2.738353648355318,
+      "grad_norm": 0.5495984257125057,
+      "learning_rate": 2.2917237068473484e-07,
+      "loss": 0.1591,
+      "step": 8200
+    },
+    {
+      "epoch": 2.7386875939221906,
+      "grad_norm": 0.543004345814433,
+      "learning_rate": 2.2859113144514055e-07,
+      "loss": 0.1581,
+      "step": 8201
+    },
+    {
+      "epoch": 2.7390215394890634,
+      "grad_norm": 0.49894758949368256,
+      "learning_rate": 2.2801061298303895e-07,
+      "loss": 0.154,
+      "step": 8202
+    },
+    {
+      "epoch": 2.7393554850559356,
+      "grad_norm": 0.5334926320488186,
+      "learning_rate": 2.2743081538612154e-07,
+      "loss": 0.1525,
+      "step": 8203
+    },
+    {
+      "epoch": 2.7396894306228083,
+      "grad_norm": 0.5413584792641948,
+      "learning_rate": 2.268517387419761e-07,
+      "loss": 0.1558,
+      "step": 8204
+    },
+    {
+      "epoch": 2.740023376189681,
+      "grad_norm": 0.5686222507446349,
+      "learning_rate": 2.2627338313807645e-07,
+      "loss": 0.1636,
+      "step": 8205
+    },
+    {
+      "epoch": 2.7403573217565538,
+      "grad_norm": 0.5348153988527491,
+      "learning_rate": 2.2569574866179166e-07,
+      "loss": 0.1592,
+      "step": 8206
+    },
+    {
+      "epoch": 2.7406912673234265,
+      "grad_norm": 0.5642140442258513,
+      "learning_rate": 2.2511883540037805e-07,
+      "loss": 0.1573,
+      "step": 8207
+    },
+    {
+      "epoch": 2.7410252128902988,
+      "grad_norm": 0.5024714154716956,
+      "learning_rate": 2.2454264344098865e-07,
+      "loss": 0.1479,
+      "step": 8208
+    },
+    {
+      "epoch": 2.7413591584571715,
+      "grad_norm": 0.517979664407899,
+      "learning_rate": 2.2396717287066106e-07,
+      "loss": 0.145,
+      "step": 8209
+    },
+    {
+      "epoch": 2.741693104024044,
+      "grad_norm": 0.5289911278223267,
+      "learning_rate": 2.233924237763291e-07,
+      "loss": 0.1537,
+      "step": 8210
+    },
+    {
+      "epoch": 2.7420270495909165,
+      "grad_norm": 0.519553175252985,
+      "learning_rate": 2.2281839624481328e-07,
+      "loss": 0.1495,
+      "step": 8211
+    },
+    {
+      "epoch": 2.742360995157789,
+      "grad_norm": 0.5335564407178233,
+      "learning_rate": 2.222450903628287e-07,
+      "loss": 0.1557,
+      "step": 8212
+    },
+    {
+      "epoch": 2.742694940724662,
+      "grad_norm": 0.5617213957103794,
+      "learning_rate": 2.2167250621697944e-07,
+      "loss": 0.1529,
+      "step": 8213
+    },
+    {
+      "epoch": 2.7430288862915346,
+      "grad_norm": 0.533394370268705,
+      "learning_rate": 2.2110064389376017e-07,
+      "loss": 0.1554,
+      "step": 8214
+    },
+    {
+      "epoch": 2.7433628318584073,
+      "grad_norm": 0.5242839412788696,
+      "learning_rate": 2.205295034795596e-07,
+      "loss": 0.1545,
+      "step": 8215
+    },
+    {
+      "epoch": 2.7436967774252796,
+      "grad_norm": 0.5180468545852273,
+      "learning_rate": 2.1995908506065366e-07,
+      "loss": 0.1562,
+      "step": 8216
+    },
+    {
+      "epoch": 2.7440307229921523,
+      "grad_norm": 0.5525667020861501,
+      "learning_rate": 2.1938938872321014e-07,
+      "loss": 0.153,
+      "step": 8217
+    },
+    {
+      "epoch": 2.744364668559025,
+      "grad_norm": 0.5713615936439899,
+      "learning_rate": 2.1882041455329073e-07,
+      "loss": 0.1606,
+      "step": 8218
+    },
+    {
+      "epoch": 2.7446986141258973,
+      "grad_norm": 0.5264399545928933,
+      "learning_rate": 2.1825216263684336e-07,
+      "loss": 0.1556,
+      "step": 8219
+    },
+    {
+      "epoch": 2.74503255969277,
+      "grad_norm": 0.4902137034825769,
+      "learning_rate": 2.176846330597099e-07,
+      "loss": 0.1515,
+      "step": 8220
+    },
+    {
+      "epoch": 2.7453665052596428,
+      "grad_norm": 0.5132192422380915,
+      "learning_rate": 2.1711782590762344e-07,
+      "loss": 0.148,
+      "step": 8221
+    },
+    {
+      "epoch": 2.745700450826515,
+      "grad_norm": 0.5177457173738288,
+      "learning_rate": 2.165517412662055e-07,
+      "loss": 0.1553,
+      "step": 8222
+    },
+    {
+      "epoch": 2.7460343963933878,
+      "grad_norm": 0.5390903052995786,
+      "learning_rate": 2.1598637922097098e-07,
+      "loss": 0.1537,
+      "step": 8223
+    },
+    {
+      "epoch": 2.7463683419602605,
+      "grad_norm": 0.5250086474438852,
+      "learning_rate": 2.1542173985732274e-07,
+      "loss": 0.1467,
+      "step": 8224
+    },
+    {
+      "epoch": 2.746702287527133,
+      "grad_norm": 0.4842237891444192,
+      "learning_rate": 2.148578232605575e-07,
+      "loss": 0.1439,
+      "step": 8225
+    },
+    {
+      "epoch": 2.747036233094006,
+      "grad_norm": 0.4949766102656577,
+      "learning_rate": 2.14294629515861e-07,
+      "loss": 0.1492,
+      "step": 8226
+    },
+    {
+      "epoch": 2.747370178660878,
+      "grad_norm": 0.5230593113441051,
+      "learning_rate": 2.137321587083119e-07,
+      "loss": 0.1517,
+      "step": 8227
+    },
+    {
+      "epoch": 2.747704124227751,
+      "grad_norm": 0.5123186979953736,
+      "learning_rate": 2.1317041092287548e-07,
+      "loss": 0.1467,
+      "step": 8228
+    },
+    {
+      "epoch": 2.7480380697946236,
+      "grad_norm": 0.5123779178667591,
+      "learning_rate": 2.126093862444123e-07,
+      "loss": 0.1604,
+      "step": 8229
+    },
+    {
+      "epoch": 2.748372015361496,
+      "grad_norm": 0.5225226900228949,
+      "learning_rate": 2.1204908475767005e-07,
+      "loss": 0.1521,
+      "step": 8230
+    },
+    {
+      "epoch": 2.7487059609283686,
+      "grad_norm": 0.5023831154513653,
+      "learning_rate": 2.114895065472905e-07,
+      "loss": 0.1449,
+      "step": 8231
+    },
+    {
+      "epoch": 2.7490399064952413,
+      "grad_norm": 0.5847888545067589,
+      "learning_rate": 2.109306516978038e-07,
+      "loss": 0.1677,
+      "step": 8232
+    },
+    {
+      "epoch": 2.7493738520621136,
+      "grad_norm": 0.49554499992889467,
+      "learning_rate": 2.1037252029363242e-07,
+      "loss": 0.1469,
+      "step": 8233
+    },
+    {
+      "epoch": 2.7497077976289863,
+      "grad_norm": 0.5556392349656654,
+      "learning_rate": 2.098151124190867e-07,
+      "loss": 0.164,
+      "step": 8234
+    },
+    {
+      "epoch": 2.750041743195859,
+      "grad_norm": 0.5186943346251153,
+      "learning_rate": 2.092584281583715e-07,
+      "loss": 0.1501,
+      "step": 8235
+    },
+    {
+      "epoch": 2.7503756887627318,
+      "grad_norm": 0.5265346475920051,
+      "learning_rate": 2.0870246759557956e-07,
+      "loss": 0.1491,
+      "step": 8236
+    },
+    {
+      "epoch": 2.7507096343296045,
+      "grad_norm": 0.5111789881374325,
+      "learning_rate": 2.0814723081469535e-07,
+      "loss": 0.1546,
+      "step": 8237
+    },
+    {
+      "epoch": 2.7510435798964767,
+      "grad_norm": 0.5401318445314278,
+      "learning_rate": 2.0759271789959513e-07,
+      "loss": 0.1534,
+      "step": 8238
+    },
+    {
+      "epoch": 2.7513775254633495,
+      "grad_norm": 0.534377040910652,
+      "learning_rate": 2.0703892893404299e-07,
+      "loss": 0.1541,
+      "step": 8239
+    },
+    {
+      "epoch": 2.751711471030222,
+      "grad_norm": 0.5721615250971579,
+      "learning_rate": 2.064858640016959e-07,
+      "loss": 0.1592,
+      "step": 8240
+    },
+    {
+      "epoch": 2.7520454165970945,
+      "grad_norm": 0.5306616003142766,
+      "learning_rate": 2.0593352318610093e-07,
+      "loss": 0.1585,
+      "step": 8241
+    },
+    {
+      "epoch": 2.752379362163967,
+      "grad_norm": 0.502808647723062,
+      "learning_rate": 2.0538190657069523e-07,
+      "loss": 0.1422,
+      "step": 8242
+    },
+    {
+      "epoch": 2.75271330773084,
+      "grad_norm": 0.5580436841293485,
+      "learning_rate": 2.048310142388077e-07,
+      "loss": 0.161,
+      "step": 8243
+    },
+    {
+      "epoch": 2.7530472532977126,
+      "grad_norm": 0.5351425052272484,
+      "learning_rate": 2.0428084627365729e-07,
+      "loss": 0.149,
+      "step": 8244
+    },
+    {
+      "epoch": 2.7533811988645853,
+      "grad_norm": 0.4946587104254831,
+      "learning_rate": 2.0373140275835203e-07,
+      "loss": 0.1361,
+      "step": 8245
+    },
+    {
+      "epoch": 2.7537151444314576,
+      "grad_norm": 0.5167893750141161,
+      "learning_rate": 2.0318268377589323e-07,
+      "loss": 0.1504,
+      "step": 8246
+    },
+    {
+      "epoch": 2.7540490899983303,
+      "grad_norm": 0.550129526570199,
+      "learning_rate": 2.026346894091702e-07,
+      "loss": 0.1593,
+      "step": 8247
+    },
+    {
+      "epoch": 2.754383035565203,
+      "grad_norm": 0.5486648519214529,
+      "learning_rate": 2.0208741974096445e-07,
+      "loss": 0.1584,
+      "step": 8248
+    },
+    {
+      "epoch": 2.7547169811320753,
+      "grad_norm": 0.5266687534141231,
+      "learning_rate": 2.0154087485394713e-07,
+      "loss": 0.1468,
+      "step": 8249
+    },
+    {
+      "epoch": 2.755050926698948,
+      "grad_norm": 0.5303620209898399,
+      "learning_rate": 2.0099505483068216e-07,
+      "loss": 0.1524,
+      "step": 8250
+    },
+    {
+      "epoch": 2.7553848722658207,
+      "grad_norm": 0.5017401492165686,
+      "learning_rate": 2.0044995975361914e-07,
+      "loss": 0.1476,
+      "step": 8251
+    },
+    {
+      "epoch": 2.755718817832693,
+      "grad_norm": 0.5512204576696247,
+      "learning_rate": 1.9990558970510388e-07,
+      "loss": 0.1627,
+      "step": 8252
+    },
+    {
+      "epoch": 2.7560527633995657,
+      "grad_norm": 0.5328549457695625,
+      "learning_rate": 1.9936194476736782e-07,
+      "loss": 0.1469,
+      "step": 8253
+    },
+    {
+      "epoch": 2.7563867089664384,
+      "grad_norm": 0.5332369435035683,
+      "learning_rate": 1.9881902502253525e-07,
+      "loss": 0.1586,
+      "step": 8254
+    },
+    {
+      "epoch": 2.756720654533311,
+      "grad_norm": 0.5442681969901398,
+      "learning_rate": 1.9827683055262114e-07,
+      "loss": 0.1625,
+      "step": 8255
+    },
+    {
+      "epoch": 2.757054600100184,
+      "grad_norm": 0.5073616134036689,
+      "learning_rate": 1.977353614395311e-07,
+      "loss": 0.1479,
+      "step": 8256
+    },
+    {
+      "epoch": 2.757388545667056,
+      "grad_norm": 0.5378551125440103,
+      "learning_rate": 1.971946177650591e-07,
+      "loss": 0.1518,
+      "step": 8257
+    },
+    {
+      "epoch": 2.757722491233929,
+      "grad_norm": 0.549121416484626,
+      "learning_rate": 1.966545996108915e-07,
+      "loss": 0.1548,
+      "step": 8258
+    },
+    {
+      "epoch": 2.7580564368008016,
+      "grad_norm": 0.550399829623678,
+      "learning_rate": 1.961153070586036e-07,
+      "loss": 0.1606,
+      "step": 8259
+    },
+    {
+      "epoch": 2.758390382367674,
+      "grad_norm": 0.5449800142650917,
+      "learning_rate": 1.9557674018966244e-07,
+      "loss": 0.1538,
+      "step": 8260
+    },
+    {
+      "epoch": 2.7587243279345466,
+      "grad_norm": 0.5752625807188275,
+      "learning_rate": 1.9503889908542572e-07,
+      "loss": 0.1673,
+      "step": 8261
+    },
+    {
+      "epoch": 2.7590582735014193,
+      "grad_norm": 0.5737368577234173,
+      "learning_rate": 1.9450178382713957e-07,
+      "loss": 0.1677,
+      "step": 8262
+    },
+    {
+      "epoch": 2.759392219068292,
+      "grad_norm": 0.5051673849530389,
+      "learning_rate": 1.9396539449594131e-07,
+      "loss": 0.1503,
+      "step": 8263
+    },
+    {
+      "epoch": 2.7597261646351647,
+      "grad_norm": 0.4914480531561124,
+      "learning_rate": 1.9342973117286056e-07,
+      "loss": 0.144,
+      "step": 8264
+    },
+    {
+      "epoch": 2.760060110202037,
+      "grad_norm": 0.5631823006449259,
+      "learning_rate": 1.9289479393881317e-07,
+      "loss": 0.165,
+      "step": 8265
+    },
+    {
+      "epoch": 2.7603940557689097,
+      "grad_norm": 0.5355383401496868,
+      "learning_rate": 1.9236058287460946e-07,
+      "loss": 0.1565,
+      "step": 8266
+    },
+    {
+      "epoch": 2.7607280013357824,
+      "grad_norm": 0.5450111267966022,
+      "learning_rate": 1.9182709806094823e-07,
+      "loss": 0.1658,
+      "step": 8267
+    },
+    {
+      "epoch": 2.7610619469026547,
+      "grad_norm": 0.5399961091071,
+      "learning_rate": 1.9129433957841781e-07,
+      "loss": 0.1513,
+      "step": 8268
+    },
+    {
+      "epoch": 2.7613958924695274,
+      "grad_norm": 0.5322165741350056,
+      "learning_rate": 1.907623075074988e-07,
+      "loss": 0.1516,
+      "step": 8269
+    },
+    {
+      "epoch": 2.7617298380364,
+      "grad_norm": 0.5481681573013868,
+      "learning_rate": 1.9023100192855914e-07,
+      "loss": 0.1608,
+      "step": 8270
+    },
+    {
+      "epoch": 2.7620637836032724,
+      "grad_norm": 0.544718465373997,
+      "learning_rate": 1.897004229218602e-07,
+      "loss": 0.1524,
+      "step": 8271
+    },
+    {
+      "epoch": 2.762397729170145,
+      "grad_norm": 0.5542344358536933,
+      "learning_rate": 1.8917057056755172e-07,
+      "loss": 0.1545,
+      "step": 8272
+    },
+    {
+      "epoch": 2.762731674737018,
+      "grad_norm": 0.4983410315703037,
+      "learning_rate": 1.8864144494567528e-07,
+      "loss": 0.1503,
+      "step": 8273
+    },
+    {
+      "epoch": 2.7630656203038906,
+      "grad_norm": 0.5290491211012668,
+      "learning_rate": 1.881130461361591e-07,
+      "loss": 0.1575,
+      "step": 8274
+    },
+    {
+      "epoch": 2.7633995658707633,
+      "grad_norm": 0.5105092858702401,
+      "learning_rate": 1.8758537421882662e-07,
+      "loss": 0.1462,
+      "step": 8275
+    },
+    {
+      "epoch": 2.7637335114376356,
+      "grad_norm": 0.5636688731155922,
+      "learning_rate": 1.870584292733868e-07,
+      "loss": 0.1596,
+      "step": 8276
+    },
+    {
+      "epoch": 2.7640674570045083,
+      "grad_norm": 0.547183058318822,
+      "learning_rate": 1.8653221137944155e-07,
+      "loss": 0.1678,
+      "step": 8277
+    },
+    {
+      "epoch": 2.764401402571381,
+      "grad_norm": 0.49128257503638373,
+      "learning_rate": 1.8600672061648283e-07,
+      "loss": 0.1442,
+      "step": 8278
+    },
+    {
+      "epoch": 2.7647353481382533,
+      "grad_norm": 0.5670280593322085,
+      "learning_rate": 1.8548195706389272e-07,
+      "loss": 0.1703,
+      "step": 8279
+    },
+    {
+      "epoch": 2.765069293705126,
+      "grad_norm": 0.4735398261466528,
+      "learning_rate": 1.849579208009411e-07,
+      "loss": 0.1331,
+      "step": 8280
+    },
+    {
+      "epoch": 2.7654032392719987,
+      "grad_norm": 0.5617939142612081,
+      "learning_rate": 1.844346119067919e-07,
+      "loss": 0.167,
+      "step": 8281
+    },
+    {
+      "epoch": 2.765737184838871,
+      "grad_norm": 0.5642449533519758,
+      "learning_rate": 1.8391203046049522e-07,
+      "loss": 0.1643,
+      "step": 8282
+    },
+    {
+      "epoch": 2.7660711304057437,
+      "grad_norm": 0.5244528138173514,
+      "learning_rate": 1.8339017654099344e-07,
+      "loss": 0.1534,
+      "step": 8283
+    },
+    {
+      "epoch": 2.7664050759726164,
+      "grad_norm": 0.5292603150792725,
+      "learning_rate": 1.828690502271202e-07,
+      "loss": 0.1611,
+      "step": 8284
+    },
+    {
+      "epoch": 2.766739021539489,
+      "grad_norm": 0.5377384654986338,
+      "learning_rate": 1.823486515975964e-07,
+      "loss": 0.1575,
+      "step": 8285
+    },
+    {
+      "epoch": 2.767072967106362,
+      "grad_norm": 0.5817154299132605,
+      "learning_rate": 1.818289807310347e-07,
+      "loss": 0.1533,
+      "step": 8286
+    },
+    {
+      "epoch": 2.767406912673234,
+      "grad_norm": 0.5228418557171426,
+      "learning_rate": 1.813100377059379e-07,
+      "loss": 0.1563,
+      "step": 8287
+    },
+    {
+      "epoch": 2.767740858240107,
+      "grad_norm": 0.5760879601371929,
+      "learning_rate": 1.8079182260069773e-07,
+      "loss": 0.1561,
+      "step": 8288
+    },
+    {
+      "epoch": 2.7680748038069796,
+      "grad_norm": 0.49105535586314847,
+      "learning_rate": 1.8027433549359764e-07,
+      "loss": 0.1542,
+      "step": 8289
+    },
+    {
+      "epoch": 2.768408749373852,
+      "grad_norm": 0.5680155038927197,
+      "learning_rate": 1.7975757646280955e-07,
+      "loss": 0.1668,
+      "step": 8290
+    },
+    {
+      "epoch": 2.7687426949407246,
+      "grad_norm": 0.5082235918459275,
+      "learning_rate": 1.792415455863955e-07,
+      "loss": 0.1557,
+      "step": 8291
+    },
+    {
+      "epoch": 2.7690766405075973,
+      "grad_norm": 0.5112458964828165,
+      "learning_rate": 1.7872624294230924e-07,
+      "loss": 0.1538,
+      "step": 8292
+    },
+    {
+      "epoch": 2.76941058607447,
+      "grad_norm": 0.5128409883230706,
+      "learning_rate": 1.7821166860839179e-07,
+      "loss": 0.1502,
+      "step": 8293
+    },
+    {
+      "epoch": 2.7697445316413427,
+      "grad_norm": 0.5237625902927323,
+      "learning_rate": 1.7769782266237767e-07,
+      "loss": 0.1508,
+      "step": 8294
+    },
+    {
+      "epoch": 2.770078477208215,
+      "grad_norm": 0.4762761343236909,
+      "learning_rate": 1.7718470518188645e-07,
+      "loss": 0.1441,
+      "step": 8295
+    },
+    {
+      "epoch": 2.7704124227750877,
+      "grad_norm": 0.5057082621318426,
+      "learning_rate": 1.7667231624443393e-07,
+      "loss": 0.1431,
+      "step": 8296
+    },
+    {
+      "epoch": 2.7707463683419604,
+      "grad_norm": 0.5366015119904906,
+      "learning_rate": 1.7616065592742038e-07,
+      "loss": 0.1633,
+      "step": 8297
+    },
+    {
+      "epoch": 2.7710803139088327,
