diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,20195 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 2879,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0003473428273706148,
+      "grad_norm": 37.23415549580066,
+      "learning_rate": 1.1494252873563219e-07,
+      "loss": 0.1191,
+      "step": 1
+    },
+    {
+      "epoch": 0.0006946856547412296,
+      "grad_norm": 43.976862685963305,
+      "learning_rate": 2.2988505747126437e-07,
+      "loss": 0.1435,
+      "step": 2
+    },
+    {
+      "epoch": 0.0010420284821118443,
+      "grad_norm": 32.512151916312156,
+      "learning_rate": 3.4482758620689656e-07,
+      "loss": 0.1155,
+      "step": 3
+    },
+    {
+      "epoch": 0.0013893713094824591,
+      "grad_norm": 48.913784050446566,
+      "learning_rate": 4.5977011494252875e-07,
+      "loss": 0.1683,
+      "step": 4
+    },
+    {
+      "epoch": 0.0017367141368530739,
+      "grad_norm": 36.577121907693055,
+      "learning_rate": 5.747126436781609e-07,
+      "loss": 0.1352,
+      "step": 5
+    },
+    {
+      "epoch": 0.0020840569642236887,
+      "grad_norm": 26.561764879370827,
+      "learning_rate": 6.896551724137931e-07,
+      "loss": 0.0911,
+      "step": 6
+    },
+    {
+      "epoch": 0.0024313997915943034,
+      "grad_norm": 26.311858988645387,
+      "learning_rate": 8.045977011494253e-07,
+      "loss": 0.095,
+      "step": 7
+    },
+    {
+      "epoch": 0.0027787426189649182,
+      "grad_norm": 4.621058125611507,
+      "learning_rate": 9.195402298850575e-07,
+      "loss": 0.0487,
+      "step": 8
+    },
+    {
+      "epoch": 0.003126085446335533,
+      "grad_norm": 5.328829287414011,
+      "learning_rate": 1.0344827586206898e-06,
+      "loss": 0.0529,
+      "step": 9
+    },
+    {
+      "epoch": 0.0034734282737061478,
+      "grad_norm": 3.148318981902749,
+      "learning_rate": 1.1494252873563219e-06,
+      "loss": 0.0402,
+      "step": 10
+    },
+    {
+      "epoch": 0.0038207711010767626,
+      "grad_norm": 3.908627374831155,
+      "learning_rate": 1.2643678160919542e-06,
+      "loss": 0.0407,
+      "step": 11
+    },
+    {
+      "epoch": 0.004168113928447377,
+      "grad_norm": 4.739284231854802,
+      "learning_rate": 1.3793103448275862e-06,
+      "loss": 0.0446,
+      "step": 12
+    },
+    {
+      "epoch": 0.0045154567558179926,
+      "grad_norm": 2.3496827004932945,
+      "learning_rate": 1.4942528735632185e-06,
+      "loss": 0.0384,
+      "step": 13
+    },
+    {
+      "epoch": 0.004862799583188607,
+      "grad_norm": 5.475271188171554,
+      "learning_rate": 1.6091954022988506e-06,
+      "loss": 0.0485,
+      "step": 14
+    },
+    {
+      "epoch": 0.005210142410559222,
+      "grad_norm": 6.547877477275296,
+      "learning_rate": 1.724137931034483e-06,
+      "loss": 0.0568,
+      "step": 15
+    },
+    {
+      "epoch": 0.0055574852379298365,
+      "grad_norm": 3.8463210912421704,
+      "learning_rate": 1.839080459770115e-06,
+      "loss": 0.0449,
+      "step": 16
+    },
+    {
+      "epoch": 0.005904828065300452,
+      "grad_norm": 5.599877467384883,
+      "learning_rate": 1.9540229885057475e-06,
+      "loss": 0.0491,
+      "step": 17
+    },
+    {
+      "epoch": 0.006252170892671066,
+      "grad_norm": 4.194689420093038,
+      "learning_rate": 2.0689655172413796e-06,
+      "loss": 0.0473,
+      "step": 18
+    },
+    {
+      "epoch": 0.006599513720041681,
+      "grad_norm": 1.8031692743490404,
+      "learning_rate": 2.1839080459770117e-06,
+      "loss": 0.042,
+      "step": 19
+    },
+    {
+      "epoch": 0.0069468565474122956,
+      "grad_norm": 4.4715637058132,
+      "learning_rate": 2.2988505747126437e-06,
+      "loss": 0.0404,
+      "step": 20
+    },
+    {
+      "epoch": 0.007294199374782911,
+      "grad_norm": 0.8236040836386482,
+      "learning_rate": 2.4137931034482762e-06,
+      "loss": 0.0382,
+      "step": 21
+    },
+    {
+      "epoch": 0.007641542202153525,
+      "grad_norm": 0.82338294259807,
+      "learning_rate": 2.5287356321839083e-06,
+      "loss": 0.0422,
+      "step": 22
+    },
+    {
+      "epoch": 0.00798888502952414,
+      "grad_norm": 1.8027969557144028,
+      "learning_rate": 2.6436781609195404e-06,
+      "loss": 0.0363,
+      "step": 23
+    },
+    {
+      "epoch": 0.008336227856894755,
+      "grad_norm": 0.727095842359559,
+      "learning_rate": 2.7586206896551725e-06,
+      "loss": 0.0351,
+      "step": 24
+    },
+    {
+      "epoch": 0.00868357068426537,
+      "grad_norm": 1.92696781932363,
+      "learning_rate": 2.8735632183908046e-06,
+      "loss": 0.0301,
+      "step": 25
+    },
+    {
+      "epoch": 0.009030913511635985,
+      "grad_norm": 0.9300122101440347,
+      "learning_rate": 2.988505747126437e-06,
+      "loss": 0.0326,
+      "step": 26
+    },
+    {
+      "epoch": 0.0093782563390066,
+      "grad_norm": 5.1665169890422495,
+      "learning_rate": 3.103448275862069e-06,
+      "loss": 0.0378,
+      "step": 27
+    },
+    {
+      "epoch": 0.009725599166377214,
+      "grad_norm": 5.236812002541303,
+      "learning_rate": 3.2183908045977012e-06,
+      "loss": 0.0408,
+      "step": 28
+    },
+    {
+      "epoch": 0.010072941993747829,
+      "grad_norm": 2.6321912137329666,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.0315,
+      "step": 29
+    },
+    {
+      "epoch": 0.010420284821118444,
+      "grad_norm": 1.7260726850635042,
+      "learning_rate": 3.448275862068966e-06,
+      "loss": 0.03,
+      "step": 30
+    },
+    {
+      "epoch": 0.01076762764848906,
+      "grad_norm": 2.359583277352366,
+      "learning_rate": 3.563218390804598e-06,
+      "loss": 0.0293,
+      "step": 31
+    },
+    {
+      "epoch": 0.011114970475859673,
+      "grad_norm": 3.4439336912247382,
+      "learning_rate": 3.67816091954023e-06,
+      "loss": 0.0391,
+      "step": 32
+    },
+    {
+      "epoch": 0.011462313303230288,
+      "grad_norm": 2.5718166529004374,
+      "learning_rate": 3.793103448275862e-06,
+      "loss": 0.0405,
+      "step": 33
+    },
+    {
+      "epoch": 0.011809656130600903,
+      "grad_norm": 1.8743075423305569,
+      "learning_rate": 3.908045977011495e-06,
+      "loss": 0.0316,
+      "step": 34
+    },
+    {
+      "epoch": 0.012156998957971519,
+      "grad_norm": 0.9641265575376259,
+      "learning_rate": 4.022988505747127e-06,
+      "loss": 0.0282,
+      "step": 35
+    },
+    {
+      "epoch": 0.012504341785342132,
+      "grad_norm": 4.625856081652024,
+      "learning_rate": 4.137931034482759e-06,
+      "loss": 0.0336,
+      "step": 36
+    },
+    {
+      "epoch": 0.012851684612712747,
+      "grad_norm": 4.956158639154869,
+      "learning_rate": 4.252873563218391e-06,
+      "loss": 0.0535,
+      "step": 37
+    },
+    {
+      "epoch": 0.013199027440083362,
+      "grad_norm": 3.8237646578914033,
+      "learning_rate": 4.367816091954023e-06,
+      "loss": 0.0321,
+      "step": 38
+    },
+    {
+      "epoch": 0.013546370267453978,
+      "grad_norm": 1.6947960633729644,
+      "learning_rate": 4.482758620689656e-06,
+      "loss": 0.0348,
+      "step": 39
+    },
+    {
+      "epoch": 0.013893713094824591,
+      "grad_norm": 0.7680647499923681,
+      "learning_rate": 4.5977011494252875e-06,
+      "loss": 0.0318,
+      "step": 40
+    },
+    {
+      "epoch": 0.014241055922195206,
+      "grad_norm": 2.1664725717573545,
+      "learning_rate": 4.71264367816092e-06,
+      "loss": 0.0314,
+      "step": 41
+    },
+    {
+      "epoch": 0.014588398749565822,
+      "grad_norm": 4.263206749577008,
+      "learning_rate": 4.8275862068965525e-06,
+      "loss": 0.0407,
+      "step": 42
+    },
+    {
+      "epoch": 0.014935741576936437,
+      "grad_norm": 4.1724594615524335,
+      "learning_rate": 4.942528735632184e-06,
+      "loss": 0.0424,
+      "step": 43
+    },
+    {
+      "epoch": 0.01528308440430705,
+      "grad_norm": 3.7167659760688303,
+      "learning_rate": 5.057471264367817e-06,
+      "loss": 0.0345,
+      "step": 44
+    },
+    {
+      "epoch": 0.015630427231677665,
+      "grad_norm": 2.463474287564055,
+      "learning_rate": 5.172413793103449e-06,
+      "loss": 0.0349,
+      "step": 45
+    },
+    {
+      "epoch": 0.01597777005904828,
+      "grad_norm": 0.9621922490266032,
+      "learning_rate": 5.287356321839081e-06,
+      "loss": 0.0308,
+      "step": 46
+    },
+    {
+      "epoch": 0.016325112886418896,
+      "grad_norm": 1.1678035904473527,
+      "learning_rate": 5.402298850574713e-06,
+      "loss": 0.0255,
+      "step": 47
+    },
+    {
+      "epoch": 0.01667245571378951,
+      "grad_norm": 2.1879662040379664,
+      "learning_rate": 5.517241379310345e-06,
+      "loss": 0.0316,
+      "step": 48
+    },
+    {
+      "epoch": 0.017019798541160126,
+      "grad_norm": 3.5945900982319015,
+      "learning_rate": 5.6321839080459775e-06,
+      "loss": 0.042,
+      "step": 49
+    },
+    {
+      "epoch": 0.01736714136853074,
+      "grad_norm": 1.5756413045669015,
+      "learning_rate": 5.747126436781609e-06,
+      "loss": 0.0255,
+      "step": 50
+    },
+    {
+      "epoch": 0.017714484195901353,
+      "grad_norm": 1.2748577272738042,
+      "learning_rate": 5.862068965517242e-06,
+      "loss": 0.03,
+      "step": 51
+    },
+    {
+      "epoch": 0.01806182702327197,
+      "grad_norm": 3.009209224361777,
+      "learning_rate": 5.977011494252874e-06,
+      "loss": 0.0341,
+      "step": 52
+    },
+    {
+      "epoch": 0.018409169850642584,
+      "grad_norm": 3.489277616266429,
+      "learning_rate": 6.091954022988507e-06,
+      "loss": 0.0317,
+      "step": 53
+    },
+    {
+      "epoch": 0.0187565126780132,
+      "grad_norm": 3.962377854439755,
+      "learning_rate": 6.206896551724138e-06,
+      "loss": 0.0383,
+      "step": 54
+    },
+    {
+      "epoch": 0.019103855505383814,
+      "grad_norm": 2.162400758859317,
+      "learning_rate": 6.321839080459771e-06,
+      "loss": 0.0327,
+      "step": 55
+    },
+    {
+      "epoch": 0.019451198332754428,
+      "grad_norm": 0.9492153536245894,
+      "learning_rate": 6.4367816091954025e-06,
+      "loss": 0.0266,
+      "step": 56
+    },
+    {
+      "epoch": 0.019798541160125045,
+      "grad_norm": 1.0187860208671786,
+      "learning_rate": 6.551724137931035e-06,
+      "loss": 0.0354,
+      "step": 57
+    },
+    {
+      "epoch": 0.020145883987495658,
+      "grad_norm": 1.1083211600458822,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.0307,
+      "step": 58
+    },
+    {
+      "epoch": 0.02049322681486627,
+      "grad_norm": 3.937460604921549,
+      "learning_rate": 6.781609195402299e-06,
+      "loss": 0.0359,
+      "step": 59
+    },
+    {
+      "epoch": 0.02084056964223689,
+      "grad_norm": 0.6965391618701994,
+      "learning_rate": 6.896551724137932e-06,
+      "loss": 0.0296,
+      "step": 60
+    },
+    {
+      "epoch": 0.021187912469607502,
+      "grad_norm": 1.7430268237608375,
+      "learning_rate": 7.011494252873564e-06,
+      "loss": 0.033,
+      "step": 61
+    },
+    {
+      "epoch": 0.02153525529697812,
+      "grad_norm": 1.6253422710766354,
+      "learning_rate": 7.126436781609196e-06,
+      "loss": 0.026,
+      "step": 62
+    },
+    {
+      "epoch": 0.021882598124348732,
+      "grad_norm": 3.037222634408878,
+      "learning_rate": 7.241379310344828e-06,
+      "loss": 0.0349,
+      "step": 63
+    },
+    {
+      "epoch": 0.022229940951719346,
+      "grad_norm": 2.145372106566845,
+      "learning_rate": 7.35632183908046e-06,
+      "loss": 0.0294,
+      "step": 64
+    },
+    {
+      "epoch": 0.022577283779089963,
+      "grad_norm": 2.6342550049280384,
+      "learning_rate": 7.4712643678160925e-06,
+      "loss": 0.038,
+      "step": 65
+    },
+    {
+      "epoch": 0.022924626606460576,
+      "grad_norm": 1.336495145846901,
+      "learning_rate": 7.586206896551724e-06,
+      "loss": 0.037,
+      "step": 66
+    },
+    {
+      "epoch": 0.02327196943383119,
+      "grad_norm": 1.3878158762500856,
+      "learning_rate": 7.701149425287356e-06,
+      "loss": 0.036,
+      "step": 67
+    },
+    {
+      "epoch": 0.023619312261201807,
+      "grad_norm": 0.7502926647886281,
+      "learning_rate": 7.81609195402299e-06,
+      "loss": 0.027,
+      "step": 68
+    },
+    {
+      "epoch": 0.02396665508857242,
+      "grad_norm": 2.7789051553384114,
+      "learning_rate": 7.93103448275862e-06,
+      "loss": 0.0385,
+      "step": 69
+    },
+    {
+      "epoch": 0.024313997915943037,
+      "grad_norm": 1.510784689203471,
+      "learning_rate": 8.045977011494253e-06,
+      "loss": 0.0434,
+      "step": 70
+    },
+    {
+      "epoch": 0.02466134074331365,
+      "grad_norm": 1.795628251322906,
+      "learning_rate": 8.160919540229886e-06,
+      "loss": 0.0308,
+      "step": 71
+    },
+    {
+      "epoch": 0.025008683570684264,
+      "grad_norm": 0.8482257639162686,
+      "learning_rate": 8.275862068965518e-06,
+      "loss": 0.0354,
+      "step": 72
+    },
+    {
+      "epoch": 0.02535602639805488,
+      "grad_norm": 1.9996412936621009,
+      "learning_rate": 8.390804597701149e-06,
+      "loss": 0.0369,
+      "step": 73
+    },
+    {
+      "epoch": 0.025703369225425494,
+      "grad_norm": 0.7654423441044093,
+      "learning_rate": 8.505747126436782e-06,
+      "loss": 0.0353,
+      "step": 74
+    },
+    {
+      "epoch": 0.02605071205279611,
+      "grad_norm": 0.6493879425365747,
+      "learning_rate": 8.620689655172414e-06,
+      "loss": 0.0291,
+      "step": 75
+    },
+    {
+      "epoch": 0.026398054880166725,
+      "grad_norm": 2.414435804159643,
+      "learning_rate": 8.735632183908047e-06,
+      "loss": 0.0317,
+      "step": 76
+    },
+    {
+      "epoch": 0.02674539770753734,
+      "grad_norm": 1.0904156611035012,
+      "learning_rate": 8.85057471264368e-06,
+      "loss": 0.0269,
+      "step": 77
+    },
+    {
+      "epoch": 0.027092740534907955,
+      "grad_norm": 1.4189653573098011,
+      "learning_rate": 8.965517241379312e-06,
+      "loss": 0.0236,
+      "step": 78
+    },
+    {
+      "epoch": 0.02744008336227857,
+      "grad_norm": 1.040668552643942,
+      "learning_rate": 9.080459770114942e-06,
+      "loss": 0.0243,
+      "step": 79
+    },
+    {
+      "epoch": 0.027787426189649182,
+      "grad_norm": 1.3437174808420649,
+      "learning_rate": 9.195402298850575e-06,
+      "loss": 0.0332,
+      "step": 80
+    },
+    {
+      "epoch": 0.0281347690170198,
+      "grad_norm": 1.4299130307272268,
+      "learning_rate": 9.310344827586207e-06,
+      "loss": 0.0308,
+      "step": 81
+    },
+    {
+      "epoch": 0.028482111844390413,
+      "grad_norm": 1.2629695612963914,
+      "learning_rate": 9.42528735632184e-06,
+      "loss": 0.0203,
+      "step": 82
+    },
+    {
+      "epoch": 0.02882945467176103,
+      "grad_norm": 1.3219643843083144,
+      "learning_rate": 9.54022988505747e-06,
+      "loss": 0.0379,
+      "step": 83
+    },
+    {
+      "epoch": 0.029176797499131643,
+      "grad_norm": 3.6871466385779783,
+      "learning_rate": 9.655172413793105e-06,
+      "loss": 0.0274,
+      "step": 84
+    },
+    {
+      "epoch": 0.029524140326502257,
+      "grad_norm": 3.9969862428228033,
+      "learning_rate": 9.770114942528738e-06,
+      "loss": 0.0323,
+      "step": 85
+    },
+    {
+      "epoch": 0.029871483153872874,
+      "grad_norm": 3.065195993982362,
+      "learning_rate": 9.885057471264368e-06,
+      "loss": 0.0309,
+      "step": 86
+    },
+    {
+      "epoch": 0.030218825981243487,
+      "grad_norm": 1.468135218974371,
+      "learning_rate": 1e-05,
+      "loss": 0.0263,
+      "step": 87
+    },
+    {
+      "epoch": 0.0305661688086141,
+      "grad_norm": 2.1306295097585353,
+      "learning_rate": 9.999996834743712e-06,
+      "loss": 0.0366,
+      "step": 88
+    },
+    {
+      "epoch": 0.030913511635984717,
+      "grad_norm": 4.561675020890797,
+      "learning_rate": 9.999987338978852e-06,
+      "loss": 0.0399,
+      "step": 89
+    },
+    {
+      "epoch": 0.03126085446335533,
+      "grad_norm": 3.0699739972744355,
+      "learning_rate": 9.999971512717445e-06,
+      "loss": 0.0361,
+      "step": 90
+    },
+    {
+      "epoch": 0.03160819729072595,
+      "grad_norm": 1.7204154640977625,
+      "learning_rate": 9.99994935597953e-06,
+      "loss": 0.023,
+      "step": 91
+    },
+    {
+      "epoch": 0.03195554011809656,
+      "grad_norm": 0.4121394943188734,
+      "learning_rate": 9.999920868793156e-06,
+      "loss": 0.0151,
+      "step": 92
+    },
+    {
+      "epoch": 0.032302882945467175,
+      "grad_norm": 1.8037684571162973,
+      "learning_rate": 9.999886051194392e-06,
+      "loss": 0.041,
+      "step": 93
+    },
+    {
+      "epoch": 0.03265022577283779,
+      "grad_norm": 1.5191359679839431,
+      "learning_rate": 9.999844903227323e-06,
+      "loss": 0.0222,
+      "step": 94
+    },
+    {
+      "epoch": 0.03299756860020841,
+      "grad_norm": 2.797587288217291,
+      "learning_rate": 9.999797424944041e-06,
+      "loss": 0.0359,
+      "step": 95
+    },
+    {
+      "epoch": 0.03334491142757902,
+      "grad_norm": 0.9848390023335929,
+      "learning_rate": 9.999743616404667e-06,
+      "loss": 0.029,
+      "step": 96
+    },
+    {
+      "epoch": 0.033692254254949636,
+      "grad_norm": 3.474922410115143,
+      "learning_rate": 9.999683477677319e-06,
+      "loss": 0.0433,
+      "step": 97
+    },
+    {
+      "epoch": 0.03403959708232025,
+      "grad_norm": 0.4352894693833356,
+      "learning_rate": 9.999617008838145e-06,
+      "loss": 0.0271,
+      "step": 98
+    },
+    {
+      "epoch": 0.03438693990969086,
+      "grad_norm": 1.0612949281619894,
+      "learning_rate": 9.999544209971299e-06,
+      "loss": 0.0262,
+      "step": 99
+    },
+    {
+      "epoch": 0.03473428273706148,
+      "grad_norm": 1.5028120141315706,
+      "learning_rate": 9.999465081168954e-06,
+      "loss": 0.023,
+      "step": 100
+    },