+      "grad_norm": 0.5063382829156947,
+      "learning_rate": 1.7564972430813899e-07,
+      "loss": 0.1511,
+      "step": 8298
+    },
+    {
+      "epoch": 2.7714142594757054,
+      "grad_norm": 0.584640844229602,
+      "learning_rate": 1.751395214637708e-07,
+      "loss": 0.165,
+      "step": 8299
+    },
+    {
+      "epoch": 2.771748205042578,
+      "grad_norm": 0.51047234479896,
+      "learning_rate": 1.7463004747138967e-07,
+      "loss": 0.1502,
+      "step": 8300
+    },
+    {
+      "epoch": 2.7720821506094504,
+      "grad_norm": 0.5145726980495454,
+      "learning_rate": 1.7412130240795578e-07,
+      "loss": 0.1496,
+      "step": 8301
+    },
+    {
+      "epoch": 2.772416096176323,
+      "grad_norm": 0.5479476488710432,
+      "learning_rate": 1.736132863503226e-07,
+      "loss": 0.1591,
+      "step": 8302
+    },
+    {
+      "epoch": 2.772750041743196,
+      "grad_norm": 0.5223980288241024,
+      "learning_rate": 1.7310599937523153e-07,
+      "loss": 0.16,
+      "step": 8303
+    },
+    {
+      "epoch": 2.7730839873100686,
+      "grad_norm": 0.5203609186908762,
+      "learning_rate": 1.7259944155931407e-07,
+      "loss": 0.154,
+      "step": 8304
+    },
+    {
+      "epoch": 2.7734179328769413,
+      "grad_norm": 0.5262438153454855,
+      "learning_rate": 1.720936129790912e-07,
+      "loss": 0.1534,
+      "step": 8305
+    },
+    {
+      "epoch": 2.7737518784438135,
+      "grad_norm": 0.5572448507311574,
+      "learning_rate": 1.7158851371097518e-07,
+      "loss": 0.164,
+      "step": 8306
+    },
+    {
+      "epoch": 2.7740858240106863,
+      "grad_norm": 0.5516458484449425,
+      "learning_rate": 1.7108414383126658e-07,
+      "loss": 0.1575,
+      "step": 8307
+    },
+    {
+      "epoch": 2.774419769577559,
+      "grad_norm": 0.5305221309834752,
+      "learning_rate": 1.7058050341615783e-07,
+      "loss": 0.1504,
+      "step": 8308
+    },
+    {
+      "epoch": 2.7747537151444313,
+      "grad_norm": 0.5465339511703754,
+      "learning_rate": 1.7007759254172752e-07,
+      "loss": 0.1597,
+      "step": 8309
+    },
+    {
+      "epoch": 2.775087660711304,
+      "grad_norm": 0.5236311349194479,
+      "learning_rate": 1.6957541128394817e-07,
+      "loss": 0.151,
+      "step": 8310
+    },
+    {
+      "epoch": 2.7754216062781767,
+      "grad_norm": 0.5794917232299244,
+      "learning_rate": 1.6907395971867858e-07,
+      "loss": 0.165,
+      "step": 8311
+    },
+    {
+      "epoch": 2.7757555518450494,
+      "grad_norm": 0.5111090214492914,
+      "learning_rate": 1.685732379216698e-07,
+      "loss": 0.1537,
+      "step": 8312
+    },
+    {
+      "epoch": 2.776089497411922,
+      "grad_norm": 0.5182270741000585,
+      "learning_rate": 1.680732459685619e-07,
+      "loss": 0.1485,
+      "step": 8313
+    },
+    {
+      "epoch": 2.7764234429787944,
+      "grad_norm": 0.5299871536028107,
+      "learning_rate": 1.6757398393488443e-07,
+      "loss": 0.1615,
+      "step": 8314
+    },
+    {
+      "epoch": 2.776757388545667,
+      "grad_norm": 0.5229619474271693,
+      "learning_rate": 1.6707545189605657e-07,
+      "loss": 0.1549,
+      "step": 8315
+    },
+    {
+      "epoch": 2.77709133411254,
+      "grad_norm": 0.5312633993540525,
+      "learning_rate": 1.6657764992738746e-07,
+      "loss": 0.1508,
+      "step": 8316
+    },
+    {
+      "epoch": 2.777425279679412,
+      "grad_norm": 0.5207042176983809,
+      "learning_rate": 1.6608057810407586e-07,
+      "loss": 0.1548,
+      "step": 8317
+    },
+    {
+      "epoch": 2.777759225246285,
+      "grad_norm": 0.509134113266697,
+      "learning_rate": 1.6558423650121003e-07,
+      "loss": 0.149,
+      "step": 8318
+    },
+    {
+      "epoch": 2.7780931708131575,
+      "grad_norm": 0.5635185316038204,
+      "learning_rate": 1.6508862519376945e-07,
+      "loss": 0.166,
+      "step": 8319
+    },
+    {
+      "epoch": 2.77842711638003,
+      "grad_norm": 0.5039009942903678,
+      "learning_rate": 1.6459374425662088e-07,
+      "loss": 0.152,
+      "step": 8320
+    },
+    {
+      "epoch": 2.7787610619469025,
+      "grad_norm": 0.5185011394458492,
+      "learning_rate": 1.6409959376452289e-07,
+      "loss": 0.1497,
+      "step": 8321
+    },
+    {
+      "epoch": 2.7790950075137753,
+      "grad_norm": 0.5035741229676389,
+      "learning_rate": 1.6360617379212185e-07,
+      "loss": 0.147,
+      "step": 8322
+    },
+    {
+      "epoch": 2.779428953080648,
+      "grad_norm": 0.5234105413015845,
+      "learning_rate": 1.6311348441395535e-07,
+      "loss": 0.1525,
+      "step": 8323
+    },
+    {
+      "epoch": 2.7797628986475207,
+      "grad_norm": 0.5402213344741763,
+      "learning_rate": 1.6262152570444777e-07,
+      "loss": 0.1567,
+      "step": 8324
+    },
+    {
+      "epoch": 2.780096844214393,
+      "grad_norm": 0.500714119408257,
+      "learning_rate": 1.6213029773791912e-07,
+      "loss": 0.1503,
+      "step": 8325
+    },
+    {
+      "epoch": 2.7804307897812657,
+      "grad_norm": 0.5327017668640379,
+      "learning_rate": 1.6163980058857164e-07,
+      "loss": 0.159,
+      "step": 8326
+    },
+    {
+      "epoch": 2.7807647353481384,
+      "grad_norm": 0.5552130569413134,
+      "learning_rate": 1.6115003433050336e-07,
+      "loss": 0.1523,
+      "step": 8327
+    },
+    {
+      "epoch": 2.7810986809150107,
+      "grad_norm": 0.5036265248479774,
+      "learning_rate": 1.6066099903769726e-07,
+      "loss": 0.1443,
+      "step": 8328
+    },
+    {
+      "epoch": 2.7814326264818834,
+      "grad_norm": 0.5280869369504387,
+      "learning_rate": 1.6017269478402875e-07,
+      "loss": 0.1594,
+      "step": 8329
+    },
+    {
+      "epoch": 2.781766572048756,
+      "grad_norm": 0.5299118807830349,
+      "learning_rate": 1.59685121643261e-07,
+      "loss": 0.1493,
+      "step": 8330
+    },
+    {
+      "epoch": 2.7821005176156284,
+      "grad_norm": 0.5115595941536983,
+      "learning_rate": 1.5919827968904955e-07,
+      "loss": 0.1498,
+      "step": 8331
+    },
+    {
+      "epoch": 2.782434463182501,
+      "grad_norm": 0.5517706631282778,
+      "learning_rate": 1.5871216899493612e-07,
+      "loss": 0.1623,
+      "step": 8332
+    },
+    {
+      "epoch": 2.782768408749374,
+      "grad_norm": 0.5322940156215018,
+      "learning_rate": 1.5822678963435479e-07,
+      "loss": 0.1637,
+      "step": 8333
+    },
+    {
+      "epoch": 2.7831023543162465,
+      "grad_norm": 0.5279044532161087,
+      "learning_rate": 1.5774214168062575e-07,
+      "loss": 0.1544,
+      "step": 8334
+    },
+    {
+      "epoch": 2.7834362998831192,
+      "grad_norm": 0.5358377006180688,
+      "learning_rate": 1.5725822520696267e-07,
+      "loss": 0.159,
+      "step": 8335
+    },
+    {
+      "epoch": 2.7837702454499915,
+      "grad_norm": 0.5181269469955557,
+      "learning_rate": 1.567750402864654e-07,
+      "loss": 0.149,
+      "step": 8336
+    },
+    {
+      "epoch": 2.7841041910168642,
+      "grad_norm": 0.5627060721708065,
+      "learning_rate": 1.5629258699212613e-07,
+      "loss": 0.1598,
+      "step": 8337
+    },
+    {
+      "epoch": 2.784438136583737,
+      "grad_norm": 0.5672074227846221,
+      "learning_rate": 1.5581086539682433e-07,
+      "loss": 0.1647,
+      "step": 8338
+    },
+    {
+      "epoch": 2.7847720821506092,
+      "grad_norm": 0.5532121190782273,
+      "learning_rate": 1.5532987557332902e-07,
+      "loss": 0.1606,
+      "step": 8339
+    },
+    {
+      "epoch": 2.785106027717482,
+      "grad_norm": 0.5144356673510838,
+      "learning_rate": 1.5484961759430095e-07,
+      "loss": 0.1483,
+      "step": 8340
+    },
+    {
+      "epoch": 2.7854399732843547,
+      "grad_norm": 0.5476944503372652,
+      "learning_rate": 1.5437009153228766e-07,
+      "loss": 0.1581,
+      "step": 8341
+    },
+    {
+      "epoch": 2.7857739188512274,
+      "grad_norm": 0.5373962416791143,
+      "learning_rate": 1.538912974597273e-07,
+      "loss": 0.1578,
+      "step": 8342
+    },
+    {
+      "epoch": 2.7861078644181,
+      "grad_norm": 0.5169478563051959,
+      "learning_rate": 1.5341323544894758e-07,
+      "loss": 0.1531,
+      "step": 8343
+    },
+    {
+      "epoch": 2.7864418099849724,
+      "grad_norm": 0.5270574826065255,
+      "learning_rate": 1.5293590557216577e-07,
+      "loss": 0.1554,
+      "step": 8344
+    },
+    {
+      "epoch": 2.786775755551845,
+      "grad_norm": 0.5322542069948488,
+      "learning_rate": 1.5245930790148743e-07,
+      "loss": 0.1533,
+      "step": 8345
+    },
+    {
+      "epoch": 2.787109701118718,
+      "grad_norm": 0.5002626708460715,
+      "learning_rate": 1.5198344250890894e-07,
+      "loss": 0.1479,
+      "step": 8346
+    },
+    {
+      "epoch": 2.78744364668559,
+      "grad_norm": 0.510836318152733,
+      "learning_rate": 1.515083094663139e-07,
+      "loss": 0.1564,
+      "step": 8347
+    },
+    {
+      "epoch": 2.787777592252463,
+      "grad_norm": 0.5024767543651935,
+      "learning_rate": 1.5103390884547931e-07,
+      "loss": 0.1496,
+      "step": 8348
+    },
+    {
+      "epoch": 2.7881115378193355,
+      "grad_norm": 0.5256020147012425,
+      "learning_rate": 1.5056024071806674e-07,
+      "loss": 0.156,
+      "step": 8349
+    },
+    {
+      "epoch": 2.788445483386208,
+      "grad_norm": 0.5513860801672124,
+      "learning_rate": 1.5008730515563064e-07,
+      "loss": 0.1702,
+      "step": 8350
+    },
+    {
+      "epoch": 2.7887794289530805,
+      "grad_norm": 0.5583302740203661,
+      "learning_rate": 1.4961510222961216e-07,
+      "loss": 0.1689,
+      "step": 8351
+    },
+    {
+      "epoch": 2.7891133745199532,
+      "grad_norm": 0.5442873203295383,
+      "learning_rate": 1.4914363201134486e-07,
+      "loss": 0.1543,
+      "step": 8352
+    },
+    {
+      "epoch": 2.789447320086826,
+      "grad_norm": 0.5524028432461694,
+      "learning_rate": 1.4867289457204726e-07,
+      "loss": 0.1726,
+      "step": 8353
+    },
+    {
+      "epoch": 2.7897812656536987,
+      "grad_norm": 0.530554499586277,
+      "learning_rate": 1.4820288998283304e-07,
+      "loss": 0.1517,
+      "step": 8354
+    },
+    {
+      "epoch": 2.790115211220571,
+      "grad_norm": 0.556591191428861,
+      "learning_rate": 1.477336183146999e-07,
+      "loss": 0.1557,
+      "step": 8355
+    },
+    {
+      "epoch": 2.7904491567874437,
+      "grad_norm": 0.544705956885302,
+      "learning_rate": 1.4726507963853776e-07,
+      "loss": 0.1627,
+      "step": 8356
+    },
+    {
+      "epoch": 2.7907831023543164,
+      "grad_norm": 0.5220600913224868,
+      "learning_rate": 1.4679727402512334e-07,
+      "loss": 0.1527,
+      "step": 8357
+    },
+    {
+      "epoch": 2.7911170479211886,
+      "grad_norm": 0.52810589422968,
+      "learning_rate": 1.4633020154512677e-07,
+      "loss": 0.1551,
+      "step": 8358
+    },
+    {
+      "epoch": 2.7914509934880614,
+      "grad_norm": 0.4970005355575271,
+      "learning_rate": 1.458638622691022e-07,
+      "loss": 0.1479,
+      "step": 8359
+    },
+    {
+      "epoch": 2.791784939054934,
+      "grad_norm": 0.5212873276869456,
+      "learning_rate": 1.4539825626749715e-07,
+      "loss": 0.1541,
+      "step": 8360
+    },
+    {
+      "epoch": 2.792118884621807,
+      "grad_norm": 0.5042382836183906,
+      "learning_rate": 1.4493338361064646e-07,
+      "loss": 0.1455,
+      "step": 8361
+    },
+    {
+      "epoch": 2.7924528301886795,
+      "grad_norm": 0.5309456541518355,
+      "learning_rate": 1.4446924436877507e-07,
+      "loss": 0.1609,
+      "step": 8362
+    },
+    {
+      "epoch": 2.792786775755552,
+      "grad_norm": 0.5118129734037632,
+      "learning_rate": 1.4400583861199636e-07,
+      "loss": 0.1541,
+      "step": 8363
+    },
+    {
+      "epoch": 2.7931207213224245,
+      "grad_norm": 0.4853226783481245,
+      "learning_rate": 1.4354316641031263e-07,
+      "loss": 0.1543,
+      "step": 8364
+    },
+    {
+      "epoch": 2.7934546668892972,
+      "grad_norm": 0.541481071236972,
+      "learning_rate": 1.4308122783361688e-07,
+      "loss": 0.1523,
+      "step": 8365
+    },
+    {
+      "epoch": 2.7937886124561695,
+      "grad_norm": 0.5278220146033721,
+      "learning_rate": 1.4262002295168997e-07,
+      "loss": 0.1547,
+      "step": 8366
+    },
+    {
+      "epoch": 2.794122558023042,
+      "grad_norm": 0.5135966918537538,
+      "learning_rate": 1.4215955183420282e-07,
+      "loss": 0.1502,
+      "step": 8367
+    },
+    {
+      "epoch": 2.794456503589915,
+      "grad_norm": 0.4920830612333697,
+      "learning_rate": 1.4169981455071368e-07,
+      "loss": 0.1571,
+      "step": 8368
+    },
+    {
+      "epoch": 2.794790449156787,
+      "grad_norm": 0.5148476681687206,
+      "learning_rate": 1.4124081117067313e-07,
+      "loss": 0.1553,
+      "step": 8369
+    },
+    {
+      "epoch": 2.79512439472366,
+      "grad_norm": 0.5234172849783939,
+      "learning_rate": 1.4078254176341788e-07,
+      "loss": 0.1583,
+      "step": 8370
+    },
+    {
+      "epoch": 2.7954583402905326,
+      "grad_norm": 0.5072703583829362,
+      "learning_rate": 1.4032500639817426e-07,
+      "loss": 0.154,
+      "step": 8371
+    },
+    {
+      "epoch": 2.7957922858574054,
+      "grad_norm": 0.5591966083853513,
+      "learning_rate": 1.3986820514405973e-07,
+      "loss": 0.1635,
+      "step": 8372
+    },
+    {
+      "epoch": 2.796126231424278,
+      "grad_norm": 0.5831281500644744,
+      "learning_rate": 1.394121380700797e-07,
+      "loss": 0.1666,
+      "step": 8373
+    },
+    {
+      "epoch": 2.7964601769911503,
+      "grad_norm": 0.5324043094043531,
+      "learning_rate": 1.3895680524512734e-07,
+      "loss": 0.1529,
+      "step": 8374
+    },
+    {
+      "epoch": 2.796794122558023,
+      "grad_norm": 0.5407819919549826,
+      "learning_rate": 1.3850220673798655e-07,
+      "loss": 0.1523,
+      "step": 8375
+    },
+    {
+      "epoch": 2.797128068124896,
+      "grad_norm": 0.571743758557889,
+      "learning_rate": 1.3804834261732957e-07,
+      "loss": 0.1604,
+      "step": 8376
+    },
+    {
+      "epoch": 2.797462013691768,
+      "grad_norm": 0.5261550026856082,
+      "learning_rate": 1.3759521295171773e-07,
+      "loss": 0.1579,
+      "step": 8377
+    },
+    {
+      "epoch": 2.7977959592586408,
+      "grad_norm": 0.5077553496486351,
+      "learning_rate": 1.3714281780960237e-07,
+      "loss": 0.1549,
+      "step": 8378
+    },
+    {
+      "epoch": 2.7981299048255135,
+      "grad_norm": 0.5512498148977449,
+      "learning_rate": 1.366911572593227e-07,
+      "loss": 0.1462,
+      "step": 8379
+    },
+    {
+      "epoch": 2.7984638503923858,
+      "grad_norm": 0.5278418783066358,
+      "learning_rate": 1.3624023136910691e-07,
+      "loss": 0.1527,
+      "step": 8380
+    },
+    {
+      "epoch": 2.7987977959592585,
+      "grad_norm": 0.5720969297382702,
+      "learning_rate": 1.3579004020707387e-07,
+      "loss": 0.1616,
+      "step": 8381
+    },
+    {
+      "epoch": 2.799131741526131,
+      "grad_norm": 0.5391404210914308,
+      "learning_rate": 1.3534058384122862e-07,
+      "loss": 0.1588,
+      "step": 8382
+    },
+    {
+      "epoch": 2.799465687093004,
+      "grad_norm": 0.5391505004911138,
+      "learning_rate": 1.3489186233946793e-07,
+      "loss": 0.1586,
+      "step": 8383
+    },
+    {
+      "epoch": 2.7997996326598766,
+      "grad_norm": 0.5405451852070394,
+      "learning_rate": 1.3444387576957706e-07,
+      "loss": 0.162,
+      "step": 8384
+    },
+    {
+      "epoch": 2.800133578226749,
+      "grad_norm": 0.5523476563053142,
+      "learning_rate": 1.33996624199228e-07,
+      "loss": 0.1588,
+      "step": 8385
+    },
+    {
+      "epoch": 2.8004675237936216,
+      "grad_norm": 0.520024555301227,
+      "learning_rate": 1.335501076959844e-07,
+      "loss": 0.1544,
+      "step": 8386
+    },
+    {
+      "epoch": 2.8008014693604943,
+      "grad_norm": 0.5772233450893358,
+      "learning_rate": 1.331043263272974e-07,
+      "loss": 0.1648,
+      "step": 8387
+    },
+    {
+      "epoch": 2.8011354149273666,
+      "grad_norm": 0.5240091069680041,
+      "learning_rate": 1.3265928016050756e-07,
+      "loss": 0.1468,
+      "step": 8388
+    },
+    {
+      "epoch": 2.8014693604942393,
+      "grad_norm": 0.5134959826858604,
+      "learning_rate": 1.3221496926284493e-07,
+      "loss": 0.1523,
+      "step": 8389
+    },
+    {
+      "epoch": 2.801803306061112,
+      "grad_norm": 0.53411701119722,
+      "learning_rate": 1.3177139370142755e-07,
+      "loss": 0.1533,
+      "step": 8390
+    },
+    {
+      "epoch": 2.8021372516279848,
+      "grad_norm": 0.5451994225507716,
+      "learning_rate": 1.3132855354326236e-07,
+      "loss": 0.1565,
+      "step": 8391
+    },
+    {
+      "epoch": 2.8024711971948575,
+      "grad_norm": 0.4988335998137933,
+      "learning_rate": 1.3088644885524637e-07,
+      "loss": 0.1479,
+      "step": 8392
+    },
+    {
+      "epoch": 2.8028051427617298,
+      "grad_norm": 0.5470946888485676,
+      "learning_rate": 1.3044507970416398e-07,
+      "loss": 0.1557,
+      "step": 8393
+    },
+    {
+      "epoch": 2.8031390883286025,
+      "grad_norm": 0.5434547017527317,
+      "learning_rate": 1.3000444615668906e-07,
+      "loss": 0.1557,
+      "step": 8394
+    },
+    {
+      "epoch": 2.803473033895475,
+      "grad_norm": 0.5210177368074225,
+      "learning_rate": 1.2956454827938557e-07,
+      "loss": 0.1471,
+      "step": 8395
+    },
+    {
+      "epoch": 2.8038069794623475,
+      "grad_norm": 0.506645320292903,