+    {
+      "epoch": 0.035081625564432097,
+      "grad_norm": 2.485311227250491,
+      "learning_rate": 9.999379622531292e-06,
+      "loss": 0.0353,
+      "step": 101
+    },
+    {
+      "epoch": 0.035428968391802707,
+      "grad_norm": 1.1277504855781095,
+      "learning_rate": 9.99928783416651e-06,
+      "loss": 0.0238,
+      "step": 102
+    },
+    {
+      "epoch": 0.03577631121917332,
+      "grad_norm": 0.9476524869349918,
+      "learning_rate": 9.99918971619083e-06,
+      "loss": 0.0311,
+      "step": 103
+    },
+    {
+      "epoch": 0.03612365404654394,
+      "grad_norm": 4.351712281867559,
+      "learning_rate": 9.999085268728473e-06,
+      "loss": 0.0371,
+      "step": 104
+    },
+    {
+      "epoch": 0.03647099687391455,
+      "grad_norm": 2.7348253113812424,
+      "learning_rate": 9.998974491911681e-06,
+      "loss": 0.0374,
+      "step": 105
+    },
+    {
+      "epoch": 0.03681833970128517,
+      "grad_norm": 2.5995603090920105,
+      "learning_rate": 9.998857385880712e-06,
+      "loss": 0.0285,
+      "step": 106
+    },
+    {
+      "epoch": 0.037165682528655784,
+      "grad_norm": 0.7311687318983651,
+      "learning_rate": 9.99873395078383e-06,
+      "loss": 0.0229,
+      "step": 107
+    },
+    {
+      "epoch": 0.0375130253560264,
+      "grad_norm": 1.2353487773193577,
+      "learning_rate": 9.998604186777318e-06,
+      "loss": 0.0322,
+      "step": 108
+    },
+    {
+      "epoch": 0.03786036818339701,
+      "grad_norm": 0.9451422134075631,
+      "learning_rate": 9.998468094025473e-06,
+      "loss": 0.0352,
+      "step": 109
+    },
+    {
+      "epoch": 0.03820771101076763,
+      "grad_norm": 1.7091841651608024,
+      "learning_rate": 9.9983256727006e-06,
+      "loss": 0.0344,
+      "step": 110
+    },
+    {
+      "epoch": 0.038555053838138245,
+      "grad_norm": 2.7065412298038387,
+      "learning_rate": 9.998176922983017e-06,
+      "loss": 0.0356,
+      "step": 111
+    },
+    {
+      "epoch": 0.038902396665508855,
+      "grad_norm": 0.7915485766007472,
+      "learning_rate": 9.998021845061059e-06,
+      "loss": 0.0278,
+      "step": 112
+    },
+    {
+      "epoch": 0.03924973949287947,
+      "grad_norm": 0.6604452686175176,
+      "learning_rate": 9.99786043913107e-06,
+      "loss": 0.0263,
+      "step": 113
+    },
+    {
+      "epoch": 0.03959708232025009,
+      "grad_norm": 2.2075871866436807,
+      "learning_rate": 9.997692705397408e-06,
+      "loss": 0.0293,
+      "step": 114
+    },
+    {
+      "epoch": 0.0399444251476207,
+      "grad_norm": 1.0990740259335516,
+      "learning_rate": 9.99751864407244e-06,
+      "loss": 0.0251,
+      "step": 115
+    },
+    {
+      "epoch": 0.040291767974991316,
+      "grad_norm": 1.1686977294017693,
+      "learning_rate": 9.997338255376545e-06,
+      "loss": 0.0313,
+      "step": 116
+    },
+    {
+      "epoch": 0.04063911080236193,
+      "grad_norm": 1.3383756136679479,
+      "learning_rate": 9.997151539538114e-06,
+      "loss": 0.0226,
+      "step": 117
+    },
+    {
+      "epoch": 0.04098645362973254,
+      "grad_norm": 0.7694885404915951,
+      "learning_rate": 9.996958496793547e-06,
+      "loss": 0.0301,
+      "step": 118
+    },
+    {
+      "epoch": 0.04133379645710316,
+      "grad_norm": 1.0594137346894543,
+      "learning_rate": 9.996759127387259e-06,
+      "loss": 0.0217,
+      "step": 119
+    },
+    {
+      "epoch": 0.04168113928447378,
+      "grad_norm": 0.7335920204997873,
+      "learning_rate": 9.996553431571669e-06,
+      "loss": 0.0225,
+      "step": 120
+    },
+    {
+      "epoch": 0.042028482111844394,
+      "grad_norm": 2.255099961830129,
+      "learning_rate": 9.99634140960721e-06,
+      "loss": 0.0248,
+      "step": 121
+    },
+    {
+      "epoch": 0.042375824939215004,
+      "grad_norm": 2.42490870668873,
+      "learning_rate": 9.996123061762324e-06,
+      "loss": 0.0424,
+      "step": 122
+    },
+    {
+      "epoch": 0.04272316776658562,
+      "grad_norm": 1.1936924708308256,
+      "learning_rate": 9.99589838831346e-06,
+      "loss": 0.036,
+      "step": 123
+    },
+    {
+      "epoch": 0.04307051059395624,
+      "grad_norm": 0.9868005673494291,
+      "learning_rate": 9.995667389545082e-06,
+      "loss": 0.0418,
+      "step": 124
+    },
+    {
+      "epoch": 0.04341785342132685,
+      "grad_norm": 2.506803561409087,
+      "learning_rate": 9.995430065749653e-06,
+      "loss": 0.0287,
+      "step": 125
+    },
+    {
+      "epoch": 0.043765196248697465,
+      "grad_norm": 0.7525519955237748,
+      "learning_rate": 9.995186417227654e-06,
+      "loss": 0.0295,
+      "step": 126
+    },
+    {
+      "epoch": 0.04411253907606808,
+      "grad_norm": 0.3504040791938595,
+      "learning_rate": 9.994936444287565e-06,
+      "loss": 0.0317,
+      "step": 127
+    },
+    {
+      "epoch": 0.04445988190343869,
+      "grad_norm": 1.0108416374201756,
+      "learning_rate": 9.99468014724588e-06,
+      "loss": 0.0281,
+      "step": 128
+    },
+    {
+      "epoch": 0.04480722473080931,
+      "grad_norm": 2.211821690034367,
+      "learning_rate": 9.994417526427094e-06,
+      "loss": 0.0351,
+      "step": 129
+    },
+    {
+      "epoch": 0.045154567558179926,
+      "grad_norm": 0.4010376448499792,
+      "learning_rate": 9.994148582163715e-06,
+      "loss": 0.0278,
+      "step": 130
+    },
+    {
+      "epoch": 0.045501910385550536,
+      "grad_norm": 1.1919924569133369,
+      "learning_rate": 9.993873314796253e-06,
+      "loss": 0.028,
+      "step": 131
+    },
+    {
+      "epoch": 0.04584925321292115,
+      "grad_norm": 1.2102441655899068,
+      "learning_rate": 9.993591724673225e-06,
+      "loss": 0.0272,
+      "step": 132
+    },
+    {
+      "epoch": 0.04619659604029177,
+      "grad_norm": 1.3423201041149735,
+      "learning_rate": 9.993303812151153e-06,
+      "loss": 0.0288,
+      "step": 133
+    },
+    {
+      "epoch": 0.04654393886766238,
+      "grad_norm": 0.9911420725512745,
+      "learning_rate": 9.993009577594564e-06,
+      "loss": 0.0266,
+      "step": 134
+    },
+    {
+      "epoch": 0.046891281695032996,
+      "grad_norm": 2.818665002325742,
+      "learning_rate": 9.992709021375987e-06,
+      "loss": 0.0416,
+      "step": 135
+    },
+    {
+      "epoch": 0.04723862452240361,
+      "grad_norm": 0.6400736482581679,
+      "learning_rate": 9.99240214387596e-06,
+      "loss": 0.0224,
+      "step": 136
+    },
+    {
+      "epoch": 0.04758596734977423,
+      "grad_norm": 1.564814253959133,
+      "learning_rate": 9.99208894548302e-06,
+      "loss": 0.0326,
+      "step": 137
+    },
+    {
+      "epoch": 0.04793331017714484,
+      "grad_norm": 1.1019040221230227,
+      "learning_rate": 9.991769426593707e-06,
+      "loss": 0.0327,
+      "step": 138
+    },
+    {
+      "epoch": 0.04828065300451546,
+      "grad_norm": 6.120623936551482,
+      "learning_rate": 9.991443587612568e-06,
+      "loss": 0.047,
+      "step": 139
+    },
+    {
+      "epoch": 0.048627995831886074,
+      "grad_norm": 5.451375099863766,
+      "learning_rate": 9.991111428952145e-06,
+      "loss": 0.0503,
+      "step": 140
+    },
+    {
+      "epoch": 0.048975338659256684,
+      "grad_norm": 4.7872510231651635,
+      "learning_rate": 9.990772951032987e-06,
+      "loss": 0.0417,
+      "step": 141
+    },
+    {
+      "epoch": 0.0493226814866273,
+      "grad_norm": 3.9394554371100337,
+      "learning_rate": 9.990428154283641e-06,
+      "loss": 0.0341,
+      "step": 142
+    },
+    {
+      "epoch": 0.04967002431399792,
+      "grad_norm": 0.8850048672337401,
+      "learning_rate": 9.990077039140655e-06,
+      "loss": 0.0319,
+      "step": 143
+    },
+    {
+      "epoch": 0.05001736714136853,
+      "grad_norm": 0.735313738269197,
+      "learning_rate": 9.989719606048578e-06,
+      "loss": 0.03,
+      "step": 144
+    },
+    {
+      "epoch": 0.050364709968739145,
+      "grad_norm": 2.380846224871553,
+      "learning_rate": 9.989355855459954e-06,
+      "loss": 0.033,
+      "step": 145
+    },
+    {
+      "epoch": 0.05071205279610976,
+      "grad_norm": 1.5964417699444167,
+      "learning_rate": 9.988985787835332e-06,
+      "loss": 0.0297,
+      "step": 146
+    },
+    {
+      "epoch": 0.05105939562348037,
+      "grad_norm": 2.2693926915330067,
+      "learning_rate": 9.988609403643254e-06,
+      "loss": 0.0355,
+      "step": 147
+    },
+    {
+      "epoch": 0.05140673845085099,
+      "grad_norm": 0.8694077145197386,
+      "learning_rate": 9.98822670336026e-06,
+      "loss": 0.0257,
+      "step": 148
+    },
+    {
+      "epoch": 0.051754081278221606,
+      "grad_norm": 2.6184498491631687,
+      "learning_rate": 9.987837687470889e-06,
+      "loss": 0.0349,
+      "step": 149
+    },
+    {
+      "epoch": 0.05210142410559222,
+      "grad_norm": 1.8812135552729436,
+      "learning_rate": 9.987442356467677e-06,
+      "loss": 0.0205,
+      "step": 150
+    },
+    {
+      "epoch": 0.05244876693296283,