+      "learning_rate": 1.291253861387043e-07,
+      "loss": 0.1385,
+      "step": 8396
+    },
+    {
+      "epoch": 2.80414092502922,
+      "grad_norm": 0.5153043838572172,
+      "learning_rate": 1.28686959800986e-07,
+      "loss": 0.1547,
+      "step": 8397
+    },
+    {
+      "epoch": 2.804474870596093,
+      "grad_norm": 0.5433374471677337,
+      "learning_rate": 1.2824926933246106e-07,
+      "loss": 0.1577,
+      "step": 8398
+    },
+    {
+      "epoch": 2.804808816162965,
+      "grad_norm": 0.5741759511423757,
+      "learning_rate": 1.2781231479924606e-07,
+      "loss": 0.1671,
+      "step": 8399
+    },
+    {
+      "epoch": 2.805142761729838,
+      "grad_norm": 0.5348545400171368,
+      "learning_rate": 1.2737609626734927e-07,
+      "loss": 0.1571,
+      "step": 8400
+    },
+    {
+      "epoch": 2.8054767072967106,
+      "grad_norm": 0.5347605161338985,
+      "learning_rate": 1.269406138026663e-07,
+      "loss": 0.1603,
+      "step": 8401
+    },
+    {
+      "epoch": 2.8058106528635833,
+      "grad_norm": 0.5372346879839113,
+      "learning_rate": 1.2650586747098238e-07,
+      "loss": 0.1609,
+      "step": 8402
+    },
+    {
+      "epoch": 2.806144598430456,
+      "grad_norm": 0.5184156394531372,
+      "learning_rate": 1.2607185733797044e-07,
+      "loss": 0.1531,
+      "step": 8403
+    },
+    {
+      "epoch": 2.8064785439973283,
+      "grad_norm": 0.5193297902452196,
+      "learning_rate": 1.2563858346919365e-07,
+      "loss": 0.1495,
+      "step": 8404
+    },
+    {
+      "epoch": 2.806812489564201,
+      "grad_norm": 0.4735671716020055,
+      "learning_rate": 1.2520604593010189e-07,
+      "loss": 0.139,
+      "step": 8405
+    },
+    {
+      "epoch": 2.8071464351310738,
+      "grad_norm": 0.5452990496424134,
+      "learning_rate": 1.247742447860356e-07,
+      "loss": 0.1518,
+      "step": 8406
+    },
+    {
+      "epoch": 2.807480380697946,
+      "grad_norm": 0.5197495861568605,
+      "learning_rate": 1.2434318010222434e-07,
+      "loss": 0.1531,
+      "step": 8407
+    },
+    {
+      "epoch": 2.8078143262648187,
+      "grad_norm": 0.5682800571844647,
+      "learning_rate": 1.2391285194378433e-07,
+      "loss": 0.1558,
+      "step": 8408
+    },
+    {
+      "epoch": 2.8081482718316915,
+      "grad_norm": 0.5476986226433352,
+      "learning_rate": 1.2348326037572244e-07,
+      "loss": 0.1642,
+      "step": 8409
+    },
+    {
+      "epoch": 2.808482217398564,
+      "grad_norm": 0.5660502282746109,
+      "learning_rate": 1.2305440546293236e-07,
+      "loss": 0.1727,
+      "step": 8410
+    },
+    {
+      "epoch": 2.808816162965437,
+      "grad_norm": 0.4987505875655572,
+      "learning_rate": 1.2262628727019942e-07,
+      "loss": 0.1414,
+      "step": 8411
+    },
+    {
+      "epoch": 2.809150108532309,
+      "grad_norm": 0.5209194368951366,
+      "learning_rate": 1.221989058621942e-07,
+      "loss": 0.1512,
+      "step": 8412
+    },
+    {
+      "epoch": 2.809484054099182,
+      "grad_norm": 0.5773338493631209,
+      "learning_rate": 1.2177226130347886e-07,
+      "loss": 0.1533,
+      "step": 8413
+    },
+    {
+      "epoch": 2.8098179996660546,
+      "grad_norm": 0.49229443873769146,
+      "learning_rate": 1.21346353658503e-07,
+      "loss": 0.1488,
+      "step": 8414
+    },
+    {
+      "epoch": 2.810151945232927,
+      "grad_norm": 0.5239314850265931,
+      "learning_rate": 1.209211829916046e-07,
+      "loss": 0.1632,
+      "step": 8415
+    },
+    {
+      "epoch": 2.8104858907997996,
+      "grad_norm": 0.5733286509903617,
+      "learning_rate": 1.204967493670106e-07,
+      "loss": 0.1632,
+      "step": 8416
+    },
+    {
+      "epoch": 2.8108198363666723,
+      "grad_norm": 0.5649617767440496,
+      "learning_rate": 1.2007305284883696e-07,
+      "loss": 0.1524,
+      "step": 8417
+    },
+    {
+      "epoch": 2.8111537819335446,
+      "grad_norm": 0.5495402597096266,
+      "learning_rate": 1.1965009350108747e-07,
+      "loss": 0.1856,
+      "step": 8418
+    },
+    {
+      "epoch": 2.8114877275004173,
+      "grad_norm": 0.4862318982701236,
+      "learning_rate": 1.1922787138765656e-07,
+      "loss": 0.1439,
+      "step": 8419
+    },
+    {
+      "epoch": 2.81182167306729,
+      "grad_norm": 0.5452192079690045,
+      "learning_rate": 1.188063865723238e-07,
+      "loss": 0.1584,
+      "step": 8420
+    },
+    {
+      "epoch": 2.8121556186341627,
+      "grad_norm": 0.5197784434274771,
+      "learning_rate": 1.1838563911876155e-07,
+      "loss": 0.1568,
+      "step": 8421
+    },
+    {
+      "epoch": 2.8124895642010355,
+      "grad_norm": 0.5583439489780697,
+      "learning_rate": 1.1796562909052734e-07,
+      "loss": 0.1571,
+      "step": 8422
+    },
+    {
+      "epoch": 2.8128235097679077,
+      "grad_norm": 0.5495414718376356,
+      "learning_rate": 1.1754635655106928e-07,
+      "loss": 0.1638,
+      "step": 8423
+    },
+    {
+      "epoch": 2.8131574553347805,
+      "grad_norm": 0.5353757303775786,
+      "learning_rate": 1.1712782156372226e-07,
+      "loss": 0.1556,
+      "step": 8424
+    },
+    {
+      "epoch": 2.813491400901653,
+      "grad_norm": 0.5316180867730131,
+      "learning_rate": 1.167100241917124e-07,
+      "loss": 0.1542,
+      "step": 8425
+    },
+    {
+      "epoch": 2.8138253464685254,
+      "grad_norm": 0.5410694261321864,
+      "learning_rate": 1.1629296449815197e-07,
+      "loss": 0.1539,
+      "step": 8426
+    },
+    {
+      "epoch": 2.814159292035398,
+      "grad_norm": 0.5255062069267001,
+      "learning_rate": 1.1587664254604336e-07,
+      "loss": 0.1558,
+      "step": 8427
+    },
+    {
+      "epoch": 2.814493237602271,
+      "grad_norm": 0.5564039864164055,
+      "learning_rate": 1.1546105839827626e-07,
+      "loss": 0.1642,
+      "step": 8428
+    },
+    {
+      "epoch": 2.814827183169143,
+      "grad_norm": 0.5451419332138157,
+      "learning_rate": 1.150462121176299e-07,
+      "loss": 0.1605,
+      "step": 8429
+    },
+    {
+      "epoch": 2.815161128736016,
+      "grad_norm": 0.5608596813939348,
+      "learning_rate": 1.1463210376677192e-07,
+      "loss": 0.1678,
+      "step": 8430
+    },
+    {
+      "epoch": 2.8154950743028886,
+      "grad_norm": 0.5426405423185976,
+      "learning_rate": 1.1421873340825729e-07,
+      "loss": 0.1549,
+      "step": 8431
+    },
+    {
+      "epoch": 2.8158290198697613,
+      "grad_norm": 0.5354590434247339,
+      "learning_rate": 1.1380610110453217e-07,
+      "loss": 0.1545,
+      "step": 8432
+    },
+    {
+      "epoch": 2.816162965436634,
+      "grad_norm": 0.5214195383604076,
+      "learning_rate": 1.133942069179278e-07,
+      "loss": 0.1524,
+      "step": 8433
+    },
+    {
+      "epoch": 2.8164969110035063,
+      "grad_norm": 0.5846181741316001,
+      "learning_rate": 1.1298305091066664e-07,
+      "loss": 0.1571,
+      "step": 8434
+    },
+    {
+      "epoch": 2.816830856570379,
+      "grad_norm": 0.5684128730306873,
+      "learning_rate": 1.1257263314485844e-07,
+      "loss": 0.1588,
+      "step": 8435
+    },
+    {
+      "epoch": 2.8171648021372517,
+      "grad_norm": 0.5148921793629571,
+      "learning_rate": 1.1216295368250196e-07,
+      "loss": 0.1561,
+      "step": 8436
+    },
+    {
+      "epoch": 2.817498747704124,
+      "grad_norm": 0.5536026332038109,
+      "learning_rate": 1.1175401258548324e-07,
+      "loss": 0.1539,
+      "step": 8437
+    },
+    {
+      "epoch": 2.8178326932709967,
+      "grad_norm": 0.5272870234520881,
+      "learning_rate": 1.1134580991557842e-07,
+      "loss": 0.1623,
+      "step": 8438
+    },
+    {
+      "epoch": 2.8181666388378694,
+      "grad_norm": 0.49663883935703573,
+      "learning_rate": 1.1093834573445094e-07,
+      "loss": 0.1494,
+      "step": 8439
+    },
+    {
+      "epoch": 2.818500584404742,
+      "grad_norm": 0.5741127853098581,
+      "learning_rate": 1.1053162010365326e-07,
+      "loss": 0.1507,
+      "step": 8440
+    },
+    {
+      "epoch": 2.818834529971615,
+      "grad_norm": 0.5546891679590489,
+      "learning_rate": 1.1012563308462565e-07,
+      "loss": 0.1574,
+      "step": 8441
+    },
+    {
+      "epoch": 2.819168475538487,
+      "grad_norm": 0.5037622547843496,
+      "learning_rate": 1.0972038473869795e-07,
+      "loss": 0.151,
+      "step": 8442
+    },
+    {
+      "epoch": 2.81950242110536,
+      "grad_norm": 0.5486643889881713,
+      "learning_rate": 1.093158751270873e-07,
+      "loss": 0.1506,
+      "step": 8443
+    },
+    {
+      "epoch": 2.8198363666722326,
+      "grad_norm": 0.5590759169843443,
+      "learning_rate": 1.0891210431089983e-07,
+      "loss": 0.1634,
+      "step": 8444
+    },
+    {
+      "epoch": 2.820170312239105,
+      "grad_norm": 0.5382380812540658,
+      "learning_rate": 1.0850907235112895e-07,
+      "loss": 0.1579,
+      "step": 8445
+    },
+    {
+      "epoch": 2.8205042578059776,
+      "grad_norm": 0.5253716527065034,
+      "learning_rate": 1.0810677930865876e-07,
+      "loss": 0.1521,
+      "step": 8446
+    },
+    {
+      "epoch": 2.8208382033728503,
+      "grad_norm": 0.48555707327772146,
+      "learning_rate": 1.0770522524425898e-07,
+      "loss": 0.1417,
+      "step": 8447
+    },
+    {
+      "epoch": 2.8211721489397226,
+      "grad_norm": 0.5528063179335293,
+      "learning_rate": 1.0730441021859106e-07,
+      "loss": 0.1491,
+      "step": 8448
+    },
+    {
+      "epoch": 2.8215060945065953,
+      "grad_norm": 0.5496044182417962,
+      "learning_rate": 1.0690433429220049e-07,
+      "loss": 0.1506,
+      "step": 8449
+    },
+    {
+      "epoch": 2.821840040073468,
+      "grad_norm": 0.5346500000652207,
+      "learning_rate": 1.0650499752552557e-07,
+      "loss": 0.1539,
+      "step": 8450
+    },
+    {
+      "epoch": 2.8221739856403407,
+      "grad_norm": 0.5479349408608167,
+      "learning_rate": 1.0610639997888917e-07,
+      "loss": 0.1557,
+      "step": 8451
+    },
+    {
+      "epoch": 2.8225079312072134,
+      "grad_norm": 0.46744982444128813,
+      "learning_rate": 1.0570854171250478e-07,
+      "loss": 0.1374,
+      "step": 8452
+    },
+    {
+      "epoch": 2.8228418767740857,
+      "grad_norm": 0.5441636040790427,
+      "learning_rate": 1.0531142278647378e-07,
+      "loss": 0.153,
+      "step": 8453
+    },
+    {
+      "epoch": 2.8231758223409584,
+      "grad_norm": 0.5717208544928984,
+      "learning_rate": 1.0491504326078483e-07,
+      "loss": 0.1606,
+      "step": 8454
+    },
+    {
+      "epoch": 2.823509767907831,
+      "grad_norm": 0.5434976551693014,
+      "learning_rate": 1.0451940319531728e-07,
+      "loss": 0.1495,
+      "step": 8455
+    },
+    {
+      "epoch": 2.8238437134747034,
+      "grad_norm": 0.5563013385219229,
+      "learning_rate": 1.0412450264983609e-07,
+      "loss": 0.1582,
+      "step": 8456
+    },
+    {
+      "epoch": 2.824177659041576,
+      "grad_norm": 0.5133006701112091,
+      "learning_rate": 1.0373034168399521e-07,
+      "loss": 0.1468,
+      "step": 8457
+    },
+    {
+      "epoch": 2.824511604608449,
+      "grad_norm": 0.5368851278335857,
+      "learning_rate": 1.0333692035733867e-07,
+      "loss": 0.1577,
+      "step": 8458
+    },
+    {
+      "epoch": 2.8248455501753216,
+      "grad_norm": 0.5658419408433192,
+      "learning_rate": 1.0294423872929615e-07,
+      "loss": 0.1458,
+      "step": 8459
+    },
+    {
+      "epoch": 2.8251794957421943,
+      "grad_norm": 0.565524948708705,
+      "learning_rate": 1.0255229685918744e-07,
+      "loss": 0.1694,
+      "step": 8460
+    },
+    {
+      "epoch": 2.8255134413090666,
+      "grad_norm": 0.5525057994914118,
+      "learning_rate": 1.0216109480622017e-07,
+      "loss": 0.1674,
+      "step": 8461
+    },
+    {
+      "epoch": 2.8258473868759393,
+      "grad_norm": 0.7857522622361894,
+      "learning_rate": 1.0177063262948927e-07,
+      "loss": 0.1641,
+      "step": 8462
+    },
+    {
+      "epoch": 2.826181332442812,
+      "grad_norm": 0.5695931540924961,
+      "learning_rate": 1.0138091038797982e-07,
+      "loss": 0.1592,
+      "step": 8463
+    },
+    {
+      "epoch": 2.8265152780096843,
+      "grad_norm": 0.5376746632499131,
+      "learning_rate": 1.0099192814056247e-07,
+      "loss": 0.1663,
+      "step": 8464
+    },
+    {
+      "epoch": 2.826849223576557,
+      "grad_norm": 0.4698801948441204,
+      "learning_rate": 1.0060368594599856e-07,
+      "loss": 0.136,
+      "step": 8465
+    },
+    {
+      "epoch": 2.8271831691434297,
+      "grad_norm": 0.544981428758501,
+      "learning_rate": 1.002161838629362e-07,
+      "loss": 0.1525,
+      "step": 8466
+    },
+    {
+      "epoch": 2.827517114710302,
+      "grad_norm": 0.5350947441046976,
+      "learning_rate": 9.982942194991297e-08,
+      "loss": 0.1503,
+      "step": 8467
+    },
+    {
+      "epoch": 2.8278510602771747,
+      "grad_norm": 0.5659243272345094,
+      "learning_rate": 9.94434002653527e-08,
+      "loss": 0.1621,
+      "step": 8468
+    },
+    {
+      "epoch": 2.8281850058440474,
+      "grad_norm": 0.4893884162320132,
+      "learning_rate": 9.905811886756933e-08,
+      "loss": 0.1444,
+      "step": 8469
+    },
+    {
+      "epoch": 2.82851895141092,
+      "grad_norm": 0.5115181229873957,
+      "learning_rate": 9.867357781476294e-08,
+      "loss": 0.1536,
+      "step": 8470
+    },
+    {
+      "epoch": 2.828852896977793,
+      "grad_norm": 0.5310645008445086,
+      "learning_rate": 9.828977716502486e-08,
+      "loss": 0.1527,
+      "step": 8471
+    },
+    {
+      "epoch": 2.829186842544665,
+      "grad_norm": 0.5390370539150535,
+      "learning_rate": 9.790671697633092e-08,
+      "loss": 0.1607,
+      "step": 8472
+    },
+    {
+      "epoch": 2.829520788111538,
+      "grad_norm": 0.5542118402025401,
+      "learning_rate": 9.752439730654872e-08,
+      "loss": 0.1641,
+      "step": 8473
+    },
+    {
+      "epoch": 2.8298547336784106,
+      "grad_norm": 0.5423612496980644,
+      "learning_rate": 9.714281821343041e-08,
+      "loss": 0.1626,
+      "step": 8474
+    },
+    {
+      "epoch": 2.830188679245283,
+      "grad_norm": 0.5342496957180441,
+      "learning_rate": 9.676197975461876e-08,
+      "loss": 0.1582,
+      "step": 8475
+    },
+    {
+      "epoch": 2.8305226248121556,
+      "grad_norm": 0.49551719740241496,
+      "learning_rate": 9.638188198764387e-08,
+      "loss": 0.1514,
+      "step": 8476
+    },
+    {
+      "epoch": 2.8308565703790283,
+      "grad_norm": 0.514845882985583,
+      "learning_rate": 9.600252496992369e-08,
+      "loss": 0.1455,
+      "step": 8477
+    },
+    {
+      "epoch": 2.8311905159459005,
+      "grad_norm": 0.5246479501288792,
+      "learning_rate": 9.562390875876515e-08,
+      "loss": 0.1571,
+      "step": 8478
+    },
+    {
+      "epoch": 2.8315244615127733,
+      "grad_norm": 0.5151252059582621,
+      "learning_rate": 9.524603341136251e-08,
+      "loss": 0.15,
+      "step": 8479
+    },
+    {
+      "epoch": 2.831858407079646,
+      "grad_norm": 0.5659310966269068,
+      "learning_rate": 9.486889898479734e-08,
+      "loss": 0.156,
+      "step": 8480
+    },
+    {
+      "epoch": 2.8321923526465187,
+      "grad_norm": 0.5146682132991759,
+      "learning_rate": 9.449250553604184e-08,
+      "loss": 0.1505,
+      "step": 8481
+    },
+    {
+      "epoch": 2.8325262982133914,
+      "grad_norm": 0.4889795191225081,
+      "learning_rate": 9.41168531219533e-08,
+      "loss": 0.1508,
+      "step": 8482
+    },
+    {
+      "epoch": 2.8328602437802637,
+      "grad_norm": 0.56495527958627,
+      "learning_rate": 9.374194179927909e-08,
+      "loss": 0.1577,
+      "step": 8483
+    },
+    {
+      "epoch": 2.8331941893471364,
+      "grad_norm": 0.5605532903002794,
+      "learning_rate": 9.336777162465449e-08,
+      "loss": 0.1633,
+      "step": 8484
+    },
+    {
+      "epoch": 2.833528134914009,
+      "grad_norm": 0.4893798584627006,
+      "learning_rate": 9.299434265460095e-08,
+      "loss": 0.1394,
+      "step": 8485
+    },
+    {
+      "epoch": 2.8338620804808814,
+      "grad_norm": 0.5367020601261826,
+      "learning_rate": 9.262165494553055e-08,
+      "loss": 0.1521,
+      "step": 8486
+    },
+    {
+      "epoch": 2.834196026047754,
+      "grad_norm": 0.5127849134301988,
+      "learning_rate": 9.22497085537416e-08,
+      "loss": 0.1504,
+      "step": 8487
+    },
+    {
+      "epoch": 2.834529971614627,
+      "grad_norm": 0.5615166401386916,
+      "learning_rate": 9.187850353542082e-08,
+      "loss": 0.1581,
+      "step": 8488
+    },
+    {
+      "epoch": 2.8348639171814995,
+      "grad_norm": 0.5169398545404523,
+      "learning_rate": 9.150803994664337e-08,
+      "loss": 0.1505,
+      "step": 8489
+    },
+    {
+      "epoch": 2.8351978627483723,
+      "grad_norm": 0.5682736343819038,
+      "learning_rate": 9.113831784337279e-08,
+      "loss": 0.1531,
+      "step": 8490
+    },
+    {
+      "epoch": 2.8355318083152445,
+      "grad_norm": 0.5340266016385549,
+      "learning_rate": 9.076933728145832e-08,
+      "loss": 0.151,
+      "step": 8491
+    },
+    {
+      "epoch": 2.8358657538821173,
+      "grad_norm": 0.5585134485577232,
+      "learning_rate": 9.040109831664035e-08,
+      "loss": 0.1619,
+      "step": 8492
+    },
+    {
+      "epoch": 2.83619969944899,
+      "grad_norm": 0.5725765003377334,
+      "learning_rate": 9.003360100454495e-08,
+      "loss": 0.1592,
+      "step": 8493
+    },
+    {
+      "epoch": 2.8365336450158622,
+      "grad_norm": 0.534273889129343,
+      "learning_rate": 8.966684540068659e-08,
+      "loss": 0.1553,