+      "grad_norm": 0.757853013334225,
+      "learning_rate": 9.987040710851148e-06,
+      "loss": 0.0361,
+      "step": 151
+    },
+    {
+      "epoch": 0.05279610976033345,
+      "grad_norm": 0.9951588342302143,
+      "learning_rate": 9.98663275112983e-06,
+      "loss": 0.0237,
+      "step": 152
+    },
+    {
+      "epoch": 0.05314345258770407,
+      "grad_norm": 1.6932805634155623,
+      "learning_rate": 9.986218477820244e-06,
+      "loss": 0.0368,
+      "step": 153
+    },
+    {
+      "epoch": 0.05349079541507468,
+      "grad_norm": 1.3139957494156915,
+      "learning_rate": 9.985797891446898e-06,
+      "loss": 0.0229,
+      "step": 154
+    },
+    {
+      "epoch": 0.053838138242445294,
+      "grad_norm": 0.5291454593174553,
+      "learning_rate": 9.9853709925423e-06,
+      "loss": 0.0228,
+      "step": 155
+    },
+    {
+      "epoch": 0.05418548106981591,
+      "grad_norm": 1.6707077604660217,
+      "learning_rate": 9.984937781646948e-06,
+      "loss": 0.03,
+      "step": 156
+    },
+    {
+      "epoch": 0.05453282389718652,
+      "grad_norm": 3.0726902017901927,
+      "learning_rate": 9.984498259309332e-06,
+      "loss": 0.0334,
+      "step": 157
+    },
+    {
+      "epoch": 0.05488016672455714,
+      "grad_norm": 0.7820818072933668,
+      "learning_rate": 9.984052426085931e-06,
+      "loss": 0.0205,
+      "step": 158
+    },
+    {
+      "epoch": 0.055227509551927755,
+      "grad_norm": 1.5369790153168252,
+      "learning_rate": 9.983600282541213e-06,
+      "loss": 0.033,
+      "step": 159
+    },
+    {
+      "epoch": 0.055574852379298365,
+      "grad_norm": 0.6130697531268696,
+      "learning_rate": 9.983141829247644e-06,
+      "loss": 0.0323,
+      "step": 160
+    },
+    {
+      "epoch": 0.05592219520666898,
+      "grad_norm": 2.0042679704987245,
+      "learning_rate": 9.982677066785667e-06,
+      "loss": 0.0362,
+      "step": 161
+    },
+    {
+      "epoch": 0.0562695380340396,
+      "grad_norm": 0.9373286974868164,
+      "learning_rate": 9.982205995743723e-06,
+      "loss": 0.027,
+      "step": 162
+    },
+    {
+      "epoch": 0.05661688086141021,
+      "grad_norm": 0.9593697401680631,
+      "learning_rate": 9.981728616718234e-06,
+      "loss": 0.0221,
+      "step": 163
+    },
+    {
+      "epoch": 0.056964223688780825,
+      "grad_norm": 0.9145114104321098,
+      "learning_rate": 9.981244930313613e-06,
+      "loss": 0.0269,
+      "step": 164
+    },
+    {
+      "epoch": 0.05731156651615144,
+      "grad_norm": 1.7022933635129043,
+      "learning_rate": 9.980754937142254e-06,
+      "loss": 0.0273,
+      "step": 165
+    },
+    {
+      "epoch": 0.05765890934352206,
+      "grad_norm": 0.5802100036143374,
+      "learning_rate": 9.98025863782454e-06,
+      "loss": 0.0197,
+      "step": 166
+    },
+    {
+      "epoch": 0.05800625217089267,
+      "grad_norm": 2.06874906530981,
+      "learning_rate": 9.979756032988837e-06,
+      "loss": 0.0292,
+      "step": 167
+    },
+    {
+      "epoch": 0.058353594998263286,
+      "grad_norm": 2.9290249045005017,
+      "learning_rate": 9.979247123271494e-06,
+      "loss": 0.0338,
+      "step": 168
+    },
+    {
+      "epoch": 0.0587009378256339,
+      "grad_norm": 0.8386866195629475,
+      "learning_rate": 9.978731909316841e-06,
+      "loss": 0.0243,
+      "step": 169
+    },
+    {
+      "epoch": 0.05904828065300451,
+      "grad_norm": 0.4502042472741338,
+      "learning_rate": 9.978210391777195e-06,
+      "loss": 0.0255,
+      "step": 170
+    },
+    {
+      "epoch": 0.05939562348037513,
+      "grad_norm": 2.248415124487682,
+      "learning_rate": 9.977682571312847e-06,
+      "loss": 0.0336,
+      "step": 171
+    },
+    {
+      "epoch": 0.05974296630774575,
+      "grad_norm": 1.0652482256976974,
+      "learning_rate": 9.977148448592077e-06,
+      "loss": 0.0224,
+      "step": 172
+    },
+    {
+      "epoch": 0.06009030913511636,
+      "grad_norm": 0.7329238551144143,
+      "learning_rate": 9.976608024291135e-06,
+      "loss": 0.0266,
+      "step": 173
+    },
+    {
+      "epoch": 0.060437651962486974,
+      "grad_norm": 0.8741338102260089,
+      "learning_rate": 9.976061299094253e-06,
+      "loss": 0.0219,
+      "step": 174
+    },
+    {
+      "epoch": 0.06078499478985759,
+      "grad_norm": 1.6779119248256096,
+      "learning_rate": 9.975508273693643e-06,
+      "loss": 0.0371,
+      "step": 175
+    },
+    {
+      "epoch": 0.0611323376172282,
+      "grad_norm": 0.7059586695874602,
+      "learning_rate": 9.974948948789492e-06,
+      "loss": 0.0255,
+      "step": 176
+    },
+    {
+      "epoch": 0.06147968044459882,
+      "grad_norm": 2.197705367927432,
+      "learning_rate": 9.974383325089962e-06,
+      "loss": 0.0222,
+      "step": 177
+    },
+    {
+      "epoch": 0.061827023271969435,
+      "grad_norm": 1.2087103303661568,
+      "learning_rate": 9.973811403311192e-06,
+      "loss": 0.0336,
+      "step": 178
+    },
+    {
+      "epoch": 0.06217436609934005,
+      "grad_norm": 1.831109776996565,
+      "learning_rate": 9.97323318417729e-06,
+      "loss": 0.0292,
+      "step": 179
+    },
+    {
+      "epoch": 0.06252170892671066,
+      "grad_norm": 0.6811940346616355,
+      "learning_rate": 9.972648668420346e-06,
+      "loss": 0.0241,
+      "step": 180
+    },
+    {
+      "epoch": 0.06286905175408128,
+      "grad_norm": 0.7273972811687258,
+      "learning_rate": 9.972057856780412e-06,
+      "loss": 0.0164,
+      "step": 181
+    },
+    {
+      "epoch": 0.0632163945814519,
+      "grad_norm": 1.795643396141142,
+      "learning_rate": 9.97146075000552e-06,
+      "loss": 0.0233,
+      "step": 182
+    },
+    {
+      "epoch": 0.06356373740882251,
+      "grad_norm": 0.7725411745774898,
+      "learning_rate": 9.970857348851667e-06,
+      "loss": 0.0352,
+      "step": 183
+    },
+    {
+      "epoch": 0.06391108023619312,
+      "grad_norm": 0.7933463105547254,
+      "learning_rate": 9.970247654082816e-06,
+      "loss": 0.0296,
+      "step": 184
+    },
+    {
+      "epoch": 0.06425842306356373,
+      "grad_norm": 2.2605994347865774,
+      "learning_rate": 9.96963166647091e-06,
+      "loss": 0.04,
+      "step": 185
+    },
+    {
+      "epoch": 0.06460576589093435,
+      "grad_norm": 1.0749382398980514,
+      "learning_rate": 9.969009386795849e-06,
+      "loss": 0.022,
+      "step": 186
+    },
+    {
+      "epoch": 0.06495310871830497,
+      "grad_norm": 0.557792334948461,
+      "learning_rate": 9.968380815845504e-06,
+      "loss": 0.0327,
+      "step": 187
+    },
+    {
+      "epoch": 0.06530045154567558,
+      "grad_norm": 3.518645514803458,
+      "learning_rate": 9.967745954415711e-06,
+      "loss": 0.0286,
+      "step": 188
+    },
+    {
+      "epoch": 0.0656477943730462,
+      "grad_norm": 0.5873301047411636,
+      "learning_rate": 9.967104803310266e-06,
+      "loss": 0.0295,
+      "step": 189
+    },
+    {
+      "epoch": 0.06599513720041682,
+      "grad_norm": 0.6753747187607116,
+      "learning_rate": 9.966457363340936e-06,
+      "loss": 0.026,
+      "step": 190
+    },
+    {
+      "epoch": 0.06634248002778742,
+      "grad_norm": 0.4603778678263685,
+      "learning_rate": 9.965803635327445e-06,
+      "loss": 0.028,
+      "step": 191
+    },
+    {
+      "epoch": 0.06668982285515804,
+      "grad_norm": 0.9556093549811782,
+      "learning_rate": 9.965143620097479e-06,
+      "loss": 0.0195,
+      "step": 192
+    },
+    {
+      "epoch": 0.06703716568252865,
+      "grad_norm": 2.1021718467486745,
+      "learning_rate": 9.964477318486687e-06,
+      "loss": 0.0256,
+      "step": 193
+    },
+    {
+      "epoch": 0.06738450850989927,
+      "grad_norm": 1.714896004446046,
+      "learning_rate": 9.963804731338674e-06,
+      "loss": 0.0245,
+      "step": 194
+    },
+    {
+      "epoch": 0.06773185133726989,
+      "grad_norm": 0.8958701314528463,
+      "learning_rate": 9.963125859505e-06,
+      "loss": 0.0367,
+      "step": 195
+    },
+    {
+      "epoch": 0.0680791941646405,
+      "grad_norm": 1.4512351067062361,
+      "learning_rate": 9.962440703845193e-06,
+      "loss": 0.0204,
+      "step": 196
+    },
+    {
+      "epoch": 0.06842653699201111,
+      "grad_norm": 0.8793852979074263,
+      "learning_rate": 9.961749265226728e-06,
+      "loss": 0.0182,
+      "step": 197
+    },
+    {
+      "epoch": 0.06877387981938173,
+      "grad_norm": 1.8655135067659925,
+      "learning_rate": 9.961051544525037e-06,
+      "loss": 0.0228,
+      "step": 198
+    },
+    {
+      "epoch": 0.06912122264675234,
+      "grad_norm": 2.2100382744344698,
+      "learning_rate": 9.960347542623506e-06,
+      "loss": 0.0344,
+      "step": 199
+    },
+    {
+      "epoch": 0.06946856547412296,
+      "grad_norm": 0.9843602786620123,
+      "learning_rate": 9.959637260413471e-06,
+      "loss": 0.0298,
+      "step": 200
+    },
+    {
+      "epoch": 0.06981590830149358,
+      "grad_norm": 3.5094590361679496,