+      "step": 8494
+    },
+    {
+      "epoch": 2.836867590582735,
+      "grad_norm": 0.5171157368775522,
+      "learning_rate": 8.930083156046931e-08,
+      "loss": 0.1517,
+      "step": 8495
+    },
+    {
+      "epoch": 2.8372015361496077,
+      "grad_norm": 0.5634182648212217,
+      "learning_rate": 8.893555953918276e-08,
+      "loss": 0.1567,
+      "step": 8496
+    },
+    {
+      "epoch": 2.83753548171648,
+      "grad_norm": 0.5264929981389637,
+      "learning_rate": 8.857102939200557e-08,
+      "loss": 0.1459,
+      "step": 8497
+    },
+    {
+      "epoch": 2.8378694272833527,
+      "grad_norm": 0.5735554438040473,
+      "learning_rate": 8.820724117400536e-08,
+      "loss": 0.1652,
+      "step": 8498
+    },
+    {
+      "epoch": 2.8382033728502254,
+      "grad_norm": 0.5039614742487958,
+      "learning_rate": 8.784419494013541e-08,
+      "loss": 0.1444,
+      "step": 8499
+    },
+    {
+      "epoch": 2.838537318417098,
+      "grad_norm": 0.5278294997838044,
+      "learning_rate": 8.74818907452385e-08,
+      "loss": 0.1614,
+      "step": 8500
+    },
+    {
+      "epoch": 2.838871263983971,
+      "grad_norm": 0.510269423877444,
+      "learning_rate": 8.712032864404529e-08,
+      "loss": 0.1497,
+      "step": 8501
+    },
+    {
+      "epoch": 2.839205209550843,
+      "grad_norm": 0.5285759359254864,
+      "learning_rate": 8.675950869117323e-08,
+      "loss": 0.1573,
+      "step": 8502
+    },
+    {
+      "epoch": 2.839539155117716,
+      "grad_norm": 0.5220521008233078,
+      "learning_rate": 8.639943094112868e-08,
+      "loss": 0.1458,
+      "step": 8503
+    },
+    {
+      "epoch": 2.8398731006845885,
+      "grad_norm": 0.5595841029538259,
+      "learning_rate": 8.604009544830705e-08,
+      "loss": 0.1601,
+      "step": 8504
+    },
+    {
+      "epoch": 2.840207046251461,
+      "grad_norm": 0.5506637831429321,
+      "learning_rate": 8.568150226698823e-08,
+      "loss": 0.1599,
+      "step": 8505
+    },
+    {
+      "epoch": 2.8405409918183335,
+      "grad_norm": 0.5610602819838691,
+      "learning_rate": 8.532365145134226e-08,
+      "loss": 0.1546,
+      "step": 8506
+    },
+    {
+      "epoch": 2.8408749373852062,
+      "grad_norm": 0.5818644312959905,
+      "learning_rate": 8.496654305542807e-08,
+      "loss": 0.1728,
+      "step": 8507
+    },
+    {
+      "epoch": 2.841208882952079,
+      "grad_norm": 0.5776511122322144,
+      "learning_rate": 8.461017713318976e-08,
+      "loss": 0.1654,
+      "step": 8508
+    },
+    {
+      "epoch": 2.8415428285189517,
+      "grad_norm": 0.5189470178377121,
+      "learning_rate": 8.425455373846147e-08,
+      "loss": 0.1497,
+      "step": 8509
+    },
+    {
+      "epoch": 2.841876774085824,
+      "grad_norm": 0.4788685040909344,
+      "learning_rate": 8.38996729249636e-08,
+      "loss": 0.1392,
+      "step": 8510
+    },
+    {
+      "epoch": 2.8422107196526967,
+      "grad_norm": 0.5386980557967677,
+      "learning_rate": 8.354553474630489e-08,
+      "loss": 0.1554,
+      "step": 8511
+    },
+    {
+      "epoch": 2.8425446652195694,
+      "grad_norm": 0.5621709868208091,
+      "learning_rate": 8.319213925598258e-08,
+      "loss": 0.1595,
+      "step": 8512
+    },
+    {
+      "epoch": 2.8428786107864417,
+      "grad_norm": 0.5321301128110231,
+      "learning_rate": 8.283948650738172e-08,
+      "loss": 0.147,
+      "step": 8513
+    },
+    {
+      "epoch": 2.8432125563533144,
+      "grad_norm": 0.521599077794686,
+      "learning_rate": 8.248757655377415e-08,
+      "loss": 0.1559,
+      "step": 8514
+    },
+    {
+      "epoch": 2.843546501920187,
+      "grad_norm": 0.5471021279762034,
+      "learning_rate": 8.213640944831957e-08,
+      "loss": 0.1635,
+      "step": 8515
+    },
+    {
+      "epoch": 2.8438804474870594,
+      "grad_norm": 0.5022501808867641,
+      "learning_rate": 8.178598524406667e-08,
+      "loss": 0.1464,
+      "step": 8516
+    },
+    {
+      "epoch": 2.844214393053932,
+      "grad_norm": 0.5173838285389819,
+      "learning_rate": 8.143630399395031e-08,
+      "loss": 0.1457,
+      "step": 8517
+    },
+    {
+      "epoch": 2.844548338620805,
+      "grad_norm": 0.5329604081610628,
+      "learning_rate": 8.108736575079434e-08,
+      "loss": 0.1615,
+      "step": 8518
+    },
+    {
+      "epoch": 2.8448822841876775,
+      "grad_norm": 0.49406524837234916,
+      "learning_rate": 8.073917056731106e-08,
+      "loss": 0.1471,
+      "step": 8519
+    },
+    {
+      "epoch": 2.8452162297545502,
+      "grad_norm": 0.503715390690022,
+      "learning_rate": 8.039171849609728e-08,
+      "loss": 0.1481,
+      "step": 8520
+    },
+    {
+      "epoch": 2.8455501753214225,
+      "grad_norm": 0.5202325840935615,
+      "learning_rate": 8.004500958964211e-08,
+      "loss": 0.1472,
+      "step": 8521
+    },
+    {
+      "epoch": 2.8458841208882952,
+      "grad_norm": 0.5296350657811596,
+      "learning_rate": 7.969904390031812e-08,
+      "loss": 0.1518,
+      "step": 8522
+    },
+    {
+      "epoch": 2.846218066455168,
+      "grad_norm": 0.5206405247107793,
+      "learning_rate": 7.935382148038794e-08,
+      "loss": 0.1583,
+      "step": 8523
+    },
+    {
+      "epoch": 2.8465520120220402,
+      "grad_norm": 0.5080001963788301,
+      "learning_rate": 7.900934238200265e-08,
+      "loss": 0.1458,
+      "step": 8524
+    },
+    {
+      "epoch": 2.846885957588913,
+      "grad_norm": 0.5533271921936429,
+      "learning_rate": 7.866560665719836e-08,
+      "loss": 0.1606,
+      "step": 8525
+    },
+    {
+      "epoch": 2.8472199031557857,
+      "grad_norm": 0.5516994173535836,
+      "learning_rate": 7.832261435790078e-08,
+      "loss": 0.1581,
+      "step": 8526
+    },
+    {
+      "epoch": 2.847553848722658,
+      "grad_norm": 0.603291791816102,
+      "learning_rate": 7.798036553592403e-08,
+      "loss": 0.1708,
+      "step": 8527
+    },
+    {
+      "epoch": 2.8478877942895306,
+      "grad_norm": 0.5226165249381783,
+      "learning_rate": 7.763886024296729e-08,
+      "loss": 0.1494,
+      "step": 8528
+    },
+    {
+      "epoch": 2.8482217398564034,
+      "grad_norm": 0.4984085875281068,
+      "learning_rate": 7.729809853061987e-08,
+      "loss": 0.1477,
+      "step": 8529
+    },
+    {
+      "epoch": 2.848555685423276,
+      "grad_norm": 0.5546006723134326,
+      "learning_rate": 7.69580804503578e-08,
+      "loss": 0.16,
+      "step": 8530
+    },
+    {
+      "epoch": 2.848889630990149,
+      "grad_norm": 0.4951006251440752,
+      "learning_rate": 7.661880605354444e-08,
+      "loss": 0.1459,
+      "step": 8531
+    },
+    {
+      "epoch": 2.849223576557021,
+      "grad_norm": 0.5475486717210286,
+      "learning_rate": 7.628027539143156e-08,
+      "loss": 0.155,
+      "step": 8532
+    },
+    {
+      "epoch": 2.849557522123894,
+      "grad_norm": 0.4853112230938831,
+      "learning_rate": 7.594248851515717e-08,
+      "loss": 0.1446,
+      "step": 8533
+    },
+    {
+      "epoch": 2.8498914676907665,
+      "grad_norm": 0.5277882995951618,
+      "learning_rate": 7.560544547574988e-08,
+      "loss": 0.157,
+      "step": 8534
+    },
+    {
+      "epoch": 2.850225413257639,
+      "grad_norm": 0.5129719043043828,
+      "learning_rate": 7.526914632412175e-08,
+      "loss": 0.1493,
+      "step": 8535
+    },
+    {
+      "epoch": 2.8505593588245115,
+      "grad_norm": 0.5254590538526176,
+      "learning_rate": 7.493359111107712e-08,
+      "loss": 0.1562,
+      "step": 8536
+    },
+    {
+      "epoch": 2.850893304391384,
+      "grad_norm": 0.5457808147998661,
+      "learning_rate": 7.459877988730325e-08,
+      "loss": 0.1568,
+      "step": 8537
+    },
+    {
+      "epoch": 2.851227249958257,
+      "grad_norm": 0.5225349529691248,
+      "learning_rate": 7.42647127033791e-08,
+      "loss": 0.1501,
+      "step": 8538
+    },
+    {
+      "epoch": 2.8515611955251297,
+      "grad_norm": 0.5583061635314059,
+      "learning_rate": 7.393138960976876e-08,
+      "loss": 0.1536,
+      "step": 8539
+    },
+    {
+      "epoch": 2.851895141092002,
+      "grad_norm": 0.5121547924256945,
+      "learning_rate": 7.359881065682473e-08,
+      "loss": 0.1438,
+      "step": 8540
+    },
+    {
+      "epoch": 2.8522290866588746,
+      "grad_norm": 0.49864090386106596,
+      "learning_rate": 7.32669758947857e-08,
+      "loss": 0.144,
+      "step": 8541
+    },
+    {
+      "epoch": 2.8525630322257474,
+      "grad_norm": 0.4893222689278706,
+      "learning_rate": 7.29358853737816e-08,
+      "loss": 0.146,
+      "step": 8542
+    },
+    {
+      "epoch": 2.8528969777926196,
+      "grad_norm": 0.5348409656064933,
+      "learning_rate": 7.260553914382573e-08,
+      "loss": 0.149,
+      "step": 8543
+    },
+    {
+      "epoch": 2.8532309233594924,
+      "grad_norm": 0.5435223373432201,
+      "learning_rate": 7.227593725482207e-08,
+      "loss": 0.1587,
+      "step": 8544
+    },
+    {
+      "epoch": 2.853564868926365,
+      "grad_norm": 0.5217911226961017,
+      "learning_rate": 7.194707975655912e-08,
+      "loss": 0.1444,
+      "step": 8545
+    },
+    {
+      "epoch": 2.8538988144932373,
+      "grad_norm": 0.6065711866534749,
+      "learning_rate": 7.161896669871605e-08,
+      "loss": 0.1735,
+      "step": 8546
+    },
+    {
+      "epoch": 2.85423276006011,
+      "grad_norm": 0.5330691052638803,
+      "learning_rate": 7.129159813085817e-08,
+      "loss": 0.158,
+      "step": 8547
+    },
+    {
+      "epoch": 2.854566705626983,
+      "grad_norm": 0.5350161629397456,
+      "learning_rate": 7.096497410243819e-08,
+      "loss": 0.1528,
+      "step": 8548
+    },
+    {
+      "epoch": 2.8549006511938555,
+      "grad_norm": 0.5560545581651952,
+      "learning_rate": 7.063909466279605e-08,
+      "loss": 0.1656,
+      "step": 8549
+    },
+    {
+      "epoch": 2.855234596760728,
+      "grad_norm": 0.585578949405154,
+      "learning_rate": 7.031395986116019e-08,
+      "loss": 0.1588,
+      "step": 8550
+    },
+    {
+      "epoch": 2.8555685423276005,
+      "grad_norm": 0.5165515533972673,
+      "learning_rate": 6.998956974664573e-08,
+      "loss": 0.1473,
+      "step": 8551
+    },
+    {
+      "epoch": 2.855902487894473,
+      "grad_norm": 0.5970286051761957,
+      "learning_rate": 6.966592436825514e-08,
+      "loss": 0.1683,
+      "step": 8552
+    },
+    {
+      "epoch": 2.856236433461346,
+      "grad_norm": 0.5173373849701015,
+      "learning_rate": 6.934302377488045e-08,
+      "loss": 0.1488,
+      "step": 8553
+    },
+    {
+      "epoch": 2.856570379028218,
+      "grad_norm": 0.5131753943094373,
+      "learning_rate": 6.902086801529817e-08,
+      "loss": 0.1519,
+      "step": 8554
+    },
+    {
+      "epoch": 2.856904324595091,
+      "grad_norm": 0.5156084894747482,
+      "learning_rate": 6.869945713817438e-08,
+      "loss": 0.1485,
+      "step": 8555
+    },
+    {
+      "epoch": 2.8572382701619636,
+      "grad_norm": 0.4849489841706419,
+      "learning_rate": 6.837879119206192e-08,
+      "loss": 0.1476,
+      "step": 8556
+    },
+    {
+      "epoch": 2.8575722157288364,
+      "grad_norm": 0.6184373363815887,
+      "learning_rate": 6.805887022540093e-08,
+      "loss": 0.1686,
+      "step": 8557
+    },
+    {
+      "epoch": 2.857906161295709,
+      "grad_norm": 0.5476203063313209,
+      "learning_rate": 6.773969428651883e-08,
+      "loss": 0.165,
+      "step": 8558
+    },
+    {
+      "epoch": 2.8582401068625813,
+      "grad_norm": 0.5373994473998241,
+      "learning_rate": 6.742126342363153e-08,
+      "loss": 0.1498,
+      "step": 8559
+    },
+    {
+      "epoch": 2.858574052429454,
+      "grad_norm": 0.5178688034759278,
+      "learning_rate": 6.710357768484165e-08,
+      "loss": 0.1571,
+      "step": 8560
+    },
+    {
+      "epoch": 2.8589079979963268,
+      "grad_norm": 0.5039114208119598,
+      "learning_rate": 6.67866371181397e-08,
+      "loss": 0.1535,
+      "step": 8561
+    },
+    {
+      "epoch": 2.859241943563199,
+      "grad_norm": 0.5332718448517993,
+      "learning_rate": 6.647044177140293e-08,
+      "loss": 0.1601,
+      "step": 8562
+    },
+    {
+      "epoch": 2.8595758891300718,
+      "grad_norm": 0.5102172652244273,
+      "learning_rate": 6.615499169239647e-08,
+      "loss": 0.1399,
+      "step": 8563
+    },
+    {
+      "epoch": 2.8599098346969445,
+      "grad_norm": 0.531015741793665,
+      "learning_rate": 6.584028692877164e-08,
+      "loss": 0.1588,
+      "step": 8564
+    },
+    {
+      "epoch": 2.8602437802638168,
+      "grad_norm": 0.523805491269051,
+      "learning_rate": 6.552632752807042e-08,
+      "loss": 0.1589,
+      "step": 8565
+    },
+    {
+      "epoch": 2.8605777258306895,
+      "grad_norm": 0.5326354340246765,
+      "learning_rate": 6.52131135377182e-08,
+      "loss": 0.1538,
+      "step": 8566
+    },
+    {
+      "epoch": 2.860911671397562,
+      "grad_norm": 0.511949106766765,
+      "learning_rate": 6.490064500503102e-08,
+      "loss": 0.1453,
+      "step": 8567
+    },
+    {
+      "epoch": 2.861245616964435,
+      "grad_norm": 0.5271829727987514,
+      "learning_rate": 6.458892197721e-08,
+      "loss": 0.1585,
+      "step": 8568
+    },
+    {
+      "epoch": 2.8615795625313076,
+      "grad_norm": 0.5430974193562058,
+      "learning_rate": 6.427794450134529e-08,
+      "loss": 0.1573,
+      "step": 8569
+    },
+    {
+      "epoch": 2.86191350809818,
+      "grad_norm": 0.5402436743016346,
+      "learning_rate": 6.396771262441259e-08,
+      "loss": 0.1587,
+      "step": 8570
+    },
+    {
+      "epoch": 2.8622474536650526,
+      "grad_norm": 0.5228404957057184,
+      "learning_rate": 6.365822639327724e-08,
+      "loss": 0.1527,
+      "step": 8571
+    },
+    {
+      "epoch": 2.8625813992319253,
+      "grad_norm": 0.5431852960693747,
+      "learning_rate": 6.334948585469014e-08,
+      "loss": 0.1516,
+      "step": 8572
+    },
+    {
+      "epoch": 2.8629153447987976,
+      "grad_norm": 0.5535152802323873,
+      "learning_rate": 6.304149105529067e-08,
+      "loss": 0.1597,
+      "step": 8573
+    },
+    {
+      "epoch": 2.8632492903656703,
+      "grad_norm": 0.503798876425636,
+      "learning_rate": 6.273424204160438e-08,
+      "loss": 0.1581,
+      "step": 8574
+    },
+    {
+      "epoch": 2.863583235932543,
+      "grad_norm": 0.4954582075562573,
+      "learning_rate": 6.242773886004583e-08,
+      "loss": 0.148,
+      "step": 8575
+    },
+    {
+      "epoch": 2.8639171814994153,
+      "grad_norm": 0.5227935230801231,
+      "learning_rate": 6.212198155691518e-08,
+      "loss": 0.1569,
+      "step": 8576
+    },
+    {
+      "epoch": 2.864251127066288,
+      "grad_norm": 0.5072845345431894,
+      "learning_rate": 6.181697017840049e-08,
+      "loss": 0.1568,
+      "step": 8577
+    },
+    {
+      "epoch": 2.8645850726331608,
+      "grad_norm": 0.5680395290888719,
+      "learning_rate": 6.151270477057825e-08,
+      "loss": 0.1665,
+      "step": 8578
+    },
+    {
+      "epoch": 2.8649190182000335,
+      "grad_norm": 0.5723718892695402,
+      "learning_rate": 6.120918537941001e-08,
+      "loss": 0.1689,
+      "step": 8579
+    },
+    {
+      "epoch": 2.865252963766906,
+      "grad_norm": 0.5543284280807889,
+      "learning_rate": 6.090641205074743e-08,
+      "loss": 0.164,
+      "step": 8580
+    },
+    {
+      "epoch": 2.8655869093337785,
+      "grad_norm": 0.5035318018400274,
+      "learning_rate": 6.060438483032671e-08,
+      "loss": 0.142,
+      "step": 8581
+    },
+    {
+      "epoch": 2.865920854900651,
+      "grad_norm": 0.5355651436216142,
+      "learning_rate": 6.030310376377302e-08,
+      "loss": 0.1612,
+      "step": 8582
+    },
+    {
+      "epoch": 2.866254800467524,
+      "grad_norm": 0.48977033513131546,
+      "learning_rate": 6.000256889659883e-08,
+      "loss": 0.144,
+      "step": 8583
+    },
+    {
+      "epoch": 2.866588746034396,
+      "grad_norm": 0.5436549484026989,
+      "learning_rate": 5.97027802742034e-08,
+      "loss": 0.168,
+      "step": 8584
+    },
+    {
+      "epoch": 2.866922691601269,
+      "grad_norm": 0.572267212305895,
+      "learning_rate": 5.940373794187326e-08,
+      "loss": 0.1636,
+      "step": 8585
+    },
+    {
+      "epoch": 2.8672566371681416,
+      "grad_norm": 0.5209482329083956,
+      "learning_rate": 5.910544194478174e-08,
+      "loss": 0.1509,
+      "step": 8586
+    },
+    {
+      "epoch": 2.8675905827350143,
+      "grad_norm": 0.5484505549507601,
+      "learning_rate": 5.880789232799e-08,
+      "loss": 0.1565,
+      "step": 8587
+    },
+    {
+      "epoch": 2.867924528301887,
+      "grad_norm": 0.5478787273088694,
+      "learning_rate": 5.851108913644765e-08,
+      "loss": 0.1641,
+      "step": 8588
+    },
+    {
+      "epoch": 2.8682584738687593,
+      "grad_norm": 0.5522406436362908,
+      "learning_rate": 5.821503241498882e-08,
+      "loss": 0.1636,
+      "step": 8589
+    },
+    {
+      "epoch": 2.868592419435632,
+      "grad_norm": 0.511907461284583,
+      "learning_rate": 5.791972220833719e-08,
+      "loss": 0.1552,
+      "step": 8590
+    },
+    {
+      "epoch": 2.8689263650025048,
+      "grad_norm": 0.5152899425702656,
+      "learning_rate": 5.762515856110262e-08,
+      "loss": 0.1517,
+      "step": 8591
+    },
+    {
+      "epoch": 2.869260310569377,
+      "grad_norm": 0.5258169726148847,
+      "learning_rate": 5.7331341517782855e-08,
+      "loss": 0.1528,
+      "step": 8592
+    },
+    {
+      "epoch": 2.8695942561362497,
+      "grad_norm": 0.4713808899737005,