+      "learning_rate": 9.958920698794226e-06,
+      "loss": 0.037,
+      "step": 201
+    },
+    {
+      "epoch": 0.07016325112886419,
+      "grad_norm": 1.5466354807280334,
+      "learning_rate": 9.958197858673009e-06,
+      "loss": 0.0294,
+      "step": 202
+    },
+    {
+      "epoch": 0.07051059395623481,
+      "grad_norm": 3.257914253480638,
+      "learning_rate": 9.95746874096501e-06,
+      "loss": 0.0343,
+      "step": 203
+    },
+    {
+      "epoch": 0.07085793678360541,
+      "grad_norm": 3.1724146363151333,
+      "learning_rate": 9.95673334659337e-06,
+      "loss": 0.046,
+      "step": 204
+    },
+    {
+      "epoch": 0.07120527961097603,
+      "grad_norm": 2.045576669926992,
+      "learning_rate": 9.95599167648917e-06,
+      "loss": 0.0328,
+      "step": 205
+    },
+    {
+      "epoch": 0.07155262243834665,
+      "grad_norm": 4.362827099663752,
+      "learning_rate": 9.95524373159144e-06,
+      "loss": 0.0333,
+      "step": 206
+    },
+    {
+      "epoch": 0.07189996526571726,
+      "grad_norm": 2.1027185240369244,
+      "learning_rate": 9.954489512847156e-06,
+      "loss": 0.0278,
+      "step": 207
+    },
+    {
+      "epoch": 0.07224730809308788,
+      "grad_norm": 1.951261682373136,
+      "learning_rate": 9.953729021211238e-06,
+      "loss": 0.0369,
+      "step": 208
+    },
+    {
+      "epoch": 0.0725946509204585,
+      "grad_norm": 2.70690992472503,
+      "learning_rate": 9.952962257646545e-06,
+      "loss": 0.0368,
+      "step": 209
+    },
+    {
+      "epoch": 0.0729419937478291,
+      "grad_norm": 1.1045388371026474,
+      "learning_rate": 9.952189223123877e-06,
+      "loss": 0.0198,
+      "step": 210
+    },
+    {
+      "epoch": 0.07328933657519972,
+      "grad_norm": 0.37890065356086955,
+      "learning_rate": 9.951409918621977e-06,
+      "loss": 0.0275,
+      "step": 211
+    },
+    {
+      "epoch": 0.07363667940257033,
+      "grad_norm": 1.27246581897078,
+      "learning_rate": 9.950624345127523e-06,
+      "loss": 0.0269,
+      "step": 212
+    },
+    {
+      "epoch": 0.07398402222994095,
+      "grad_norm": 0.6532636236631134,
+      "learning_rate": 9.949832503635133e-06,
+      "loss": 0.0275,
+      "step": 213
+    },
+    {
+      "epoch": 0.07433136505731157,
+      "grad_norm": 2.5733332280836576,
+      "learning_rate": 9.949034395147357e-06,
+      "loss": 0.0274,
+      "step": 214
+    },
+    {
+      "epoch": 0.07467870788468219,
+      "grad_norm": 2.652931462756075,
+      "learning_rate": 9.948230020674685e-06,
+      "loss": 0.0321,
+      "step": 215
+    },
+    {
+      "epoch": 0.0750260507120528,
+      "grad_norm": 1.2497045154490993,
+      "learning_rate": 9.947419381235538e-06,
+      "loss": 0.0277,
+      "step": 216
+    },
+    {
+      "epoch": 0.0753733935394234,
+      "grad_norm": 1.029916760819159,
+      "learning_rate": 9.946602477856262e-06,
+      "loss": 0.0223,
+      "step": 217
+    },
+    {
+      "epoch": 0.07572073636679402,
+      "grad_norm": 1.0496525720606742,
+      "learning_rate": 9.94577931157115e-06,
+      "loss": 0.0309,
+      "step": 218
+    },
+    {
+      "epoch": 0.07606807919416464,
+      "grad_norm": 0.8813614432848731,
+      "learning_rate": 9.944949883422409e-06,
+      "loss": 0.0223,
+      "step": 219
+    },
+    {
+      "epoch": 0.07641542202153526,
+      "grad_norm": 1.048099444449534,
+      "learning_rate": 9.944114194460181e-06,
+      "loss": 0.0358,
+      "step": 220
+    },
+    {
+      "epoch": 0.07676276484890587,
+      "grad_norm": 0.5608593650182933,
+      "learning_rate": 9.943272245742534e-06,
+      "loss": 0.0197,
+      "step": 221
+    },
+    {
+      "epoch": 0.07711010767627649,
+      "grad_norm": 1.6998010859628299,
+      "learning_rate": 9.942424038335462e-06,
+      "loss": 0.0255,
+      "step": 222
+    },
+    {
+      "epoch": 0.0774574505036471,
+      "grad_norm": 2.345709307993445,
+      "learning_rate": 9.941569573312882e-06,
+      "loss": 0.0341,
+      "step": 223
+    },
+    {
+      "epoch": 0.07780479333101771,
+      "grad_norm": 0.7888283535068908,
+      "learning_rate": 9.940708851756633e-06,
+      "loss": 0.0204,
+      "step": 224
+    },
+    {
+      "epoch": 0.07815213615838833,
+      "grad_norm": 0.5490007316162747,
+      "learning_rate": 9.939841874756481e-06,
+      "loss": 0.0273,
+      "step": 225
+    },
+    {
+      "epoch": 0.07849947898575894,
+      "grad_norm": 1.1934836714677952,
+      "learning_rate": 9.938968643410103e-06,
+      "loss": 0.0206,
+      "step": 226
+    },
+    {
+      "epoch": 0.07884682181312956,
+      "grad_norm": 0.7241877331482239,
+      "learning_rate": 9.938089158823101e-06,
+      "loss": 0.0221,
+      "step": 227
+    },
+    {
+      "epoch": 0.07919416464050018,
+      "grad_norm": 2.4778189410310696,
+      "learning_rate": 9.937203422108995e-06,
+      "loss": 0.0308,
+      "step": 228
+    },
+    {
+      "epoch": 0.0795415074678708,
+      "grad_norm": 3.748172451740872,
+      "learning_rate": 9.936311434389216e-06,
+      "loss": 0.0427,
+      "step": 229
+    },
+    {
+      "epoch": 0.0798888502952414,
+      "grad_norm": 0.7889067721600285,
+      "learning_rate": 9.935413196793111e-06,
+      "loss": 0.0193,
+      "step": 230
+    },
+    {
+      "epoch": 0.08023619312261202,
+      "grad_norm": 0.704831305338057,
+      "learning_rate": 9.934508710457944e-06,
+      "loss": 0.0194,
+      "step": 231
+    },
+    {
+      "epoch": 0.08058353594998263,
+      "grad_norm": 0.40717838994648536,
+      "learning_rate": 9.933597976528883e-06,
+      "loss": 0.0254,
+      "step": 232
+    },
+    {
+      "epoch": 0.08093087877735325,
+      "grad_norm": 1.964638344320029,
+      "learning_rate": 9.932680996159016e-06,
+      "loss": 0.0283,
+      "step": 233
+    },
+    {
+      "epoch": 0.08127822160472387,
+      "grad_norm": 2.4616476396806473,
+      "learning_rate": 9.931757770509332e-06,
+      "loss": 0.0364,
+      "step": 234
+    },
+    {
+      "epoch": 0.08162556443209448,
+      "grad_norm": 0.6589471069941858,
+      "learning_rate": 9.930828300748726e-06,
+      "loss": 0.0152,
+      "step": 235
+    },
+    {
+      "epoch": 0.08197290725946509,
+      "grad_norm": 1.5170996006478779,
+      "learning_rate": 9.929892588054007e-06,
+      "loss": 0.0456,
+      "step": 236
+    },
+    {
+      "epoch": 0.0823202500868357,
+      "grad_norm": 0.49344733685789915,
+      "learning_rate": 9.928950633609878e-06,
+      "loss": 0.0157,
+      "step": 237
+    },
+    {
+      "epoch": 0.08266759291420632,
+      "grad_norm": 1.2510013978714254,
+      "learning_rate": 9.928002438608955e-06,
+      "loss": 0.0264,
+      "step": 238
+    },
+    {
+      "epoch": 0.08301493574157694,
+      "grad_norm": 0.6672029612706915,
+      "learning_rate": 9.927048004251748e-06,
+      "loss": 0.0232,
+      "step": 239
+    },
+    {
+      "epoch": 0.08336227856894755,
+      "grad_norm": 0.7870864791530758,
+      "learning_rate": 9.926087331746668e-06,
+      "loss": 0.0241,
+      "step": 240
+    },
+    {
+      "epoch": 0.08370962139631817,
+      "grad_norm": 0.4430411461060668,
+      "learning_rate": 9.925120422310023e-06,
+      "loss": 0.0239,
+      "step": 241
+    },
+    {
+      "epoch": 0.08405696422368879,
+      "grad_norm": 0.7393154201127319,
+      "learning_rate": 9.924147277166025e-06,
+      "loss": 0.0221,
+      "step": 242
+    },
+    {
+      "epoch": 0.08440430705105939,
+      "grad_norm": 1.5211011444968148,
+      "learning_rate": 9.923167897546773e-06,
+      "loss": 0.0361,
+      "step": 243
+    },
+    {
+      "epoch": 0.08475164987843001,
+      "grad_norm": 0.306402457203348,
+      "learning_rate": 9.92218228469226e-06,
+      "loss": 0.0161,
+      "step": 244
+    },
+    {
+      "epoch": 0.08509899270580062,
+      "grad_norm": 0.6425545998545745,
+      "learning_rate": 9.921190439850374e-06,
+      "loss": 0.0336,
+      "step": 245
+    },
+    {
+      "epoch": 0.08544633553317124,
+      "grad_norm": 1.1621242417732545,
+      "learning_rate": 9.920192364276894e-06,
+      "loss": 0.0346,
+      "step": 246
+    },
+    {
+      "epoch": 0.08579367836054186,
+      "grad_norm": 0.8594232560472324,
+      "learning_rate": 9.919188059235483e-06,
+      "loss": 0.0276,
+      "step": 247
+    },
+    {
+      "epoch": 0.08614102118791248,
+      "grad_norm": 1.1817899977122954,
+      "learning_rate": 9.918177525997697e-06,
+      "loss": 0.0237,
+      "step": 248
+    },
+    {
+      "epoch": 0.08648836401528308,
+      "grad_norm": 0.663618476895076,
+      "learning_rate": 9.917160765842972e-06,
+      "loss": 0.0147,
+      "step": 249
+    },
+    {
+      "epoch": 0.0868357068426537,
+      "grad_norm": 2.1119160052926906,
+      "learning_rate": 9.916137780058634e-06,
+      "loss": 0.0372,
+      "step": 250
+    },
+    {
+      "epoch": 0.08718304967002431,
+      "grad_norm": 0.5192464513916062,
+      "learning_rate": 9.915108569939884e-06,
+      "loss": 0.0226,
+      "step": 251
+    },
+    {
+      "epoch": 0.08753039249739493,
+      "grad_norm": 1.4192863431762925,
+      "learning_rate": 9.914073136789812e-06,
+      "loss": 0.0208,
+      "step": 252
+    },
+    {
+      "epoch": 0.08787773532476555,
+      "grad_norm": 1.4952635090772373,
+      "learning_rate": 9.913031481919378e-06,
+      "loss": 0.0256,
+      "step": 253
+    },
+    {
+      "epoch": 0.08822507815213616,
+      "grad_norm": 0.7782630656189196,
+      "learning_rate": 9.911983606647426e-06,
+      "loss": 0.0301,
+      "step": 254
+    },
+    {
+      "epoch": 0.08857242097950677,
+      "grad_norm": 1.0227442959844868,
+      "learning_rate": 9.910929512300673e-06,
+      "loss": 0.0205,
+      "step": 255
+    },
+    {
+      "epoch": 0.08891976380687738,
+      "grad_norm": 1.0984883231423739,
+      "learning_rate": 9.909869200213711e-06,
+      "loss": 0.0272,
+      "step": 256
+    },
+    {
+      "epoch": 0.089267106634248,
+      "grad_norm": 2.616121786507333,
+      "learning_rate": 9.908802671729004e-06,
+      "loss": 0.0298,
+      "step": 257
+    },
+    {
+      "epoch": 0.08961444946161862,
+      "grad_norm": 1.0424691898538212,
+      "learning_rate": 9.907729928196885e-06,
+      "loss": 0.0221,
+      "step": 258
+    },
+    {
+      "epoch": 0.08996179228898923,
+      "grad_norm": 1.2136434228663253,
+      "learning_rate": 9.90665097097556e-06,
+      "loss": 0.0344,
+      "step": 259
+    },
+    {
+      "epoch": 0.09030913511635985,
+      "grad_norm": 1.0607165560461278,
+      "learning_rate": 9.905565801431097e-06,
+      "loss": 0.0278,
+      "step": 260
+    },
+    {
+      "epoch": 0.09065647794373047,
+      "grad_norm": 0.7592364646233872,
+      "learning_rate": 9.904474420937431e-06,
+      "loss": 0.0241,
+      "step": 261
+    },
+    {
+      "epoch": 0.09100382077110107,
+      "grad_norm": 1.142833500760027,
+      "learning_rate": 9.903376830876363e-06,
+      "loss": 0.0194,
+      "step": 262
+    },
+    {
+      "epoch": 0.09135116359847169,
+      "grad_norm": 1.5792096178170156,
+      "learning_rate": 9.902273032637558e-06,
+      "loss": 0.0285,
+      "step": 263
+    },
+    {
+      "epoch": 0.0916985064258423,
+      "grad_norm": 1.2090200476218058,
+      "learning_rate": 9.901163027618532e-06,
+      "loss": 0.027,
+      "step": 264
+    },
+    {
+      "epoch": 0.09204584925321292,
+      "grad_norm": 0.7247661863395669,
+      "learning_rate": 9.90004681722467e-06,
+      "loss": 0.0381,
+      "step": 265
+    },
+    {
+      "epoch": 0.09239319208058354,
+      "grad_norm": 1.3359604948328887,
+      "learning_rate": 9.898924402869204e-06,
+      "loss": 0.025,
+      "step": 266
+    },
+    {
+      "epoch": 0.09274053490795416,
+      "grad_norm": 1.5054331995013246,
+      "learning_rate": 9.897795785973227e-06,
+      "loss": 0.0176,
+      "step": 267
+    },
+    {
+      "epoch": 0.09308787773532476,
+      "grad_norm": 0.3800080517239375,
+      "learning_rate": 9.896660967965688e-06,
+      "loss": 0.0206,
+      "step": 268
+    },
+    {
+      "epoch": 0.09343522056269538,
+      "grad_norm": 2.2058762679111212,
+      "learning_rate": 9.895519950283378e-06,
+      "loss": 0.0382,
+      "step": 269
+    },
+    {
+      "epoch": 0.09378256339006599,
+      "grad_norm": 1.987671837514159,
+      "learning_rate": 9.894372734370945e-06,
+      "loss": 0.0245,
+      "step": 270
+    },
+    {
+      "epoch": 0.09412990621743661,
+      "grad_norm": 1.1525093904407058,
+      "learning_rate": 9.89321932168088e-06,
+      "loss": 0.0294,
+      "step": 271
+    },
+    {
+      "epoch": 0.09447724904480723,
+      "grad_norm": 0.5288867918479366,
+      "learning_rate": 9.892059713673521e-06,
+      "loss": 0.0329,
+      "step": 272
+    },
+    {
+      "epoch": 0.09482459187217784,
+      "grad_norm": 2.3156241409626537,
+      "learning_rate": 9.890893911817056e-06,
+      "loss": 0.0281,
+      "step": 273
+    },
+    {
+      "epoch": 0.09517193469954846,
+      "grad_norm": 1.6457301083303564,
+      "learning_rate": 9.889721917587504e-06,
+      "loss": 0.0246,
+      "step": 274
+    },
+    {
+      "epoch": 0.09551927752691906,
+      "grad_norm": 1.3357334045720493,
+      "learning_rate": 9.888543732468732e-06,
+      "loss": 0.0285,
+      "step": 275
+    },
+    {
+      "epoch": 0.09586662035428968,
+      "grad_norm": 1.283091226837623,
+      "learning_rate": 9.887359357952441e-06,
+      "loss": 0.0354,
+      "step": 276
+    },
+    {
+      "epoch": 0.0962139631816603,
+      "grad_norm": 1.468754392397442,
+      "learning_rate": 9.886168795538175e-06,
+      "loss": 0.0243,
+      "step": 277
+    },
+    {
+      "epoch": 0.09656130600903091,
+      "grad_norm": 0.5847171116164384,
+      "learning_rate": 9.884972046733306e-06,
+      "loss": 0.0187,
+      "step": 278
+    },
+    {
+      "epoch": 0.09690864883640153,
+      "grad_norm": 0.7176769597295244,
+      "learning_rate": 9.883769113053039e-06,
+      "loss": 0.0242,
+      "step": 279
+    },
+    {
+      "epoch": 0.09725599166377215,
+      "grad_norm": 0.4262926228040994,
+      "learning_rate": 9.882559996020414e-06,
+      "loss": 0.0179,
+      "step": 280
+    },
+    {
+      "epoch": 0.09760333449114275,
+      "grad_norm": 0.9525497182278096,
+      "learning_rate": 9.881344697166293e-06,
+      "loss": 0.0349,
+      "step": 281
+    },
+    {
+      "epoch": 0.09795067731851337,
+      "grad_norm": 0.9791433669149336,
+      "learning_rate": 9.880123218029374e-06,
+      "loss": 0.0308,
+      "step": 282
+    },
+    {
+      "epoch": 0.09829802014588399,
+      "grad_norm": 0.7630066342045814,
+      "learning_rate": 9.878895560156172e-06,
+      "loss": 0.0205,
+      "step": 283
+    },
+    {
+      "epoch": 0.0986453629732546,
+      "grad_norm": 0.364457070730968,
+      "learning_rate": 9.877661725101028e-06,
+      "loss": 0.0195,
+      "step": 284
+    },
+    {
+      "epoch": 0.09899270580062522,
+      "grad_norm": 1.2950920833202644,
+      "learning_rate": 9.876421714426104e-06,
+      "loss": 0.0211,
+      "step": 285
+    },
+    {
+      "epoch": 0.09934004862799584,
+      "grad_norm": 3.594479893409194,
+      "learning_rate": 9.87517552970138e-06,
+      "loss": 0.0375,
+      "step": 286
+    },
+    {
+      "epoch": 0.09968739145536645,
+      "grad_norm": 1.3559833171841018,
+      "learning_rate": 9.873923172504653e-06,
+      "loss": 0.0306,
+      "step": 287
+    },
+    {
+      "epoch": 0.10003473428273706,
+      "grad_norm": 1.0574786377389542,
+      "learning_rate": 9.872664644421539e-06,
+      "loss": 0.0378,
+      "step": 288
+    },
+    {
+      "epoch": 0.10038207711010767,
+      "grad_norm": 0.8553158210643477,
+      "learning_rate": 9.87139994704546e-06,
+      "loss": 0.017,
+      "step": 289
+    },
+    {
+      "epoch": 0.10072941993747829,
+      "grad_norm": 1.2327340151649626,
+      "learning_rate": 9.870129081977654e-06,
+      "loss": 0.0291,
+      "step": 290
+    },
+    {
+      "epoch": 0.10107676276484891,
+      "grad_norm": 0.6120639599947546,
+      "learning_rate": 9.868852050827167e-06,
+      "loss": 0.0248,
+      "step": 291
+    },
+    {
+      "epoch": 0.10142410559221952,
+      "grad_norm": 1.3504767761554735,
+      "learning_rate": 9.86756885521085e-06,
+      "loss": 0.0196,
+      "step": 292
+    },
+    {
+      "epoch": 0.10177144841959014,
+      "grad_norm": 2.130326569000447,
+      "learning_rate": 9.866279496753361e-06,
+      "loss": 0.0272,
+      "step": 293
+    },
+    {
+      "epoch": 0.10211879124696074,
+      "grad_norm": 2.961313545763907,
+      "learning_rate": 9.86498397708716e-06,
+      "loss": 0.0306,
+      "step": 294
+    },
+    {
+      "epoch": 0.10246613407433136,
+      "grad_norm": 0.6185598205757522,
+      "learning_rate": 9.863682297852506e-06,
+      "loss": 0.0268,
+      "step": 295
+    },
+    {
+      "epoch": 0.10281347690170198,
+      "grad_norm": 2.900646593623201,
+      "learning_rate": 9.862374460697462e-06,
+      "loss": 0.0299,
+      "step": 296
+    },
+    {
+      "epoch": 0.1031608197290726,
+      "grad_norm": 4.815048093426214,
+      "learning_rate": 9.86106046727788e-06,
+      "loss": 0.0413,
+      "step": 297
+    },
+    {
+      "epoch": 0.10350816255644321,
+      "grad_norm": 3.6854621061195236,
+      "learning_rate": 9.859740319257413e-06,
+      "loss": 0.0315,
+      "step": 298
+    },
+    {
+      "epoch": 0.10385550538381383,
+      "grad_norm": 3.1485787317926315,
+      "learning_rate": 9.858414018307503e-06,
+      "loss": 0.0395,
+      "step": 299
+    },
+    {
+      "epoch": 0.10420284821118445,
+      "grad_norm": 1.6857361849761061,
+      "learning_rate": 9.857081566107383e-06,
+      "loss": 0.0306,
+      "step": 300
+    },
+    {
+      "epoch": 0.10455019103855505,
+      "grad_norm": 1.1203216330091716,
+      "learning_rate": 9.855742964344074e-06,
+      "loss": 0.0243,
+      "step": 301
+    },