+      "learning_rate": 5.703827112276128e-08,
+      "loss": 0.1416,
+      "step": 8593
+    },
+    {
+      "epoch": 2.8699282017031225,
+      "grad_norm": 0.5273185402336777,
+      "learning_rate": 5.674594742031081e-08,
+      "loss": 0.1601,
+      "step": 8594
+    },
+    {
+      "epoch": 2.8702621472699947,
+      "grad_norm": 0.5220436578594836,
+      "learning_rate": 5.6454370454589456e-08,
+      "loss": 0.1562,
+      "step": 8595
+    },
+    {
+      "epoch": 2.8705960928368675,
+      "grad_norm": 0.5221063380938278,
+      "learning_rate": 5.6163540269644215e-08,
+      "loss": 0.1518,
+      "step": 8596
+    },
+    {
+      "epoch": 2.87093003840374,
+      "grad_norm": 0.4974464271881414,
+      "learning_rate": 5.5873456909407706e-08,
+      "loss": 0.1505,
+      "step": 8597
+    },
+    {
+      "epoch": 2.871263983970613,
+      "grad_norm": 0.5046804081978551,
+      "learning_rate": 5.5584120417701005e-08,
+      "loss": 0.1511,
+      "step": 8598
+    },
+    {
+      "epoch": 2.8715979295374856,
+      "grad_norm": 0.5053148608639023,
+      "learning_rate": 5.529553083823136e-08,
+      "loss": 0.1516,
+      "step": 8599
+    },
+    {
+      "epoch": 2.871931875104358,
+      "grad_norm": 0.5777837178110677,
+      "learning_rate": 5.50076882145939e-08,
+      "loss": 0.1639,
+      "step": 8600
+    },
+    {
+      "epoch": 2.8722658206712306,
+      "grad_norm": 0.5620172653957106,
+      "learning_rate": 5.472059259027051e-08,
+      "loss": 0.1632,
+      "step": 8601
+    },
+    {
+      "epoch": 2.8725997662381033,
+      "grad_norm": 0.5347887245031796,
+      "learning_rate": 5.44342440086304e-08,
+      "loss": 0.1568,
+      "step": 8602
+    },
+    {
+      "epoch": 2.8729337118049756,
+      "grad_norm": 0.5236727280193502,
+      "learning_rate": 5.414864251293006e-08,
+      "loss": 0.1484,
+      "step": 8603
+    },
+    {
+      "epoch": 2.8732676573718483,
+      "grad_norm": 0.5322026712399329,
+      "learning_rate": 5.386378814631277e-08,
+      "loss": 0.1536,
+      "step": 8604
+    },
+    {
+      "epoch": 2.873601602938721,
+      "grad_norm": 0.534740394002748,
+      "learning_rate": 5.3579680951808545e-08,
+      "loss": 0.1604,
+      "step": 8605
+    },
+    {
+      "epoch": 2.8739355485055937,
+      "grad_norm": 0.5853877385693443,
+      "learning_rate": 5.329632097233639e-08,
+      "loss": 0.1597,
+      "step": 8606
+    },
+    {
+      "epoch": 2.8742694940724665,
+      "grad_norm": 0.5011755113078209,
+      "learning_rate": 5.3013708250700405e-08,
+      "loss": 0.1491,
+      "step": 8607
+    },
+    {
+      "epoch": 2.8746034396393387,
+      "grad_norm": 0.5460105318096763,
+      "learning_rate": 5.2731842829591984e-08,
+      "loss": 0.1603,
+      "step": 8608
+    },
+    {
+      "epoch": 2.8749373852062114,
+      "grad_norm": 0.5535251387041829,
+      "learning_rate": 5.2450724751592076e-08,
+      "loss": 0.1646,
+      "step": 8609
+    },
+    {
+      "epoch": 2.875271330773084,
+      "grad_norm": 0.5412106966495831,
+      "learning_rate": 5.217035405916449e-08,
+      "loss": 0.1524,
+      "step": 8610
+    },
+    {
+      "epoch": 2.8756052763399564,
+      "grad_norm": 0.48812577783849903,
+      "learning_rate": 5.1890730794664227e-08,
+      "loss": 0.1438,
+      "step": 8611
+    },
+    {
+      "epoch": 2.875939221906829,
+      "grad_norm": 0.5630739912686819,
+      "learning_rate": 5.161185500033139e-08,
+      "loss": 0.1621,
+      "step": 8612
+    },
+    {
+      "epoch": 2.876273167473702,
+      "grad_norm": 0.5637930034961525,
+      "learning_rate": 5.1333726718293396e-08,
+      "loss": 0.1584,
+      "step": 8613
+    },
+    {
+      "epoch": 2.876607113040574,
+      "grad_norm": 0.5413995620609211,
+      "learning_rate": 5.105634599056386e-08,
+      "loss": 0.1517,
+      "step": 8614
+    },
+    {
+      "epoch": 2.876941058607447,
+      "grad_norm": 0.5284011452682914,
+      "learning_rate": 5.077971285904593e-08,
+      "loss": 0.1532,
+      "step": 8615
+    },
+    {
+      "epoch": 2.8772750041743196,
+      "grad_norm": 0.5373211068578061,
+      "learning_rate": 5.050382736552728e-08,
+      "loss": 0.1567,
+      "step": 8616
+    },
+    {
+      "epoch": 2.8776089497411923,
+      "grad_norm": 0.5164966070831943,
+      "learning_rate": 5.022868955168403e-08,
+      "loss": 0.1498,
+      "step": 8617
+    },
+    {
+      "epoch": 2.877942895308065,
+      "grad_norm": 0.5016749655968777,
+      "learning_rate": 4.995429945907848e-08,
+      "loss": 0.148,
+      "step": 8618
+    },
+    {
+      "epoch": 2.8782768408749373,
+      "grad_norm": 0.5350771758808319,
+      "learning_rate": 4.968065712916137e-08,
+      "loss": 0.1571,
+      "step": 8619
+    },
+    {
+      "epoch": 2.87861078644181,
+      "grad_norm": 0.5315077919152822,
+      "learning_rate": 4.940776260326907e-08,
+      "loss": 0.149,
+      "step": 8620
+    },
+    {
+      "epoch": 2.8789447320086827,
+      "grad_norm": 0.530178375587157,
+      "learning_rate": 4.913561592262528e-08,
+      "loss": 0.1555,
+      "step": 8621
+    },
+    {
+      "epoch": 2.879278677575555,
+      "grad_norm": 0.5217846163655511,
+      "learning_rate": 4.886421712834155e-08,
+      "loss": 0.1571,
+      "step": 8622
+    },
+    {
+      "epoch": 2.8796126231424277,
+      "grad_norm": 0.5115490799371355,
+      "learning_rate": 4.859356626141509e-08,
+      "loss": 0.1491,
+      "step": 8623
+    },
+    {
+      "epoch": 2.8799465687093004,
+      "grad_norm": 0.5516763395859796,
+      "learning_rate": 4.8323663362732084e-08,
+      "loss": 0.166,
+      "step": 8624
+    },
+    {
+      "epoch": 2.8802805142761727,
+      "grad_norm": 0.5599558340991891,
+      "learning_rate": 4.8054508473063253e-08,
+      "loss": 0.1685,
+      "step": 8625
+    },
+    {
+      "epoch": 2.8806144598430454,
+      "grad_norm": 0.4982081440561017,
+      "learning_rate": 4.778610163306885e-08,
+      "loss": 0.1551,
+      "step": 8626
+    },
+    {
+      "epoch": 2.880948405409918,
+      "grad_norm": 0.5606710703278645,
+      "learning_rate": 4.751844288329366e-08,
+      "loss": 0.1567,
+      "step": 8627
+    },
+    {
+      "epoch": 2.881282350976791,
+      "grad_norm": 0.5087763751751658,
+      "learning_rate": 4.72515322641709e-08,
+      "loss": 0.1464,
+      "step": 8628
+    },
+    {
+      "epoch": 2.8816162965436636,
+      "grad_norm": 0.552549417686956,
+      "learning_rate": 4.6985369816021644e-08,
+      "loss": 0.1565,
+      "step": 8629
+    },
+    {
+      "epoch": 2.881950242110536,
+      "grad_norm": 0.49592917632561173,
+      "learning_rate": 4.6719955579052064e-08,
+      "loss": 0.1448,
+      "step": 8630
+    },
+    {
+      "epoch": 2.8822841876774086,
+      "grad_norm": 0.5255353569803617,
+      "learning_rate": 4.6455289593355656e-08,
+      "loss": 0.1557,
+      "step": 8631
+    },
+    {
+      "epoch": 2.8826181332442813,
+      "grad_norm": 0.5534574457295879,
+      "learning_rate": 4.619137189891432e-08,
+      "loss": 0.163,
+      "step": 8632
+    },
+    {
+      "epoch": 2.8829520788111536,
+      "grad_norm": 0.5397072465178986,
+      "learning_rate": 4.5928202535595044e-08,
+      "loss": 0.157,
+      "step": 8633
+    },
+    {
+      "epoch": 2.8832860243780263,
+      "grad_norm": 0.549349296130443,
+      "learning_rate": 4.5665781543153266e-08,
+      "loss": 0.1645,
+      "step": 8634
+    },
+    {
+      "epoch": 2.883619969944899,
+      "grad_norm": 0.5420717576627341,
+      "learning_rate": 4.54041089612306e-08,
+      "loss": 0.1656,
+      "step": 8635
+    },
+    {
+      "epoch": 2.8839539155117717,
+      "grad_norm": 0.5363732368933584,
+      "learning_rate": 4.514318482935598e-08,
+      "loss": 0.1617,
+      "step": 8636
+    },
+    {
+      "epoch": 2.8842878610786444,
+      "grad_norm": 0.5350097682625145,
+      "learning_rate": 4.488300918694455e-08,
+      "loss": 0.1562,
+      "step": 8637
+    },
+    {
+      "epoch": 2.8846218066455167,
+      "grad_norm": 0.5246389948891459,
+      "learning_rate": 4.4623582073299864e-08,
+      "loss": 0.158,
+      "step": 8638
+    },
+    {
+      "epoch": 2.8849557522123894,
+      "grad_norm": 0.5436712211339741,
+      "learning_rate": 4.4364903527610026e-08,
+      "loss": 0.1545,
+      "step": 8639
+    },
+    {
+      "epoch": 2.885289697779262,
+      "grad_norm": 0.5009484631759727,
+      "learning_rate": 4.410697358895211e-08,
+      "loss": 0.1464,
+      "step": 8640
+    },
+    {
+      "epoch": 2.8856236433461344,
+      "grad_norm": 0.5290374848110081,
+      "learning_rate": 4.384979229628994e-08,
+      "loss": 0.1482,
+      "step": 8641
+    },
+    {
+      "epoch": 2.885957588913007,
+      "grad_norm": 0.5504796169207347,
+      "learning_rate": 4.359335968847356e-08,
+      "loss": 0.1583,
+      "step": 8642
+    },
+    {
+      "epoch": 2.88629153447988,
+      "grad_norm": 0.5589515588770203,
+      "learning_rate": 4.333767580423976e-08,
+      "loss": 0.1672,
+      "step": 8643
+    },
+    {
+      "epoch": 2.886625480046752,
+      "grad_norm": 0.5394714452979373,
+      "learning_rate": 4.3082740682213186e-08,
+      "loss": 0.1581,
+      "step": 8644
+    },
+    {
+      "epoch": 2.886959425613625,
+      "grad_norm": 0.5311277472998632,
+      "learning_rate": 4.2828554360904165e-08,
+      "loss": 0.16,
+      "step": 8645
+    },
+    {
+      "epoch": 2.8872933711804976,
+      "grad_norm": 0.5214855770881811,
+      "learning_rate": 4.25751168787103e-08,
+      "loss": 0.1584,
+      "step": 8646
+    },
+    {
+      "epoch": 2.8876273167473703,
+      "grad_norm": 0.5196864326489907,
+      "learning_rate": 4.2322428273917635e-08,
+      "loss": 0.1505,
+      "step": 8647
+    },
+    {
+      "epoch": 2.887961262314243,
+      "grad_norm": 0.5395236365751814,
+      "learning_rate": 4.2070488584696754e-08,
+      "loss": 0.1583,
+      "step": 8648
+    },
+    {
+      "epoch": 2.8882952078811153,
+      "grad_norm": 0.5281187130550312,
+      "learning_rate": 4.18192978491061e-08,
+      "loss": 0.157,
+      "step": 8649
+    },
+    {
+      "epoch": 2.888629153447988,
+      "grad_norm": 0.5335230980635554,
+      "learning_rate": 4.1568856105091424e-08,
+      "loss": 0.1605,
+      "step": 8650
+    },
+    {
+      "epoch": 2.8889630990148607,
+      "grad_norm": 0.5585441771953708,
+      "learning_rate": 4.1319163390484693e-08,
+      "loss": 0.1614,
+      "step": 8651
+    },
+    {
+      "epoch": 2.889297044581733,
+      "grad_norm": 0.5469020513513515,
+      "learning_rate": 4.107021974300407e-08,
+      "loss": 0.1586,
+      "step": 8652
+    },
+    {
+      "epoch": 2.8896309901486057,
+      "grad_norm": 0.507051026666806,
+      "learning_rate": 4.082202520025724e-08,
+      "loss": 0.151,
+      "step": 8653
+    },
+    {
+      "epoch": 2.8899649357154784,
+      "grad_norm": 0.4997794435735754,
+      "learning_rate": 4.0574579799735335e-08,
+      "loss": 0.1458,
+      "step": 8654
+    },
+    {
+      "epoch": 2.890298881282351,
+      "grad_norm": 0.530000754000752,
+      "learning_rate": 4.0327883578819006e-08,
+      "loss": 0.1558,
+      "step": 8655
+    },
+    {
+      "epoch": 2.890632826849224,
+      "grad_norm": 0.5068361461273592,
+      "learning_rate": 4.008193657477399e-08,
+      "loss": 0.1485,
+      "step": 8656
+    },
+    {
+      "epoch": 2.890966772416096,
+      "grad_norm": 0.5529734401920643,
+      "learning_rate": 3.9836738824753364e-08,
+      "loss": 0.1637,
+      "step": 8657
+    },
+    {
+      "epoch": 2.891300717982969,
+      "grad_norm": 0.5096699557010229,
+      "learning_rate": 3.959229036579748e-08,
+      "loss": 0.1483,
+      "step": 8658
+    },
+    {
+      "epoch": 2.8916346635498416,
+      "grad_norm": 0.4798650693812542,
+      "learning_rate": 3.9348591234832926e-08,
+      "loss": 0.1448,
+      "step": 8659
+    },
+    {
+      "epoch": 2.891968609116714,
+      "grad_norm": 0.4990085154143496,
+      "learning_rate": 3.9105641468673574e-08,
+      "loss": 0.1561,
+      "step": 8660
+    },
+    {
+      "epoch": 2.8923025546835865,
+      "grad_norm": 0.5224750928551145,
+      "learning_rate": 3.886344110402007e-08,
+      "loss": 0.1531,
+      "step": 8661
+    },
+    {
+      "epoch": 2.8926365002504593,
+      "grad_norm": 0.5338496892667479,
+      "learning_rate": 3.862199017745871e-08,
+      "loss": 0.1585,
+      "step": 8662
+    },
+    {
+      "epoch": 2.8929704458173315,
+      "grad_norm": 0.5134775255661097,
+      "learning_rate": 3.838128872546421e-08,
+      "loss": 0.1547,
+      "step": 8663
+    },
+    {
+      "epoch": 2.8933043913842043,
+      "grad_norm": 0.5331918577725364,
+      "learning_rate": 3.814133678439691e-08,
+      "loss": 0.1587,
+      "step": 8664
+    },
+    {
+      "epoch": 2.893638336951077,
+      "grad_norm": 0.517842146198162,
+      "learning_rate": 3.790213439050561e-08,
+      "loss": 0.1533,
+      "step": 8665
+    },
+    {
+      "epoch": 2.8939722825179497,
+      "grad_norm": 0.5189498284952392,
+      "learning_rate": 3.766368157992306e-08,
+      "loss": 0.1523,
+      "step": 8666
+    },
+    {
+      "epoch": 2.8943062280848224,
+      "grad_norm": 0.49860727440804087,
+      "learning_rate": 3.7425978388671014e-08,
+      "loss": 0.1468,
+      "step": 8667
+    },
+    {
+      "epoch": 2.8946401736516947,
+      "grad_norm": 0.5297330753233082,
+      "learning_rate": 3.718902485265741e-08,
+      "loss": 0.1577,
+      "step": 8668
+    },
+    {
+      "epoch": 2.8949741192185674,
+      "grad_norm": 0.5334255049213774,
+      "learning_rate": 3.6952821007676943e-08,
+      "loss": 0.1537,
+      "step": 8669
+    },
+    {
+      "epoch": 2.89530806478544,
+      "grad_norm": 0.5289833683098253,
+      "learning_rate": 3.671736688941108e-08,
+      "loss": 0.152,
+      "step": 8670
+    },
+    {
+      "epoch": 2.8956420103523124,
+      "grad_norm": 0.4915608352001022,
+      "learning_rate": 3.6482662533426914e-08,
+      "loss": 0.1436,
+      "step": 8671
+    },
+    {
+      "epoch": 2.895975955919185,
+      "grad_norm": 0.5486892456552841,
+      "learning_rate": 3.6248707975181096e-08,
+      "loss": 0.1543,
+      "step": 8672
+    },
+    {
+      "epoch": 2.896309901486058,
+      "grad_norm": 0.5503337548743263,
+      "learning_rate": 3.601550325001313e-08,
+      "loss": 0.1567,
+      "step": 8673
+    },
+    {
+      "epoch": 2.89664384705293,
+      "grad_norm": 0.5034194587207542,
+      "learning_rate": 3.578304839315316e-08,
+      "loss": 0.1445,
+      "step": 8674
+    },
+    {
+      "epoch": 2.896977792619803,
+      "grad_norm": 0.5146376386794662,
+      "learning_rate": 3.5551343439715336e-08,
+      "loss": 0.1441,
+      "step": 8675
+    },
+    {
+      "epoch": 2.8973117381866755,
+      "grad_norm": 0.5207461435783136,
+      "learning_rate": 3.5320388424701644e-08,
+      "loss": 0.1455,
+      "step": 8676
+    },
+    {
+      "epoch": 2.8976456837535483,
+      "grad_norm": 0.5822846711151547,
+      "learning_rate": 3.50901833830003e-08,
+      "loss": 0.1609,
+      "step": 8677
+    },
+    {
+      "epoch": 2.897979629320421,
+      "grad_norm": 0.4881700169042898,
+      "learning_rate": 3.4860728349386807e-08,
+      "loss": 0.151,
+      "step": 8678
+    },
+    {
+      "epoch": 2.8983135748872932,
+      "grad_norm": 0.5621883536149694,
+      "learning_rate": 3.4632023358522894e-08,
+      "loss": 0.1619,
+      "step": 8679
+    },
+    {
+      "epoch": 2.898647520454166,
+      "grad_norm": 0.5110749883128612,
+      "learning_rate": 3.440406844495758e-08,
+      "loss": 0.1549,
+      "step": 8680
+    },
+    {
+      "epoch": 2.8989814660210387,
+      "grad_norm": 0.4713601733022798,
+      "learning_rate": 3.4176863643125e-08,
+      "loss": 0.1366,
+      "step": 8681
+    },
+    {
+      "epoch": 2.899315411587911,
+      "grad_norm": 0.5227456137198871,
+      "learning_rate": 3.395040898734825e-08,
+      "loss": 0.1552,
+      "step": 8682
+    },
+    {
+      "epoch": 2.8996493571547837,
+      "grad_norm": 0.6136173309903541,
+      "learning_rate": 3.372470451183496e-08,
+      "loss": 0.1657,
+      "step": 8683
+    },
+    {
+      "epoch": 2.8999833027216564,
+      "grad_norm": 0.5297325697192652,
+      "learning_rate": 3.349975025068175e-08,
+      "loss": 0.1527,
+      "step": 8684
+    },
+    {
+      "epoch": 2.900317248288529,
+      "grad_norm": 0.6288686454375647,
+      "learning_rate": 3.327554623786977e-08,
+      "loss": 0.161,
+      "step": 8685
+    },
+    {
+      "epoch": 2.900651193855402,
+      "grad_norm": 0.5269797630905554,
+      "learning_rate": 3.305209250726804e-08,
+      "loss": 0.1501,
+      "step": 8686
+    },
+    {
+      "epoch": 2.900985139422274,
+      "grad_norm": 0.5554879926954489,
+      "learning_rate": 3.282938909263122e-08,
+      "loss": 0.1624,
+      "step": 8687
+    },
+    {
+      "epoch": 2.901319084989147,
+      "grad_norm": 0.5208615726777216,
+      "learning_rate": 3.2607436027601854e-08,
+      "loss": 0.1534,
+      "step": 8688
+    },
+    {
+      "epoch": 2.9016530305560195,
+      "grad_norm": 0.5420289935437654,
+      "learning_rate": 3.238623334570812e-08,
+      "loss": 0.1535,
+      "step": 8689
+    },
+    {
+      "epoch": 2.901986976122892,
+      "grad_norm": 0.538732737361838,
+      "learning_rate": 3.2165781080366054e-08,
+      "loss": 0.1569,
+      "step": 8690
+    },
+    {
+      "epoch": 2.9023209216897645,
+      "grad_norm": 0.5313485948505864,
+      "learning_rate": 3.194607926487681e-08,
+      "loss": 0.1586,