+    {
+      "epoch": 0.10489753386592567,
+      "grad_norm": 2.214128202481394,
+      "learning_rate": 9.854398214712382e-06,
+      "loss": 0.0234,
+      "step": 302
+    },
+    {
+      "epoch": 0.10524487669329628,
+      "grad_norm": 1.6017019194654412,
+      "learning_rate": 9.853047318914898e-06,
+      "loss": 0.0291,
+      "step": 303
+    },
+    {
+      "epoch": 0.1055922195206669,
+      "grad_norm": 2.2711519721461917,
+      "learning_rate": 9.851690278661998e-06,
+      "loss": 0.0271,
+      "step": 304
+    },
+    {
+      "epoch": 0.10593956234803752,
+      "grad_norm": 0.4684895446152942,
+      "learning_rate": 9.850327095671831e-06,
+      "loss": 0.0177,
+      "step": 305
+    },
+    {
+      "epoch": 0.10628690517540813,
+      "grad_norm": 1.8303208610929107,
+      "learning_rate": 9.848957771670326e-06,
+      "loss": 0.0286,
+      "step": 306
+    },
+    {
+      "epoch": 0.10663424800277874,
+      "grad_norm": 0.547820372795727,
+      "learning_rate": 9.847582308391189e-06,
+      "loss": 0.0328,
+      "step": 307
+    },
+    {
+      "epoch": 0.10698159083014935,
+      "grad_norm": 1.7911990714879877,
+      "learning_rate": 9.846200707575897e-06,
+      "loss": 0.0153,
+      "step": 308
+    },
+    {
+      "epoch": 0.10732893365751997,
+      "grad_norm": 2.725196937839101,
+      "learning_rate": 9.844812970973699e-06,
+      "loss": 0.0228,
+      "step": 309
+    },
+    {
+      "epoch": 0.10767627648489059,
+      "grad_norm": 2.375094543001517,
+      "learning_rate": 9.843419100341608e-06,
+      "loss": 0.0203,
+      "step": 310
+    },
+    {
+      "epoch": 0.1080236193122612,
+      "grad_norm": 0.732852515115049,
+      "learning_rate": 9.842019097444414e-06,
+      "loss": 0.0214,
+      "step": 311
+    },
+    {
+      "epoch": 0.10837096213963182,
+      "grad_norm": 0.8565652562532247,
+      "learning_rate": 9.840612964054658e-06,
+      "loss": 0.031,
+      "step": 312
+    },
+    {
+      "epoch": 0.10871830496700244,
+      "grad_norm": 1.046089822343109,
+      "learning_rate": 9.839200701952653e-06,
+      "loss": 0.0273,
+      "step": 313
+    },
+    {
+      "epoch": 0.10906564779437304,
+      "grad_norm": 1.4037462533263418,
+      "learning_rate": 9.837782312926465e-06,
+      "loss": 0.0247,
+      "step": 314
+    },
+    {
+      "epoch": 0.10941299062174366,
+      "grad_norm": 0.724135174812571,
+      "learning_rate": 9.836357798771922e-06,
+      "loss": 0.0197,
+      "step": 315
+    },
+    {
+      "epoch": 0.10976033344911428,
+      "grad_norm": 0.8979754709718626,
+      "learning_rate": 9.834927161292604e-06,
+      "loss": 0.0135,
+      "step": 316
+    },
+    {
+      "epoch": 0.11010767627648489,
+      "grad_norm": 0.818961402615803,
+      "learning_rate": 9.833490402299844e-06,
+      "loss": 0.024,
+      "step": 317
+    },
+    {
+      "epoch": 0.11045501910385551,
+      "grad_norm": 1.8250252957901378,
+      "learning_rate": 9.832047523612726e-06,
+      "loss": 0.0305,
+      "step": 318
+    },
+    {
+      "epoch": 0.11080236193122613,
+      "grad_norm": 0.898470564935298,
+      "learning_rate": 9.830598527058083e-06,
+      "loss": 0.0262,
+      "step": 319
+    },
+    {
+      "epoch": 0.11114970475859673,
+      "grad_norm": 0.7300373503368474,
+      "learning_rate": 9.829143414470495e-06,
+      "loss": 0.0286,
+      "step": 320
+    },
+    {
+      "epoch": 0.11149704758596735,
+      "grad_norm": 1.0198415430023702,
+      "learning_rate": 9.82768218769228e-06,
+      "loss": 0.0245,
+      "step": 321
+    },
+    {
+      "epoch": 0.11184439041333796,
+      "grad_norm": 1.5853784707754817,
+      "learning_rate": 9.826214848573503e-06,
+      "loss": 0.0272,
+      "step": 322
+    },
+    {
+      "epoch": 0.11219173324070858,
+      "grad_norm": 0.5862824974087557,
+      "learning_rate": 9.824741398971966e-06,
+      "loss": 0.0168,
+      "step": 323
+    },
+    {
+      "epoch": 0.1125390760680792,
+      "grad_norm": 1.0344261835540582,
+      "learning_rate": 9.823261840753209e-06,
+      "loss": 0.0186,
+      "step": 324
+    },
+    {
+      "epoch": 0.11288641889544981,
+      "grad_norm": 1.3434435771133537,
+      "learning_rate": 9.821776175790501e-06,
+      "loss": 0.0287,
+      "step": 325
+    },
+    {
+      "epoch": 0.11323376172282042,
+      "grad_norm": 0.6130576808513916,
+      "learning_rate": 9.820284405964846e-06,
+      "loss": 0.018,
+      "step": 326
+    },
+    {
+      "epoch": 0.11358110455019103,
+      "grad_norm": 0.6570948194085574,
+      "learning_rate": 9.81878653316498e-06,
+      "loss": 0.0136,
+      "step": 327
+    },
+    {
+      "epoch": 0.11392844737756165,
+      "grad_norm": 1.6328034399729674,
+      "learning_rate": 9.817282559287362e-06,
+      "loss": 0.0277,
+      "step": 328
+    },
+    {
+      "epoch": 0.11427579020493227,
+      "grad_norm": 0.3948241510052037,
+      "learning_rate": 9.815772486236179e-06,
+      "loss": 0.0126,
+      "step": 329
+    },
+    {
+      "epoch": 0.11462313303230288,
+      "grad_norm": 0.5069144619443747,
+      "learning_rate": 9.814256315923335e-06,
+      "loss": 0.0119,
+      "step": 330
+    },
+    {
+      "epoch": 0.1149704758596735,
+      "grad_norm": 1.1749675379829028,
+      "learning_rate": 9.81273405026846e-06,
+      "loss": 0.0401,
+      "step": 331
+    },
+    {
+      "epoch": 0.11531781868704412,
+      "grad_norm": 1.38512145823474,
+      "learning_rate": 9.811205691198897e-06,
+      "loss": 0.0255,
+      "step": 332
+    },
+    {
+      "epoch": 0.11566516151441472,
+      "grad_norm": 1.6588777415442508,
+      "learning_rate": 9.809671240649705e-06,
+      "loss": 0.0232,
+      "step": 333
+    },
+    {
+      "epoch": 0.11601250434178534,
+      "grad_norm": 0.891351380406614,
+      "learning_rate": 9.808130700563658e-06,
+      "loss": 0.0182,
+      "step": 334
+    },
+    {
+      "epoch": 0.11635984716915596,
+      "grad_norm": 1.9182917391789063,
+      "learning_rate": 9.806584072891234e-06,
+      "loss": 0.0295,
+      "step": 335
+    },
+    {
+      "epoch": 0.11670718999652657,
+      "grad_norm": 1.414292519566524,
+      "learning_rate": 9.805031359590626e-06,
+      "loss": 0.0182,
+      "step": 336
+    },
+    {
+      "epoch": 0.11705453282389719,
+      "grad_norm": 1.048723673722958,
+      "learning_rate": 9.803472562627726e-06,
+      "loss": 0.0119,
+      "step": 337
+    },
+    {
+      "epoch": 0.1174018756512678,
+      "grad_norm": 1.2747705868686736,
+      "learning_rate": 9.801907683976128e-06,
+      "loss": 0.0272,
+      "step": 338
+    },
+    {
+      "epoch": 0.11774921847863841,
+      "grad_norm": 0.9457245602215513,
+      "learning_rate": 9.800336725617136e-06,
+      "loss": 0.0201,
+      "step": 339
+    },
+    {
+      "epoch": 0.11809656130600903,
+      "grad_norm": 1.2510122100134684,
+      "learning_rate": 9.798759689539739e-06,
+      "loss": 0.0257,
+      "step": 340
+    },
+    {
+      "epoch": 0.11844390413337964,
+      "grad_norm": 0.8319010285486187,
+      "learning_rate": 9.797176577740625e-06,
+      "loss": 0.0254,
+      "step": 341
+    },
+    {
+      "epoch": 0.11879124696075026,
+      "grad_norm": 0.7850673506313961,
+      "learning_rate": 9.795587392224182e-06,
+      "loss": 0.0234,
+      "step": 342
+    },
+    {
+      "epoch": 0.11913858978812088,
+      "grad_norm": 0.587424666600968,
+      "learning_rate": 9.793992135002476e-06,
+      "loss": 0.0179,
+      "step": 343
+    },
+    {
+      "epoch": 0.1194859326154915,
+      "grad_norm": 1.2534232883052836,
+      "learning_rate": 9.792390808095268e-06,
+      "loss": 0.0228,
+      "step": 344
+    },
+    {
+      "epoch": 0.11983327544286211,
+      "grad_norm": 1.1177620903458203,
+      "learning_rate": 9.790783413530006e-06,
+      "loss": 0.0283,
+      "step": 345
+    },
+    {
+      "epoch": 0.12018061827023271,
+      "grad_norm": 0.7956916247471605,
+      "learning_rate": 9.789169953341809e-06,
+      "loss": 0.026,
+      "step": 346
+    },
+    {
+      "epoch": 0.12052796109760333,
+      "grad_norm": 1.4672529265658403,
+      "learning_rate": 9.787550429573487e-06,
+      "loss": 0.0242,
+      "step": 347
+    },
+    {
+      "epoch": 0.12087530392497395,
+      "grad_norm": 0.499722630858149,
+      "learning_rate": 9.785924844275523e-06,
+      "loss": 0.0243,
+      "step": 348
+    },
+    {
+      "epoch": 0.12122264675234456,
+      "grad_norm": 0.7528884315208382,
+      "learning_rate": 9.784293199506076e-06,
+      "loss": 0.0312,
+      "step": 349
+    },
+    {
+      "epoch": 0.12156998957971518,
+      "grad_norm": 0.9565348415093637,
+      "learning_rate": 9.782655497330972e-06,
+      "loss": 0.0149,
+      "step": 350
+    },
+    {
+      "epoch": 0.1219173324070858,
+      "grad_norm": 4.700227177143458,
+      "learning_rate": 9.781011739823715e-06,
+      "loss": 0.0459,
+      "step": 351
+    },
+    {
+      "epoch": 0.1222646752344564,