+      "step": 8691
+    },
+    {
+      "epoch": 2.9026548672566372,
+      "grad_norm": 0.5356772127350496,
+      "learning_rate": 3.1727127932429936e-08,
+      "loss": 0.1568,
+      "step": 8692
+    },
+    {
+      "epoch": 2.9029888128235095,
+      "grad_norm": 0.5169393673994584,
+      "learning_rate": 3.150892711609899e-08,
+      "loss": 0.1603,
+      "step": 8693
+    },
+    {
+      "epoch": 2.9033227583903822,
+      "grad_norm": 0.4894913655189434,
+      "learning_rate": 3.129147684884704e-08,
+      "loss": 0.1385,
+      "step": 8694
+    },
+    {
+      "epoch": 2.903656703957255,
+      "grad_norm": 0.53281078755593,
+      "learning_rate": 3.107477716352225e-08,
+      "loss": 0.1507,
+      "step": 8695
+    },
+    {
+      "epoch": 2.9039906495241277,
+      "grad_norm": 0.5359916182265713,
+      "learning_rate": 3.0858828092859564e-08,
+      "loss": 0.1701,
+      "step": 8696
+    },
+    {
+      "epoch": 2.9043245950910004,
+      "grad_norm": 0.5262814413620647,
+      "learning_rate": 3.0643629669480644e-08,
+      "loss": 0.1534,
+      "step": 8697
+    },
+    {
+      "epoch": 2.9046585406578727,
+      "grad_norm": 0.5497355759559763,
+      "learning_rate": 3.042918192589395e-08,
+      "loss": 0.1596,
+      "step": 8698
+    },
+    {
+      "epoch": 2.9049924862247454,
+      "grad_norm": 0.5412939441164044,
+      "learning_rate": 3.021548489449355e-08,
+      "loss": 0.1588,
+      "step": 8699
+    },
+    {
+      "epoch": 2.905326431791618,
+      "grad_norm": 0.49003677617408303,
+      "learning_rate": 3.000253860756197e-08,
+      "loss": 0.1559,
+      "step": 8700
+    },
+    {
+      "epoch": 2.9056603773584904,
+      "grad_norm": 0.5473785424325217,
+      "learning_rate": 2.979034309726625e-08,
+      "loss": 0.1618,
+      "step": 8701
+    },
+    {
+      "epoch": 2.905994322925363,
+      "grad_norm": 0.5097771881556972,
+      "learning_rate": 2.9578898395661858e-08,
+      "loss": 0.1488,
+      "step": 8702
+    },
+    {
+      "epoch": 2.906328268492236,
+      "grad_norm": 0.5149654857058888,
+      "learning_rate": 2.9368204534689916e-08,
+      "loss": 0.1468,
+      "step": 8703
+    },
+    {
+      "epoch": 2.9066622140591085,
+      "grad_norm": 0.5259625454544891,
+      "learning_rate": 2.915826154617718e-08,
+      "loss": 0.1545,
+      "step": 8704
+    },
+    {
+      "epoch": 2.9069961596259812,
+      "grad_norm": 0.48948710540364265,
+      "learning_rate": 2.8949069461839952e-08,
+      "loss": 0.1401,
+      "step": 8705
+    },
+    {
+      "epoch": 2.9073301051928535,
+      "grad_norm": 0.5531487170068914,
+      "learning_rate": 2.8740628313276842e-08,
+      "loss": 0.1606,
+      "step": 8706
+    },
+    {
+      "epoch": 2.9076640507597262,
+      "grad_norm": 0.5032223065794879,
+      "learning_rate": 2.853293813197766e-08,
+      "loss": 0.1505,
+      "step": 8707
+    },
+    {
+      "epoch": 2.907997996326599,
+      "grad_norm": 0.5328715714346113,
+      "learning_rate": 2.8325998949314536e-08,
+      "loss": 0.1536,
+      "step": 8708
+    },
+    {
+      "epoch": 2.908331941893471,
+      "grad_norm": 0.6225983310360557,
+      "learning_rate": 2.811981079654913e-08,
+      "loss": 0.1652,
+      "step": 8709
+    },
+    {
+      "epoch": 2.908665887460344,
+      "grad_norm": 0.5526958652803692,
+      "learning_rate": 2.7914373704827634e-08,
+      "loss": 0.1601,
+      "step": 8710
+    },
+    {
+      "epoch": 2.9089998330272167,
+      "grad_norm": 0.5632023342482809,
+      "learning_rate": 2.7709687705185227e-08,
+      "loss": 0.1673,
+      "step": 8711
+    },
+    {
+      "epoch": 2.909333778594089,
+      "grad_norm": 0.5155594811296534,
+      "learning_rate": 2.7505752828541065e-08,
+      "loss": 0.1518,
+      "step": 8712
+    },
+    {
+      "epoch": 2.9096677241609616,
+      "grad_norm": 0.5339483867535059,
+      "learning_rate": 2.730256910570217e-08,
+      "loss": 0.1641,
+      "step": 8713
+    },
+    {
+      "epoch": 2.9100016697278344,
+      "grad_norm": 0.5461665859120574,
+      "learning_rate": 2.7100136567361767e-08,
+      "loss": 0.1576,
+      "step": 8714
+    },
+    {
+      "epoch": 2.910335615294707,
+      "grad_norm": 0.5336837655705691,
+      "learning_rate": 2.689845524409984e-08,
+      "loss": 0.1493,
+      "step": 8715
+    },
+    {
+      "epoch": 2.91066956086158,
+      "grad_norm": 0.4644299519054088,
+      "learning_rate": 2.6697525166382575e-08,
+      "loss": 0.1365,
+      "step": 8716
+    },
+    {
+      "epoch": 2.911003506428452,
+      "grad_norm": 0.5243383754514328,
+      "learning_rate": 2.649734636456236e-08,
+      "loss": 0.1487,
+      "step": 8717
+    },
+    {
+      "epoch": 2.911337451995325,
+      "grad_norm": 0.566255311577455,
+      "learning_rate": 2.629791886888e-08,
+      "loss": 0.1657,
+      "step": 8718
+    },
+    {
+      "epoch": 2.9116713975621975,
+      "grad_norm": 0.5151497217026064,
+      "learning_rate": 2.6099242709459737e-08,
+      "loss": 0.1535,
+      "step": 8719
+    },
+    {
+      "epoch": 2.91200534312907,
+      "grad_norm": 0.5286692237715818,
+      "learning_rate": 2.5901317916314783e-08,
+      "loss": 0.1574,
+      "step": 8720
+    },
+    {
+      "epoch": 2.9123392886959425,
+      "grad_norm": 0.5286053495921429,
+      "learning_rate": 2.5704144519344e-08,
+      "loss": 0.1609,
+      "step": 8721
+    },
+    {
+      "epoch": 2.912673234262815,
+      "grad_norm": 0.5265287526525314,
+      "learning_rate": 2.5507722548332446e-08,
+      "loss": 0.151,
+      "step": 8722
+    },
+    {
+      "epoch": 2.9130071798296875,
+      "grad_norm": 0.5179557694097449,
+      "learning_rate": 2.5312052032952505e-08,
+      "loss": 0.1517,
+      "step": 8723
+    },
+    {
+      "epoch": 2.91334112539656,
+      "grad_norm": 0.5090029191612081,
+      "learning_rate": 2.5117133002762196e-08,
+      "loss": 0.1455,
+      "step": 8724
+    },
+    {
+      "epoch": 2.913675070963433,
+      "grad_norm": 0.5325582522074706,
+      "learning_rate": 2.492296548720574e-08,
+      "loss": 0.1581,
+      "step": 8725
+    },
+    {
+      "epoch": 2.9140090165303056,
+      "grad_norm": 0.5281192943466827,
+      "learning_rate": 2.4729549515615235e-08,
+      "loss": 0.1561,
+      "step": 8726
+    },
+    {
+      "epoch": 2.9143429620971784,
+      "grad_norm": 0.5038006994679268,
+      "learning_rate": 2.453688511720842e-08,
+      "loss": 0.1418,
+      "step": 8727
+    },
+    {
+      "epoch": 2.9146769076640506,
+      "grad_norm": 0.5549235611646274,
+      "learning_rate": 2.4344972321089234e-08,
+      "loss": 0.1599,
+      "step": 8728
+    },
+    {
+      "epoch": 2.9150108532309233,
+      "grad_norm": 0.5302992496522636,
+      "learning_rate": 2.415381115624782e-08,
+      "loss": 0.1574,
+      "step": 8729
+    },
+    {
+      "epoch": 2.915344798797796,
+      "grad_norm": 0.5110893643807364,
+      "learning_rate": 2.3963401651562747e-08,
+      "loss": 0.1544,
+      "step": 8730
+    },
+    {
+      "epoch": 2.9156787443646683,
+      "grad_norm": 0.48901569574272535,
+      "learning_rate": 2.3773743835796558e-08,
+      "loss": 0.148,
+      "step": 8731
+    },
+    {
+      "epoch": 2.916012689931541,
+      "grad_norm": 0.5057738968119625,
+      "learning_rate": 2.358483773759912e-08,
+      "loss": 0.148,
+      "step": 8732
+    },
+    {
+      "epoch": 2.9163466354984138,
+      "grad_norm": 0.5454906371257616,
+      "learning_rate": 2.33966833855076e-08,
+      "loss": 0.1649,
+      "step": 8733
+    },
+    {
+      "epoch": 2.9166805810652865,
+      "grad_norm": 0.5538324525215251,
+      "learning_rate": 2.320928080794482e-08,
+      "loss": 0.1628,
+      "step": 8734
+    },
+    {
+      "epoch": 2.917014526632159,
+      "grad_norm": 0.5046291060510665,
+      "learning_rate": 2.3022630033219807e-08,
+      "loss": 0.1389,
+      "step": 8735
+    },
+    {
+      "epoch": 2.9173484721990315,
+      "grad_norm": 0.5542770135094989,
+      "learning_rate": 2.2836731089528886e-08,
+      "loss": 0.1655,
+      "step": 8736
+    },
+    {
+      "epoch": 2.917682417765904,
+      "grad_norm": 0.5699978406392153,
+      "learning_rate": 2.2651584004953485e-08,
+      "loss": 0.1642,
+      "step": 8737
+    },
+    {
+      "epoch": 2.918016363332777,
+      "grad_norm": 0.5293384260135789,
+      "learning_rate": 2.2467188807462902e-08,
+      "loss": 0.1564,
+      "step": 8738
+    },
+    {
+      "epoch": 2.918350308899649,
+      "grad_norm": 0.551306712890341,
+      "learning_rate": 2.2283545524912075e-08,
+      "loss": 0.1614,
+      "step": 8739
+    },
+    {
+      "epoch": 2.918684254466522,
+      "grad_norm": 0.5530682179532525,
+      "learning_rate": 2.210065418504215e-08,
+      "loss": 0.1504,
+      "step": 8740
+    },
+    {
+      "epoch": 2.9190182000333946,
+      "grad_norm": 0.5133652618900586,
+      "learning_rate": 2.1918514815481572e-08,
+      "loss": 0.1513,
+      "step": 8741
+    },
+    {
+      "epoch": 2.919352145600267,
+      "grad_norm": 0.48003189542915814,
+      "learning_rate": 2.17371274437439e-08,
+      "loss": 0.139,
+      "step": 8742
+    },
+    {
+      "epoch": 2.9196860911671396,
+      "grad_norm": 0.5691839116317102,
+      "learning_rate": 2.155649209723054e-08,
+      "loss": 0.1588,
+      "step": 8743
+    },
+    {
+      "epoch": 2.9200200367340123,
+      "grad_norm": 0.5663322722634725,
+      "learning_rate": 2.137660880322856e-08,
+      "loss": 0.1551,
+      "step": 8744
+    },
+    {
+      "epoch": 2.920353982300885,
+      "grad_norm": 0.522117785518442,
+      "learning_rate": 2.1197477588910666e-08,
+      "loss": 0.1534,
+      "step": 8745
+    },
+    {
+      "epoch": 2.9206879278677578,
+      "grad_norm": 0.5412269304337229,
+      "learning_rate": 2.101909848133743e-08,
+      "loss": 0.1605,
+      "step": 8746
+    },
+    {
+      "epoch": 2.92102187343463,
+      "grad_norm": 0.5590537588726184,
+      "learning_rate": 2.0841471507455635e-08,
+      "loss": 0.1613,
+      "step": 8747
+    },
+    {
+      "epoch": 2.9213558190015028,
+      "grad_norm": 0.5697624900762938,
+      "learning_rate": 2.0664596694096596e-08,
+      "loss": 0.1606,
+      "step": 8748
+    },
+    {
+      "epoch": 2.9216897645683755,
+      "grad_norm": 0.5509931109216466,
+      "learning_rate": 2.0488474067980045e-08,
+      "loss": 0.1608,
+      "step": 8749
+    },
+    {
+      "epoch": 2.9220237101352478,
+      "grad_norm": 0.5328571613716071,
+      "learning_rate": 2.0313103655711373e-08,
+      "loss": 0.1549,
+      "step": 8750
+    },
+    {
+      "epoch": 2.9223576557021205,
+      "grad_norm": 0.513781721288196,
+      "learning_rate": 2.0138485483782723e-08,
+      "loss": 0.1522,
+      "step": 8751
+    },
+    {
+      "epoch": 2.922691601268993,
+      "grad_norm": 0.5293720468326389,
+      "learning_rate": 1.996461957857132e-08,
+      "loss": 0.1501,
+      "step": 8752
+    },
+    {
+      "epoch": 2.923025546835866,
+      "grad_norm": 0.5633355764295512,
+      "learning_rate": 1.9791505966342273e-08,
+      "loss": 0.1624,
+      "step": 8753
+    },
+    {
+      "epoch": 2.9233594924027386,
+      "grad_norm": 0.5170586045028958,
+      "learning_rate": 1.9619144673246325e-08,
+      "loss": 0.1503,
+      "step": 8754
+    },
+    {
+      "epoch": 2.923693437969611,
+      "grad_norm": 0.5323221698384542,
+      "learning_rate": 1.9447535725320987e-08,
+      "loss": 0.1494,
+      "step": 8755
+    },
+    {
+      "epoch": 2.9240273835364836,
+      "grad_norm": 0.5356685252034754,
+      "learning_rate": 1.9276679148488854e-08,
+      "loss": 0.1478,
+      "step": 8756
+    },
+    {
+      "epoch": 2.9243613291033563,
+      "grad_norm": 0.5517955598586819,
+      "learning_rate": 1.9106574968560943e-08,
+      "loss": 0.1537,
+      "step": 8757
+    },
+    {
+      "epoch": 2.9246952746702286,
+      "grad_norm": 0.5070202914078584,
+      "learning_rate": 1.8937223211232257e-08,
+      "loss": 0.1442,
+      "step": 8758
+    },
+    {
+      "epoch": 2.9250292202371013,
+      "grad_norm": 0.5321126462836772,
+      "learning_rate": 1.876862390208678e-08,
+      "loss": 0.1546,
+      "step": 8759
+    },
+    {
+      "epoch": 2.925363165803974,
+      "grad_norm": 0.5484493215810594,
+      "learning_rate": 1.8600777066593023e-08,
+      "loss": 0.1554,
+      "step": 8760
+    },
+    {
+      "epoch": 2.9256971113708463,
+      "grad_norm": 0.5291482520966377,
+      "learning_rate": 1.8433682730105706e-08,
+      "loss": 0.1558,
+      "step": 8761
+    },
+    {
+      "epoch": 2.926031056937719,
+      "grad_norm": 0.5176849375689132,
+      "learning_rate": 1.8267340917866306e-08,
+      "loss": 0.1506,
+      "step": 8762
+    },
+    {
+      "epoch": 2.9263650025045918,
+      "grad_norm": 0.5205766280424573,
+      "learning_rate": 1.8101751655003053e-08,
+      "loss": 0.1494,
+      "step": 8763
+    },
+    {
+      "epoch": 2.9266989480714645,
+      "grad_norm": 0.5050130872096809,
+      "learning_rate": 1.793691496653094e-08,
+      "loss": 0.1546,
+      "step": 8764
+    },
+    {
+      "epoch": 2.927032893638337,
+      "grad_norm": 0.5509444172234437,
+      "learning_rate": 1.7772830877348933e-08,
+      "loss": 0.159,
+      "step": 8765
+    },
+    {
+      "epoch": 2.9273668392052095,
+      "grad_norm": 0.5237455644993183,
+      "learning_rate": 1.760949941224499e-08,
+      "loss": 0.1498,
+      "step": 8766
+    },
+    {
+      "epoch": 2.927700784772082,
+      "grad_norm": 0.5474546675198331,
+      "learning_rate": 1.7446920595892147e-08,
+      "loss": 0.1599,
+      "step": 8767
+    },
+    {
+      "epoch": 2.928034730338955,
+      "grad_norm": 0.5292033875636737,
+      "learning_rate": 1.7285094452849095e-08,
+      "loss": 0.1528,
+      "step": 8768
+    },
+    {
+      "epoch": 2.928368675905827,
+      "grad_norm": 0.5326258288778125,
+      "learning_rate": 1.7124021007562385e-08,
+      "loss": 0.1541,
+      "step": 8769
+    },
+    {
+      "epoch": 2.9287026214727,
+      "grad_norm": 0.505259236701369,
+      "learning_rate": 1.696370028436367e-08,
+      "loss": 0.1504,
+      "step": 8770
+    },
+    {
+      "epoch": 2.9290365670395726,
+      "grad_norm": 0.5295968720784965,
+      "learning_rate": 1.6804132307471354e-08,
+      "loss": 0.1554,
+      "step": 8771
+    },
+    {
+      "epoch": 2.929370512606445,
+      "grad_norm": 0.5855530439478099,
+      "learning_rate": 1.6645317100990044e-08,
+      "loss": 0.1655,
+      "step": 8772
+    },
+    {
+      "epoch": 2.9297044581733176,
+      "grad_norm": 0.5341122406061138,
+      "learning_rate": 1.6487254688910546e-08,
+      "loss": 0.1496,
+      "step": 8773
+    },
+    {
+      "epoch": 2.9300384037401903,
+      "grad_norm": 0.4966688117422215,
+      "learning_rate": 1.6329945095110435e-08,
+      "loss": 0.1437,
+      "step": 8774
+    },
+    {
+      "epoch": 2.930372349307063,
+      "grad_norm": 0.5169723744277109,
+      "learning_rate": 1.6173388343352915e-08,
+      "loss": 0.154,
+      "step": 8775
+    },
+    {
+      "epoch": 2.9307062948739357,
+      "grad_norm": 0.5287153064523404,
+      "learning_rate": 1.601758445728796e-08,
+      "loss": 0.146,
+      "step": 8776
+    },
+    {
+      "epoch": 2.931040240440808,
+      "grad_norm": 0.5645974728665406,
+      "learning_rate": 1.586253346045119e-08,
+      "loss": 0.1636,
+      "step": 8777
+    },
+    {
+      "epoch": 2.9313741860076807,
+      "grad_norm": 0.5405173803625365,
+      "learning_rate": 1.570823537626498e-08,
+      "loss": 0.1619,
+      "step": 8778
+    },
+    {
+      "epoch": 2.9317081315745535,
+      "grad_norm": 0.47878908052664143,
+      "learning_rate": 1.5554690228037905e-08,
+      "loss": 0.1496,
+      "step": 8779
+    },
+    {
+      "epoch": 2.9320420771414257,
+      "grad_norm": 0.5136540347637661,
+      "learning_rate": 1.5401898038964748e-08,
+      "loss": 0.1505,
+      "step": 8780
+    },
+    {
+      "epoch": 2.9323760227082984,
+      "grad_norm": 0.5066164231201058,
+      "learning_rate": 1.5249858832126486e-08,
+      "loss": 0.1515,
+      "step": 8781
+    },
+    {
+      "epoch": 2.932709968275171,
+      "grad_norm": 0.5494650012983797,
+      "learning_rate": 1.5098572630491414e-08,
+      "loss": 0.1557,
+      "step": 8782
+    },
+    {
+      "epoch": 2.933043913842044,
+      "grad_norm": 0.5356394443781805,
+      "learning_rate": 1.4948039456911256e-08,
+      "loss": 0.1598,
+      "step": 8783
+    },
+    {
+      "epoch": 2.9333778594089166,
+      "grad_norm": 0.579231974043971,
+      "learning_rate": 1.4798259334127263e-08,
+      "loss": 0.1622,
+      "step": 8784
+    },
+    {
+      "epoch": 2.933711804975789,
+      "grad_norm": 0.549857407176255,
+      "learning_rate": 1.4649232284765225e-08,
+      "loss": 0.1566,
+      "step": 8785
+    },
+    {
+      "epoch": 2.9340457505426616,
+      "grad_norm": 0.4940294824980054,
+      "learning_rate": 1.4500958331337134e-08,
+      "loss": 0.1479,
+      "step": 8786
+    },
+    {
+      "epoch": 2.9343796961095343,
+      "grad_norm": 0.5158826721565871,
+      "learning_rate": 1.435343749624174e-08,
+      "loss": 0.1515,
+      "step": 8787
+    },
+    {
+      "epoch": 2.9347136416764066,
+      "grad_norm": 0.5551253298998934,
+      "learning_rate": 1.420666980176344e-08,
+      "loss": 0.1695,
+      "step": 8788
+    },
+    {
+      "epoch": 2.9350475872432793,
+      "grad_norm": 0.5605107676992664,
+      "learning_rate": 1.4060655270073387e-08,
+      "loss": 0.1479,
+      "step": 8789
+    },
+    {
+      "epoch": 2.935381532810152,
+      "grad_norm": 0.535586571303752,
+      "learning_rate": 1.3915393923228936e-08,
+      "loss": 0.1575,
+      "step": 8790
+    },
+    {
+      "epoch": 2.9357154783770243,
+      "grad_norm": 0.4790608505081332,
+      "learning_rate": 1.3770885783173649e-08,
+      "loss": 0.1492,
+      "step": 8791
+    },
+    {
+      "epoch": 2.936049423943897,
+      "grad_norm": 0.5189679586079701,
+      "learning_rate": 1.3627130871737282e-08,
+      "loss": 0.1468,
+      "step": 8792
+    },
+    {
+      "epoch": 2.9363833695107697,
+      "grad_norm": 0.5060738164366134,
+      "learning_rate": 1.3484129210635243e-08,
+      "loss": 0.1486,
+      "step": 8793
+    },
+    {
+      "epoch": 2.9367173150776424,
+      "grad_norm": 0.5236931826293424,
+      "learning_rate": 1.3341880821469699e-08,
+      "loss": 0.1517,
+      "step": 8794
+    },
+    {
+      "epoch": 2.937051260644515,
+      "grad_norm": 0.5122034983251454,
+      "learning_rate": 1.3200385725729014e-08,
+      "loss": 0.146,
+      "step": 8795
+    },
+    {
+      "epoch": 2.9373852062113874,
+      "grad_norm": 0.5644397660384538,
+      "learning_rate": 1.3059643944787759e-08,
+      "loss": 0.1627,
+      "step": 8796
+    },
+    {
+      "epoch": 2.93771915177826,
+      "grad_norm": 0.5436746008349438,
+      "learning_rate": 1.2919655499906703e-08,
+      "loss": 0.1595,
+      "step": 8797
+    },
+    {
+      "epoch": 2.938053097345133,
+      "grad_norm": 0.5430037231902027,
+      "learning_rate": 1.2780420412232263e-08,
+      "loss": 0.1577,
+      "step": 8798
+    },
+    {
+      "epoch": 2.938387042912005,
+      "grad_norm": 0.5761697034083223,
+      "learning_rate": 1.2641938702798174e-08,
+      "loss": 0.1583,
+      "step": 8799
+    },
+    {
+      "epoch": 2.938720988478878,
+      "grad_norm": 0.47887126004228137,
+      "learning_rate": 1.2504210392523808e-08,
+      "loss": 0.1374,
+      "step": 8800
+    },
+    {
+      "epoch": 2.9390549340457506,
+      "grad_norm": 0.5759489993737833,
+      "learning_rate": 1.2367235502214192e-08,
+      "loss": 0.1603,
+      "step": 8801
+    },
+    {
+      "epoch": 2.9393888796126233,
+      "grad_norm": 0.5153572086231668,
+      "learning_rate": 1.2231014052560553e-08,
+      "loss": 0.1525,
+      "step": 8802
+    },
+    {
+      "epoch": 2.939722825179496,
+      "grad_norm": 0.518999090589087,
+      "learning_rate": 1.2095546064141982e-08,
+      "loss": 0.16,
+      "step": 8803
+    },
+    {
+      "epoch": 2.9400567707463683,
+      "grad_norm": 0.54913576945233,
+      "learning_rate": 1.196083155742156e-08,
+      "loss": 0.159,
+      "step": 8804
+    },
+    {
+      "epoch": 2.940390716313241,
+      "grad_norm": 0.5129445655764181,
+      "learning_rate": 1.1826870552749669e-08,
+      "loss": 0.1531,
+      "step": 8805
+    },
+    {
+      "epoch": 2.9407246618801137,
+      "grad_norm": 0.5302553484489443,
+      "learning_rate": 1.169366307036346e-08,
+      "loss": 0.1511,
+      "step": 8806
+    },
+    {
+      "epoch": 2.941058607446986,
+      "grad_norm": 0.5156122133349168,
+      "learning_rate": 1.1561209130384055e-08,
+      "loss": 0.145,
+      "step": 8807
+    },
+    {
+      "epoch": 2.9413925530138587,
+      "grad_norm": 0.5127356121469071,
+      "learning_rate": 1.1429508752821561e-08,
+      "loss": 0.1478,
+      "step": 8808
+    },
+    {
+      "epoch": 2.9417264985807314,
+      "grad_norm": 0.522416127983697,
+      "learning_rate": 1.1298561957570065e-08,
+      "loss": 0.156,
+      "step": 8809
+    },
+    {
+      "epoch": 2.9420604441476037,
+      "grad_norm": 0.5832906496004197,
+      "learning_rate": 1.1168368764410408e-08,
+      "loss": 0.1676,
+      "step": 8810
+    },
+    {
+      "epoch": 2.9423943897144764,
+      "grad_norm": 0.5065280786514194,
+      "learning_rate": 1.103892919301075e-08,
+      "loss": 0.1494,
+      "step": 8811
+    },
+    {
+      "epoch": 2.942728335281349,
+      "grad_norm": 0.4869535167280794,
+      "learning_rate": 1.0910243262923781e-08,
+      "loss": 0.146,
+      "step": 8812
+    },
+    {
+      "epoch": 2.943062280848222,
+      "grad_norm": 0.5054299642033462,
+      "learning_rate": 1.0782310993589506e-08,
+      "loss": 0.1485,
+      "step": 8813
+    },
+    {
+      "epoch": 2.9433962264150946,
+      "grad_norm": 0.48002626148076355,
+      "learning_rate": 1.0655132404333024e-08,
+      "loss": 0.1442,
+      "step": 8814
+    },
+    {
+      "epoch": 2.943730171981967,
+      "grad_norm": 0.5758905705157378,
+      "learning_rate": 1.0528707514366743e-08,
+      "loss": 0.168,
+      "step": 8815
+    },
+    {
+      "epoch": 2.9440641175488396,
+      "grad_norm": 0.5102727690819755,
+      "learning_rate": 1.0403036342787609e-08,
+      "loss": 0.1505,
+      "step": 8816
+    },
+    {
+      "epoch": 2.9443980631157123,
+      "grad_norm": 0.48725259548600885,
+      "learning_rate": 1.0278118908580992e-08,
+      "loss": 0.1423,
+      "step": 8817
+    },
+    {
+      "epoch": 2.9447320086825846,
+      "grad_norm": 0.5339536559315551,
+      "learning_rate": 1.0153955230616241e-08,
+      "loss": 0.1593,
+      "step": 8818
+    },
+    {
+      "epoch": 2.9450659542494573,
+      "grad_norm": 0.5337907365193634,
+      "learning_rate": 1.0030545327650576e-08,
+      "loss": 0.1554,
+      "step": 8819
+    },
+    {
+      "epoch": 2.94539989981633,
+      "grad_norm": 0.5371286184405106,
+      "learning_rate": 9.907889218325751e-09,
+      "loss": 0.162,
+      "step": 8820
+    },
+    {
+      "epoch": 2.9457338453832023,
+      "grad_norm": 0.5106944556580268,
+      "learning_rate": 9.78598692117083e-09,
+      "loss": 0.1527,
+      "step": 8821
+    },
+    {
+      "epoch": 2.946067790950075,
+      "grad_norm": 0.4891694550107267,
+      "learning_rate": 9.664838454599978e-09,
+      "loss": 0.1417,
+      "step": 8822
+    },
+    {
+      "epoch": 2.9464017365169477,
+      "grad_norm": 0.5067859025666626,
+      "learning_rate": 9.544443836914664e-09,
+      "loss": 0.1561,
+      "step": 8823
+    },
+    {
+      "epoch": 2.9467356820838204,
+      "grad_norm": 0.6033702392881961,
+      "learning_rate": 9.42480308630256e-09,
+      "loss": 0.1666,
+      "step": 8824
+    },
+    {
+      "epoch": 2.947069627650693,
+      "grad_norm": 0.5274076271656751,
+      "learning_rate": 9.30591622083532e-09,
+      "loss": 0.1581,
+      "step": 8825
+    },
+    {
+      "epoch": 2.9474035732175654,
+      "grad_norm": 0.6016450274515811,
+      "learning_rate": 9.187783258473027e-09,
+      "loss": 0.1543,
+      "step": 8826
+    },
+    {
+      "epoch": 2.947737518784438,
+      "grad_norm": 0.629833846802068,
+      "learning_rate": 9.070404217061402e-09,
+      "loss": 0.1559,
+      "step": 8827
+    },
+    {
+      "epoch": 2.948071464351311,
+      "grad_norm": 0.49603332101775693,
+      "learning_rate": 8.953779114331262e-09,
+      "loss": 0.1417,
+      "step": 8828
+    },
+    {
+      "epoch": 2.948405409918183,
+      "grad_norm": 0.5690481056431702,
+      "learning_rate": 8.837907967900183e-09,
+      "loss": 0.1615,
+      "step": 8829
+    },
+    {
+      "epoch": 2.948739355485056,
+      "grad_norm": 0.5286074793185255,
+      "learning_rate": 8.722790795272495e-09,
+      "loss": 0.1498,
+      "step": 8830
+    },
+    {
+      "epoch": 2.9490733010519286,
+      "grad_norm": 0.5394688875104848,
+      "learning_rate": 8.608427613837622e-09,
+      "loss": 0.1599,
+      "step": 8831
+    },
+    {
+      "epoch": 2.9494072466188013,
+      "grad_norm": 0.5232591082172079,
+      "learning_rate": 8.494818440871189e-09,
+      "loss": 0.1518,
+      "step": 8832
+    },
+    {
+      "epoch": 2.949741192185674,
+      "grad_norm": 0.5323106596009022,
+      "learning_rate": 8.381963293535577e-09,
+      "loss": 0.1489,
+      "step": 8833
+    },
+    {
+      "epoch": 2.9500751377525463,
+      "grad_norm": 0.5059971519222234,
+      "learning_rate": 8.269862188879374e-09,
+      "loss": 0.1531,
+      "step": 8834
+    },
+    {
+      "epoch": 2.950409083319419,
+      "grad_norm": 0.5337561349373425,
+      "learning_rate": 8.158515143835698e-09,
+      "loss": 0.1498,
+      "step": 8835
+    },
+    {
+      "epoch": 2.9507430288862917,
+      "grad_norm": 0.5203134409775217,
+      "learning_rate": 8.047922175225542e-09,
+      "loss": 0.1616,
+      "step": 8836
+    },
+    {
+      "epoch": 2.951076974453164,
+      "grad_norm": 0.5669691571198704,
+      "learning_rate": 7.938083299754984e-09,
+      "loss": 0.1622,
+      "step": 8837
+    },
+    {
+      "epoch": 2.9514109200200367,
+      "grad_norm": 0.5413633005491444,
+      "learning_rate": 7.828998534016308e-09,
+      "loss": 0.1566,
+      "step": 8838
+    },
+    {
+      "epoch": 2.9517448655869094,
+      "grad_norm": 0.5485356620391724,
+      "learning_rate": 7.720667894488554e-09,
+      "loss": 0.1603,
+      "step": 8839
+    },
+    {
+      "epoch": 2.9520788111537817,
+      "grad_norm": 0.5412801017017537,
+      "learning_rate": 7.613091397535855e-09,
+      "loss": 0.1529,
+      "step": 8840
+    },
+    {
+      "epoch": 2.9524127567206544,
+      "grad_norm": 0.5082343717572196,
+      "learning_rate": 7.506269059409654e-09,
+      "loss": 0.1489,
+      "step": 8841
+    },
+    {
+      "epoch": 2.952746702287527,
+      "grad_norm": 0.5188298301684736,
+      "learning_rate": 7.400200896245935e-09,
+      "loss": 0.1548,
+      "step": 8842
+    },
+    {
+      "epoch": 2.9530806478544,
+      "grad_norm": 0.5401849776800605,
+      "learning_rate": 7.29488692406799e-09,
+      "loss": 0.1637,
+      "step": 8843
+    },
+    {
+      "epoch": 2.9534145934212725,
+      "grad_norm": 0.6017819588969197,
+      "learning_rate": 7.190327158784205e-09,
+      "loss": 0.17,
+      "step": 8844
+    },
+    {
+      "epoch": 2.953748538988145,
+      "grad_norm": 0.5050764086292866,
+      "learning_rate": 7.0865216161902785e-09,
+      "loss": 0.1495,
+      "step": 8845
+    },
+    {
+      "epoch": 2.9540824845550175,
+      "grad_norm": 0.5397065334704347,
+      "learning_rate": 6.983470311967e-09,
+      "loss": 0.1542,
+      "step": 8846
+    },
+    {
+      "epoch": 2.9544164301218903,
+      "grad_norm": 0.49879861658122404,
+      "learning_rate": 6.881173261680807e-09,
+      "loss": 0.1403,
+      "step": 8847
+    },
+    {
+      "epoch": 2.9547503756887625,
+      "grad_norm": 0.5275264970743618,
+      "learning_rate": 6.779630480786004e-09,
+      "loss": 0.1495,
+      "step": 8848
+    },
+    {
+      "epoch": 2.9550843212556352,
+      "grad_norm": 0.5011084154145502,
+      "learning_rate": 6.678841984621432e-09,
+      "loss": 0.1522,
+      "step": 8849
+    },
+    {
+      "epoch": 2.955418266822508,
+      "grad_norm": 0.5681686799273573,
+      "learning_rate": 6.578807788411579e-09,
+      "loss": 0.1636,
+      "step": 8850
+    },
+    {
+      "epoch": 2.9557522123893807,
+      "grad_norm": 0.531060734134599,
+      "learning_rate": 6.479527907268801e-09,
+      "loss": 0.1514,
+      "step": 8851
+    },
+    {
+      "epoch": 2.9560861579562534,
+      "grad_norm": 0.5219698663221229,
+      "learning_rate": 6.381002356189991e-09,
+      "loss": 0.1504,
+      "step": 8852
+    },
+    {
+      "epoch": 2.9564201035231257,
+      "grad_norm": 0.5363223857696832,
+      "learning_rate": 6.283231150058799e-09,
+      "loss": 0.156,
+      "step": 8853
+    },
+    {
+      "epoch": 2.9567540490899984,
+      "grad_norm": 0.5360700439648506,
+      "learning_rate": 6.186214303645077e-09,
+      "loss": 0.1574,
+      "step": 8854
+    },
+    {
+      "epoch": 2.957087994656871,
+      "grad_norm": 0.5512505946034206,
+      "learning_rate": 6.0899518316032135e-09,
+      "loss": 0.1598,
+      "step": 8855
+    },
+    {
+      "epoch": 2.9574219402237434,
+      "grad_norm": 0.5071274370861918,
+      "learning_rate": 5.99444374847602e-09,
+      "loss": 0.1502,
+      "step": 8856
+    },
+    {
+      "epoch": 2.957755885790616,
+      "grad_norm": 0.5092307827434925,
+      "learning_rate": 5.899690068690289e-09,
+      "loss": 0.1435,
+      "step": 8857
+    },
+    {
+      "epoch": 2.958089831357489,
+      "grad_norm": 0.49280916083565207,
+      "learning_rate": 5.805690806560127e-09,
+      "loss": 0.1444,
+      "step": 8858
+    },
+    {
+      "epoch": 2.958423776924361,
+      "grad_norm": 0.5065398927260356,
+      "learning_rate": 5.712445976285286e-09,
+      "loss": 0.1511,
+      "step": 8859
+    },
+    {
+      "epoch": 2.958757722491234,
+      "grad_norm": 0.5279817308600407,
+      "learning_rate": 5.619955591951165e-09,
+      "loss": 0.1616,
+      "step": 8860
+    },
+    {
+      "epoch": 2.9590916680581065,
+      "grad_norm": 0.5624691691259425,
+      "learning_rate": 5.528219667529921e-09,
+      "loss": 0.1622,
+      "step": 8861
+    },
+    {
+      "epoch": 2.9594256136249792,
+      "grad_norm": 0.5183889213776066,
+      "learning_rate": 5.437238216878804e-09,
+      "loss": 0.1538,
+      "step": 8862
+    },
+    {
+      "epoch": 2.959759559191852,
+      "grad_norm": 0.5406835519360788,
+      "learning_rate": 5.347011253741819e-09,
+      "loss": 0.1587,
+      "step": 8863
+    },
+    {
+      "epoch": 2.9600935047587242,
+      "grad_norm": 0.49802577951748117,
+      "learning_rate": 5.257538791749173e-09,
+      "loss": 0.1481,
+      "step": 8864
+    },
+    {
+      "epoch": 2.960427450325597,
+      "grad_norm": 0.5261683688675758,
+      "learning_rate": 5.168820844416167e-09,
+      "loss": 0.1565,
+      "step": 8865
+    },
+    {
+      "epoch": 2.9607613958924697,
+      "grad_norm": 0.5114224521968415,
+      "learning_rate": 5.080857425145413e-09,
+      "loss": 0.1521,
+      "step": 8866
+    },
+    {
+      "epoch": 2.961095341459342,
+      "grad_norm": 0.5208384086955586,
+      "learning_rate": 4.993648547224062e-09,
+      "loss": 0.1551,
+      "step": 8867
+    },
+    {
+      "epoch": 2.9614292870262147,
+      "grad_norm": 0.5482296679225996,
+      "learning_rate": 4.907194223826572e-09,
+      "loss": 0.1543,
+      "step": 8868
+    },
+    {
+      "epoch": 2.9617632325930874,
+      "grad_norm": 0.5106259843517691,
+      "learning_rate": 4.8214944680125e-09,
+      "loss": 0.1499,
+      "step": 8869
+    },
+    {
+      "epoch": 2.9620971781599597,
+      "grad_norm": 0.5261301025915729,
+      "learning_rate": 4.736549292728154e-09,
+      "loss": 0.1586,
+      "step": 8870
+    },
+    {
+      "epoch": 2.9624311237268324,
+      "grad_norm": 0.5858625891292126,
+      "learning_rate": 4.652358710805494e-09,
+      "loss": 0.1639,
+      "step": 8871
+    },
+    {
+      "epoch": 2.962765069293705,
+      "grad_norm": 0.5259863655766891,
+      "learning_rate": 4.5689227349626775e-09,
+      "loss": 0.1551,
+      "step": 8872
+    },
+    {
+      "epoch": 2.963099014860578,
+      "grad_norm": 0.5277440908987092,
+      "learning_rate": 4.486241377802958e-09,
+      "loss": 0.1527,
+      "step": 8873
+    },
+    {
+      "epoch": 2.9634329604274505,
+      "grad_norm": 0.5230736899290115,
+      "learning_rate": 4.404314651816344e-09,
+      "loss": 0.1591,
+      "step": 8874
+    },
+    {
+      "epoch": 2.963766905994323,
+      "grad_norm": 0.5575061010473056,
+      "learning_rate": 4.323142569379602e-09,
+      "loss": 0.1604,
+      "step": 8875
+    },
+    {
+      "epoch": 2.9641008515611955,
+      "grad_norm": 0.5401053105604766,
+      "learning_rate": 4.242725142754589e-09,
+      "loss": 0.1464,
+      "step": 8876
+    },
+    {
+      "epoch": 2.9644347971280682,
+      "grad_norm": 0.5475423256849493,
+      "learning_rate": 4.163062384088812e-09,
+      "loss": 0.1624,
+      "step": 8877
+    },
+    {
+      "epoch": 2.9647687426949405,
+      "grad_norm": 0.5765524254562182,
+      "learning_rate": 4.0841543054165324e-09,
+      "loss": 0.1742,
+      "step": 8878
+    },
+    {
+      "epoch": 2.9651026882618132,
+      "grad_norm": 0.5045226252131164,
+      "learning_rate": 4.006000918658215e-09,
+      "loss": 0.1519,
+      "step": 8879
+    },
+    {
+      "epoch": 2.965436633828686,
+      "grad_norm": 0.5294171058153377,
+      "learning_rate": 3.928602235618861e-09,
+      "loss": 0.146,
+      "step": 8880
+    },
+    {
+      "epoch": 2.9657705793955587,
+      "grad_norm": 0.5557587243982196,
+      "learning_rate": 3.851958267990785e-09,
+      "loss": 0.16,
+      "step": 8881
+    },
+    {
+      "epoch": 2.9661045249624314,
+      "grad_norm": 0.516710441466049,
+      "learning_rate": 3.776069027352503e-09,
+      "loss": 0.1487,
+      "step": 8882
+    },
+    {
+      "epoch": 2.9664384705293037,
+      "grad_norm": 0.5615655633903573,
+      "learning_rate": 3.700934525167621e-09,
+      "loss": 0.1602,
+      "step": 8883
+    },
+    {
+      "epoch": 2.9667724160961764,
+      "grad_norm": 0.5047773838898094,
+      "learning_rate": 3.626554772786506e-09,
+      "loss": 0.1444,
+      "step": 8884
+    },
+    {
+      "epoch": 2.967106361663049,
+      "grad_norm": 0.515974634919327,
+      "learning_rate": 3.5529297814440587e-09,
+      "loss": 0.1498,
+      "step": 8885
+    },
+    {
+      "epoch": 2.9674403072299214,
+      "grad_norm": 0.536967546601452,
+      "learning_rate": 3.4800595622630497e-09,
+      "loss": 0.1507,
+      "step": 8886
+    },
+    {
+      "epoch": 2.967774252796794,
+      "grad_norm": 0.5415943444305699,
+      "learning_rate": 3.407944126251339e-09,
+      "loss": 0.1511,
+      "step": 8887
+    },
+    {
+      "epoch": 2.968108198363667,
+      "grad_norm": 0.5237385427929732,
+      "learning_rate": 3.336583484301881e-09,
+      "loss": 0.1463,
+      "step": 8888
+    },
+    {
+      "epoch": 2.968442143930539,
+      "grad_norm": 0.5448904790540473,
+      "learning_rate": 3.2659776471960505e-09,
+      "loss": 0.1659,
+      "step": 8889
+    },
+    {
+      "epoch": 2.968776089497412,
+      "grad_norm": 0.47703613603288947,
+      "learning_rate": 3.19612662559865e-09,
+      "loss": 0.1421,
+      "step": 8890
+    },
+    {
+      "epoch": 2.9691100350642845,
+      "grad_norm": 0.5685768307596133,
+      "learning_rate": 3.1270304300617947e-09,
+      "loss": 0.1746,
+      "step": 8891
+    },
+    {
+      "epoch": 2.969443980631157,
+      "grad_norm": 0.5260057299774317,
+      "learning_rate": 3.0586890710232465e-09,
+      "loss": 0.1579,
+      "step": 8892
+    },
+    {
+      "epoch": 2.96977792619803,
+      "grad_norm": 0.48821693615311545,
+      "learning_rate": 2.9911025588069685e-09,
+      "loss": 0.1446,
+      "step": 8893
+    },
+    {
+      "epoch": 2.970111871764902,
+      "grad_norm": 0.4848027115717547,
+      "learning_rate": 2.9242709036225723e-09,
+      "loss": 0.1406,
+      "step": 8894
+    },
+    {
+      "epoch": 2.970445817331775,
+      "grad_norm": 0.5188283228711802,
+      "learning_rate": 2.858194115565871e-09,
+      "loss": 0.151,
+      "step": 8895
+    },
+    {
+      "epoch": 2.9707797628986476,
+      "grad_norm": 0.5553633231710993,
+      "learning_rate": 2.7928722046177692e-09,
+      "loss": 0.1645,
+      "step": 8896
+    },
+    {
+      "epoch": 2.97111370846552,
+      "grad_norm": 0.510796810430822,
+      "learning_rate": 2.7283051806470394e-09,
+      "loss": 0.1477,
+      "step": 8897
+    },
+    {
+      "epoch": 2.9714476540323926,
+      "grad_norm": 0.565538056069778,
+      "learning_rate": 2.664493053406436e-09,
+      "loss": 0.1614,
+      "step": 8898
+    },
+    {
+      "epoch": 2.9717815995992654,
+      "grad_norm": 0.5381306694870884,
+      "learning_rate": 2.6014358325360256e-09,
+      "loss": 0.1585,
+      "step": 8899
+    },
+    {
+      "epoch": 2.972115545166138,
+      "grad_norm": 0.5257043175764273,
+      "learning_rate": 2.5391335275609665e-09,
+      "loss": 0.1528,
+      "step": 8900
+    },
+    {
+      "epoch": 2.972449490733011,
+      "grad_norm": 0.5353702583181583,
+      "learning_rate": 2.4775861478937293e-09,
+      "loss": 0.1523,
+      "step": 8901
+    },
+    {
+      "epoch": 2.972783436299883,
+      "grad_norm": 0.508839273196237,
+      "learning_rate": 2.416793702830211e-09,
+      "loss": 0.1445,
+      "step": 8902
+    },
+    {
+      "epoch": 2.973117381866756,
+      "grad_norm": 0.5087754840730636,
+      "learning_rate": 2.3567562015547328e-09,
+      "loss": 0.1427,
+      "step": 8903
+    },
+    {
+      "epoch": 2.9734513274336285,
+      "grad_norm": 0.524446864735963,
+      "learning_rate": 2.297473653136706e-09,
+      "loss": 0.1457,
+      "step": 8904
+    },
+    {
+      "epoch": 2.9737852730005008,
+      "grad_norm": 0.5317021238776584,
+      "learning_rate": 2.2389460665317443e-09,
+      "loss": 0.1547,
+      "step": 8905
+    },
+    {
+      "epoch": 2.9741192185673735,
+      "grad_norm": 0.5407645623936888,
+      "learning_rate": 2.1811734505799985e-09,
+      "loss": 0.1641,
+      "step": 8906
+    },
+    {
+      "epoch": 2.974453164134246,
+      "grad_norm": 0.5124507419550098,
+      "learning_rate": 2.1241558140100426e-09,
+      "loss": 0.1574,
+      "step": 8907
+    },
+    {
+      "epoch": 2.9747871097011185,
+      "grad_norm": 0.5441593837282322,
+      "learning_rate": 2.0678931654344314e-09,
+      "loss": 0.158,
+      "step": 8908
+    },
+    {
+      "epoch": 2.975121055267991,
+      "grad_norm": 0.5132818149239089,
+      "learning_rate": 2.012385513351922e-09,
+      "loss": 0.15,
+      "step": 8909
+    },
+    {
+      "epoch": 2.975455000834864,
+      "grad_norm": 0.5508960448044162,
+      "learning_rate": 1.9576328661480293e-09,
+      "loss": 0.1608,
+      "step": 8910
+    },
+    {
+      "epoch": 2.9757889464017366,
+      "grad_norm": 0.5478676465623783,
+      "learning_rate": 1.9036352320939146e-09,
+      "loss": 0.1563,
+      "step": 8911
+    },
+    {
+      "epoch": 2.9761228919686094,
+      "grad_norm": 0.5743289834297103,
+      "learning_rate": 1.850392619345831e-09,
+      "loss": 0.169,
+      "step": 8912
+    },
+    {
+      "epoch": 2.9764568375354816,
+      "grad_norm": 0.540526385238044,
+      "learning_rate": 1.7979050359479e-09,
+      "loss": 0.1572,
+      "step": 8913
+    },
+    {
+      "epoch": 2.9767907831023543,
+      "grad_norm": 0.5118515001947324,
+      "learning_rate": 1.746172489828224e-09,
+      "loss": 0.1496,
+      "step": 8914
+    },
+    {
+      "epoch": 2.977124728669227,
+      "grad_norm": 0.5295962054993424,
+      "learning_rate": 1.6951949888016627e-09,
+      "loss": 0.1488,
+      "step": 8915
+    },
+    {
+      "epoch": 2.9774586742360993,
+      "grad_norm": 0.5556596310017836,
+      "learning_rate": 1.6449725405687234e-09,
+      "loss": 0.1555,
+      "step": 8916
+    },
+    {
+      "epoch": 2.977792619802972,
+      "grad_norm": 0.568570152916496,
+      "learning_rate": 1.59550515271667e-09,
+      "loss": 0.1674,
+      "step": 8917
+    },
+    {
+      "epoch": 2.9781265653698448,
+      "grad_norm": 0.5121218556638442,
+      "learning_rate": 1.5467928327178582e-09,
+      "loss": 0.1512,
+      "step": 8918
+    },
+    {
+      "epoch": 2.978460510936717,
+      "grad_norm": 0.545092074861627,
+      "learning_rate": 1.498835587930847e-09,
+      "loss": 0.1579,
+      "step": 8919
+    },
+    {
+      "epoch": 2.9787944565035898,
+      "grad_norm": 0.5238858042064578,
+      "learning_rate": 1.4516334256003962e-09,
+      "loss": 0.1558,
+      "step": 8920
+    },
+    {
+      "epoch": 2.9791284020704625,
+      "grad_norm": 0.5058952627939988,
+      "learning_rate": 1.4051863528563581e-09,
+      "loss": 0.1447,
+      "step": 8921
+    },
+    {
+      "epoch": 2.979462347637335,
+      "grad_norm": 0.48109655383730054,
+      "learning_rate": 1.3594943767158974e-09,
+      "loss": 0.1414,
+      "step": 8922
+    },
+    {
+      "epoch": 2.979796293204208,
+      "grad_norm": 0.5352658964999676,
+      "learning_rate": 1.3145575040801605e-09,
+      "loss": 0.1466,
+      "step": 8923
+    },
+    {
+      "epoch": 2.98013023877108,
+      "grad_norm": 0.5336342161222689,
+      "learning_rate": 1.2703757417387164e-09,
+      "loss": 0.1514,
+      "step": 8924
+    },
+    {
+      "epoch": 2.980464184337953,
+      "grad_norm": 0.526133531148846,
+      "learning_rate": 1.2269490963651154e-09,
+      "loss": 0.154,
+      "step": 8925
+    },
+    {
+      "epoch": 2.9807981299048256,
+      "grad_norm": 0.5280976403546764,
+      "learning_rate": 1.1842775745196655e-09,
+      "loss": 0.1564,
+      "step": 8926
+    },
+    {
+      "epoch": 2.981132075471698,
+      "grad_norm": 0.49864255780159394,
+      "learning_rate": 1.1423611826477665e-09,
+      "loss": 0.153,
+      "step": 8927
+    },
+    {
+      "epoch": 2.9814660210385706,
+      "grad_norm": 0.4990574803432966,
+      "learning_rate": 1.1011999270821305e-09,
+      "loss": 0.1519,
+      "step": 8928
+    },
+    {
+      "epoch": 2.9817999666054433,
+      "grad_norm": 0.5206535132804454,
+      "learning_rate": 1.0607938140400064e-09,
+      "loss": 0.1481,
+      "step": 8929
+    },
+    {
+      "epoch": 2.982133912172316,
+      "grad_norm": 0.5203593281225992,
+      "learning_rate": 1.0211428496259557e-09,
+      "loss": 0.1585,
+      "step": 8930
+    },
+    {
+      "epoch": 2.9824678577391888,
+      "grad_norm": 0.511652837070169,
+      "learning_rate": 9.822470398296312e-10,
+      "loss": 0.1445,
+      "step": 8931
+    },
+    {
+      "epoch": 2.982801803306061,
+      "grad_norm": 0.5299622217401528,
+      "learning_rate": 9.441063905257785e-10,
+      "loss": 0.1523,
+      "step": 8932
+    },
+    {
+      "epoch": 2.9831357488729338,
+      "grad_norm": 0.5267298343073171,
+      "learning_rate": 9.067209074770101e-10,
+      "loss": 0.1556,
+      "step": 8933
+    },
+    {
+      "epoch": 2.9834696944398065,
+      "grad_norm": 0.5149660634700207,
+      "learning_rate": 8.700905963304751e-10,
+      "loss": 0.1552,
+      "step": 8934
+    },
+    {
+      "epoch": 2.9838036400066787,
+      "grad_norm": 0.6143707298607329,
+      "learning_rate": 8.342154626195254e-10,
+      "loss": 0.1573,
+      "step": 8935
+    },
+    {
+      "epoch": 2.9841375855735515,
+      "grad_norm": 0.5435314943549732,
+      "learning_rate": 7.990955117631594e-10,
+      "loss": 0.1528,
+      "step": 8936
+    },
+    {
+      "epoch": 2.984471531140424,
+      "grad_norm": 0.5484366140014967,
+      "learning_rate": 7.647307490676881e-10,
+      "loss": 0.1598,
+      "step": 8937
+    },
+    {
+      "epoch": 2.9848054767072965,
+      "grad_norm": 0.55852719855818,
+      "learning_rate": 7.311211797234041e-10,
+      "loss": 0.1551,
+      "step": 8938
+    },
+    {
+      "epoch": 2.985139422274169,
+      "grad_norm": 0.557832342106755,
+      "learning_rate": 6.982668088079126e-10,
+      "loss": 0.1673,
+      "step": 8939
+    },
+    {
+      "epoch": 2.985473367841042,
+      "grad_norm": 0.5086856936416371,
+      "learning_rate": 6.661676412844653e-10,
+      "loss": 0.1518,
+      "step": 8940
+    },
+    {
+      "epoch": 2.9858073134079146,
+      "grad_norm": 0.5437075245689668,
+      "learning_rate": 6.348236820008513e-10,
+      "loss": 0.1728,
+      "step": 8941
+    },
+    {
+      "epoch": 2.9861412589747873,
+      "grad_norm": 0.5162048454653639,
+      "learning_rate": 6.042349356932819e-10,
+      "loss": 0.1616,
+      "step": 8942
+    },
+    {
+      "epoch": 2.9864752045416596,
+      "grad_norm": 0.539657063914861,
+      "learning_rate": 5.744014069819503e-10,
+      "loss": 0.155,
+      "step": 8943
+    },
+    {
+      "epoch": 2.9868091501085323,
+      "grad_norm": 0.5276528948415139,
+      "learning_rate": 5.453231003732518e-10,
+      "loss": 0.1614,
+      "step": 8944
+    },
+    {
+      "epoch": 2.987143095675405,
+      "grad_norm": 0.5228415015085567,
+      "learning_rate": 5.170000202608938e-10,
+      "loss": 0.1488,
+      "step": 8945
+    },
+    {
+      "epoch": 2.9874770412422773,
+      "grad_norm": 0.5328035030755964,
+      "learning_rate": 4.894321709220106e-10,
+      "loss": 0.1496,
+      "step": 8946
+    },
+    {
+      "epoch": 2.98781098680915,
+      "grad_norm": 0.5211736551909211,
+      "learning_rate": 4.626195565221592e-10,
+      "loss": 0.156,
+      "step": 8947
+    },
+    {
+      "epoch": 2.9881449323760227,
+      "grad_norm": 0.5664070819139843,
+      "learning_rate": 4.365621811108778e-10,
+      "loss": 0.162,
+      "step": 8948
+    },
+    {
+      "epoch": 2.9884788779428955,
+      "grad_norm": 0.5453801039366839,
+      "learning_rate": 4.112600486250173e-10,
+      "loss": 0.1543,
+      "step": 8949
+    },
+    {
+      "epoch": 2.988812823509768,
+      "grad_norm": 0.5163009694570976,
+      "learning_rate": 3.867131628865206e-10,
+      "loss": 0.15,
+      "step": 8950
+    },
+    {
+      "epoch": 2.9891467690766405,
+      "grad_norm": 0.5144527921379415,
+      "learning_rate": 3.629215276035325e-10,
+      "loss": 0.1479,
+      "step": 8951
+    },
+    {
+      "epoch": 2.989480714643513,
+      "grad_norm": 0.4984509452806607,
+      "learning_rate": 3.3988514637040003e-10,
+      "loss": 0.1503,
+      "step": 8952
+    },
+    {
+      "epoch": 2.989814660210386,
+      "grad_norm": 0.530959054033469,
+      "learning_rate": 3.176040226660071e-10,
+      "loss": 0.1607,
+      "step": 8953
+    },
+    {
+      "epoch": 2.990148605777258,
+      "grad_norm": 0.5323475806559018,
+      "learning_rate": 2.960781598576601e-10,
+      "loss": 0.1558,
+      "step": 8954
+    },
+    {
+      "epoch": 2.990482551344131,
+      "grad_norm": 0.525452993656753,
+      "learning_rate": 2.7530756119609204e-10,
+      "loss": 0.1534,
+      "step": 8955
+    },
+    {
+      "epoch": 2.9908164969110036,
+      "grad_norm": 0.5141292907473098,
+      "learning_rate": 2.5529222981879323e-10,
+      "loss": 0.1516,
+      "step": 8956
+    },
+    {
+      "epoch": 2.991150442477876,
+      "grad_norm": 0.5160401121615004,
+      "learning_rate": 2.360321687500111e-10,
+      "loss": 0.1578,
+      "step": 8957
+    },
+    {
+      "epoch": 2.9914843880447486,
+      "grad_norm": 0.5226041005544119,
+      "learning_rate": 2.175273808985301e-10,
+      "loss": 0.1587,
+      "step": 8958
+    },
+    {
+      "epoch": 2.9918183336116213,
+      "grad_norm": 0.5481442045964605,
+      "learning_rate": 1.9977786906044683e-10,
+      "loss": 0.1589,
+      "step": 8959
+    },
+    {
+      "epoch": 2.992152279178494,
+      "grad_norm": 0.5130316124877334,
+      "learning_rate": 1.827836359163948e-10,
+      "loss": 0.1543,
+      "step": 8960
+    },
+    {
+      "epoch": 2.9924862247453667,
+      "grad_norm": 0.49218211497168046,
+      "learning_rate": 1.665446840343199e-10,
+      "loss": 0.1487,
+      "step": 8961
+    },
+    {
+      "epoch": 2.992820170312239,
+      "grad_norm": 0.5173019376331516,
+      "learning_rate": 1.5106101586614963e-10,
+      "loss": 0.1496,
+      "step": 8962
+    },
+    {
+      "epoch": 2.9931541158791117,
+      "grad_norm": 0.5519389841712395,
+      "learning_rate": 1.3633263375223414e-10,
+      "loss": 0.1646,
+      "step": 8963
+    },
+    {
+      "epoch": 2.9934880614459844,
+      "grad_norm": 0.49634411519978117,
+      "learning_rate": 1.223595399163502e-10,
+      "loss": 0.1522,
+      "step": 8964
+    },
+    {
+      "epoch": 2.9938220070128567,
+      "grad_norm": 0.5190963445321033,
+      "learning_rate": 1.091417364695868e-10,
+      "loss": 0.1518,
+      "step": 8965
+    },
+    {
+      "epoch": 2.9941559525797294,
+      "grad_norm": 0.5046318641135656,
+      "learning_rate": 9.667922540868013e-11,
+      "loss": 0.1506,
+      "step": 8966
+    },
+    {
+      "epoch": 2.994489898146602,
+      "grad_norm": 0.5811234054996668,
+      "learning_rate": 8.49720086165684e-11,
+      "loss": 0.1624,
+      "step": 8967
+    },
+    {
+      "epoch": 2.9948238437134744,
+      "grad_norm": 0.5187060796598542,
+      "learning_rate": 7.40200878618369e-11,
+      "loss": 0.1494,
+      "step": 8968
+    },
+    {
+      "epoch": 2.995157789280347,
+      "grad_norm": 0.5210939621866848,
+      "learning_rate": 6.382346479816282e-11,
+      "loss": 0.1562,
+      "step": 8969
+    },
+    {
+      "epoch": 2.99549173484722,
+      "grad_norm": 0.5147922038095906,
+      "learning_rate": 5.438214096653571e-11,
+      "loss": 0.1591,
+      "step": 8970
+    },
+    {
+      "epoch": 2.9958256804140926,
+      "grad_norm": 0.5091227792055766,
+      "learning_rate": 4.569611779248195e-11,
+      "loss": 0.1474,
+      "step": 8971
+    },
+    {
+      "epoch": 2.9961596259809653,
+      "grad_norm": 0.52462660718752,
+      "learning_rate": 3.776539658939538e-11,
+      "loss": 0.1587,
+      "step": 8972
+    },
+    {
+      "epoch": 2.9964935715478376,
+      "grad_norm": 0.5151458507703925,
+      "learning_rate": 3.0589978553541286e-11,
+      "loss": 0.1616,
+      "step": 8973
+    },
+    {
+      "epoch": 2.9968275171147103,
+      "grad_norm": 0.5064544199820976,
+      "learning_rate": 2.416986477071781e-11,
+      "loss": 0.1483,
+      "step": 8974
+    },
+    {
+      "epoch": 2.997161462681583,
+      "grad_norm": 0.5835195691906906,
+      "learning_rate": 1.850505620903942e-11,
+      "loss": 0.1629,
+      "step": 8975
+    },
+    {
+      "epoch": 2.9974954082484553,
+      "grad_norm": 0.5602503239016194,
+      "learning_rate": 1.3595553725598287e-11,
+      "loss": 0.1666,
+      "step": 8976
+    },
+    {
+      "epoch": 2.997829353815328,
+      "grad_norm": 0.5032549204664997,
+      "learning_rate": 9.441358061468286e-12,
+      "loss": 0.156,
+      "step": 8977
+    },
+    {
+      "epoch": 2.9981632993822007,
+      "grad_norm": 0.5091638685400584,
+      "learning_rate": 6.042469843925425e-12,
+      "loss": 0.1564,
+      "step": 8978
+    },
+    {
+      "epoch": 2.9984972449490734,
+      "grad_norm": 0.5291218098615191,
+      "learning_rate": 3.398889586447851e-12,
+      "loss": 0.1521,
+      "step": 8979
+    },
+    {
+      "epoch": 2.998831190515946,
+      "grad_norm": 0.5183009155912531,
+      "learning_rate": 1.5106176892709656e-12,
+      "loss": 0.1503,
+      "step": 8980
+    },
+    {
+      "epoch": 2.9991651360828184,
+      "grad_norm": 0.5279125117346501,
+      "learning_rate": 3.7765443661186283e-13,
+      "loss": 0.1586,
+      "step": 8981
+    },
+    {
+      "epoch": 2.999499081649691,
+      "grad_norm": 0.4923848154444246,
+      "learning_rate": 0.0,
+      "loss": 0.1483,
+      "step": 8982
+    },
+    {
+      "epoch": 2.999499081649691,
+      "eval_loss": 0.21598096191883087,
+      "eval_runtime": 183.8798,
+      "eval_samples_per_second": 109.708,
+      "eval_steps_per_second": 1.719,
+      "step": 8982
+    },
+    {
+      "epoch": 2.999499081649691,
+      "step": 8982,
+      "total_flos": 2.8091411711025807e+18,
+      "train_loss": 0.2020581066641609,
+      "train_runtime": 42907.438,
+      "train_samples_per_second": 26.797,
+      "train_steps_per_second": 0.209
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 8982,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "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": 2.8091411711025807e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}