+      "grad_norm": 2.0797417224686545,
+      "learning_rate": 9.779361929065462e-06,
+      "loss": 0.0248,
+      "step": 352
+    },
+    {
+      "epoch": 0.12261201806182702,
+      "grad_norm": 1.5530326186566639,
+      "learning_rate": 9.777706067145052e-06,
+      "loss": 0.0278,
+      "step": 353
+    },
+    {
+      "epoch": 0.12295936088919764,
+      "grad_norm": 0.8614425997016937,
+      "learning_rate": 9.77604415615897e-06,
+      "loss": 0.0171,
+      "step": 354
+    },
+    {
+      "epoch": 0.12330670371656825,
+      "grad_norm": 0.798882409804884,
+      "learning_rate": 9.77437619821137e-06,
+      "loss": 0.0221,
+      "step": 355
+    },
+    {
+      "epoch": 0.12365404654393887,
+      "grad_norm": 1.309963463600437,
+      "learning_rate": 9.772702195414053e-06,
+      "loss": 0.0227,
+      "step": 356
+    },
+    {
+      "epoch": 0.12400138937130949,
+      "grad_norm": 1.0239470706246998,
+      "learning_rate": 9.771022149886482e-06,
+      "loss": 0.0247,
+      "step": 357
+    },
+    {
+      "epoch": 0.1243487321986801,
+      "grad_norm": 0.694165293574145,
+      "learning_rate": 9.769336063755763e-06,
+      "loss": 0.0165,
+      "step": 358
+    },
+    {
+      "epoch": 0.1246960750260507,
+      "grad_norm": 0.8806399954727403,
+      "learning_rate": 9.767643939156658e-06,
+      "loss": 0.0315,
+      "step": 359
+    },
+    {
+      "epoch": 0.12504341785342132,
+      "grad_norm": 1.4519111212390032,
+      "learning_rate": 9.765945778231568e-06,
+      "loss": 0.0206,
+      "step": 360
+    },
+    {
+      "epoch": 0.12539076068079194,
+      "grad_norm": 0.7103007161003656,
+      "learning_rate": 9.76424158313054e-06,
+      "loss": 0.0206,
+      "step": 361
+    },
+    {
+      "epoch": 0.12573810350816256,
+      "grad_norm": 1.4369518423012835,
+      "learning_rate": 9.762531356011258e-06,
+      "loss": 0.0309,
+      "step": 362
+    },
+    {
+      "epoch": 0.12608544633553317,
+      "grad_norm": 1.9088679954477914,
+      "learning_rate": 9.760815099039045e-06,
+      "loss": 0.025,
+      "step": 363
+    },
+    {
+      "epoch": 0.1264327891629038,
+      "grad_norm": 1.2218949903799274,
+      "learning_rate": 9.75909281438686e-06,
+      "loss": 0.0241,
+      "step": 364
+    },
+    {
+      "epoch": 0.1267801319902744,
+      "grad_norm": 0.8840693103252929,
+      "learning_rate": 9.757364504235292e-06,
+      "loss": 0.0256,
+      "step": 365
+    },
+    {
+      "epoch": 0.12712747481764503,
+      "grad_norm": 2.3783985810393853,
+      "learning_rate": 9.755630170772556e-06,
+      "loss": 0.0279,
+      "step": 366
+    },
+    {
+      "epoch": 0.12747481764501564,
+      "grad_norm": 2.1206965462444987,
+      "learning_rate": 9.753889816194498e-06,
+      "loss": 0.0371,
+      "step": 367
+    },
+    {
+      "epoch": 0.12782216047238623,
+      "grad_norm": 3.735386143479719,
+      "learning_rate": 9.752143442704586e-06,
+      "loss": 0.0474,
+      "step": 368
+    },
+    {
+      "epoch": 0.12816950329975685,
+      "grad_norm": 0.6664661005983077,
+      "learning_rate": 9.750391052513906e-06,
+      "loss": 0.026,
+      "step": 369
+    },
+    {
+      "epoch": 0.12851684612712747,
+      "grad_norm": 0.3453207995141737,
+      "learning_rate": 9.748632647841165e-06,
+      "loss": 0.0189,
+      "step": 370
+    },
+    {
+      "epoch": 0.12886418895449808,
+      "grad_norm": 1.1376711577741057,
+      "learning_rate": 9.746868230912683e-06,
+      "loss": 0.0347,
+      "step": 371
+    },
+    {
+      "epoch": 0.1292115317818687,
+      "grad_norm": 1.0621834917092619,
+      "learning_rate": 9.745097803962394e-06,
+      "loss": 0.0212,
+      "step": 372
+    },
+    {
+      "epoch": 0.12955887460923932,
+      "grad_norm": 1.4420764224102458,
+      "learning_rate": 9.743321369231837e-06,
+      "loss": 0.0288,
+      "step": 373
+    },
+    {
+      "epoch": 0.12990621743660993,
+      "grad_norm": 0.47180137756368545,
+      "learning_rate": 9.741538928970163e-06,
+      "loss": 0.0212,
+      "step": 374
+    },
+    {
+      "epoch": 0.13025356026398055,
+      "grad_norm": 1.0223869124759941,
+      "learning_rate": 9.739750485434126e-06,
+      "loss": 0.0275,
+      "step": 375
+    },
+    {
+      "epoch": 0.13060090309135117,
+      "grad_norm": 0.3467757694541222,
+      "learning_rate": 9.737956040888073e-06,
+      "loss": 0.0212,
+      "step": 376
+    },
+    {
+      "epoch": 0.13094824591872178,
+      "grad_norm": 0.4743085064117592,
+      "learning_rate": 9.736155597603959e-06,
+      "loss": 0.0238,
+      "step": 377
+    },
+    {
+      "epoch": 0.1312955887460924,
+      "grad_norm": 0.5939635024690287,
+      "learning_rate": 9.734349157861329e-06,
+      "loss": 0.0187,
+      "step": 378
+    },
+    {
+      "epoch": 0.13164293157346302,
+      "grad_norm": 0.8656882396061,
+      "learning_rate": 9.73253672394732e-06,
+      "loss": 0.0267,
+      "step": 379
+    },
+    {
+      "epoch": 0.13199027440083363,
+      "grad_norm": 0.9530332353603665,
+      "learning_rate": 9.73071829815666e-06,
+      "loss": 0.0226,
+      "step": 380
+    },
+    {
+      "epoch": 0.13233761722820422,
+      "grad_norm": 1.126170471334089,
+      "learning_rate": 9.728893882791663e-06,
+      "loss": 0.0264,
+      "step": 381
+    },
+    {
+      "epoch": 0.13268496005557484,
+      "grad_norm": 1.0981019989116185,
+      "learning_rate": 9.727063480162226e-06,
+      "loss": 0.0192,
+      "step": 382
+    },
+    {
+      "epoch": 0.13303230288294546,
+      "grad_norm": 0.890921570685852,
+      "learning_rate": 9.725227092585824e-06,
+      "loss": 0.0291,
+      "step": 383
+    },
+    {
+      "epoch": 0.13337964571031607,
+      "grad_norm": 0.785030282186115,
+      "learning_rate": 9.723384722387516e-06,
+      "loss": 0.0239,
+      "step": 384
+    },
+    {
+      "epoch": 0.1337269885376867,
+      "grad_norm": 2.0060229292652942,
+      "learning_rate": 9.721536371899928e-06,
+      "loss": 0.0291,
+      "step": 385
+    },
+    {
+      "epoch": 0.1340743313650573,
+      "grad_norm": 1.0118306100174919,
+      "learning_rate": 9.719682043463261e-06,
+      "loss": 0.0151,
+      "step": 386
+    },
+    {
+      "epoch": 0.13442167419242793,
+      "grad_norm": 1.1099679117319492,
+      "learning_rate": 9.717821739425286e-06,
+      "loss": 0.0257,
+      "step": 387
+    },
+    {
+      "epoch": 0.13476901701979854,
+      "grad_norm": 0.6925372348746138,
+      "learning_rate": 9.71595546214134e-06,
+      "loss": 0.0302,
+      "step": 388
+    },
+    {
+      "epoch": 0.13511635984716916,
+      "grad_norm": 1.1588103017208862,
+      "learning_rate": 9.714083213974322e-06,
+      "loss": 0.0159,
+      "step": 389
+    },
+    {
+      "epoch": 0.13546370267453978,
+      "grad_norm": 1.6463962782244659,
+      "learning_rate": 9.712204997294685e-06,
+      "loss": 0.028,
+      "step": 390
+    },
+    {
+      "epoch": 0.1358110455019104,
+      "grad_norm": 0.740625572838366,
+      "learning_rate": 9.710320814480448e-06,
+      "loss": 0.0291,
+      "step": 391
+    },
+    {
+      "epoch": 0.136158388329281,
+      "grad_norm": 1.461026830177916,
+      "learning_rate": 9.708430667917179e-06,
+      "loss": 0.0266,
+      "step": 392
+    },
+    {
+      "epoch": 0.13650573115665163,
+      "grad_norm": 1.173157945166447,
+      "learning_rate": 9.706534559997997e-06,
+      "loss": 0.0183,
+      "step": 393
+    },
+    {
+      "epoch": 0.13685307398402222,
+      "grad_norm": 2.7283819002594334,
+      "learning_rate": 9.704632493123569e-06,
+      "loss": 0.031,
+      "step": 394
+    },
+    {
+      "epoch": 0.13720041681139283,
+      "grad_norm": 0.804775724073791,
+      "learning_rate": 9.702724469702107e-06,
+      "loss": 0.0209,
+      "step": 395
+    },
+    {
+      "epoch": 0.13754775963876345,
+      "grad_norm": 0.7136569776548899,
+      "learning_rate": 9.700810492149364e-06,
+      "loss": 0.0247,
+      "step": 396
+    },
+    {
+      "epoch": 0.13789510246613407,
+      "grad_norm": 0.8364128273663843,
+      "learning_rate": 9.698890562888632e-06,
+      "loss": 0.0208,
+      "step": 397
+    },
+    {
+      "epoch": 0.13824244529350468,
+      "grad_norm": 1.0792662917432296,
+      "learning_rate": 9.696964684350738e-06,
+      "loss": 0.0283,
+      "step": 398
+    },
+    {
+      "epoch": 0.1385897881208753,
+      "grad_norm": 1.0704903922122424,
+      "learning_rate": 9.695032858974042e-06,
+      "loss": 0.0233,
+      "step": 399
+    },
+    {
+      "epoch": 0.13893713094824592,
+      "grad_norm": 0.8609262980285204,
+      "learning_rate": 9.693095089204431e-06,
+      "loss": 0.0277,
+      "step": 400
+    },
+    {
+      "epoch": 0.13928447377561654,
+      "grad_norm": 0.38166608517210565,
+      "learning_rate": 9.691151377495324e-06,
+      "loss": 0.0198,
+      "step": 401
+    },
+    {
+      "epoch": 0.13963181660298715,
+      "grad_norm": 0.5729627674226186,
+      "learning_rate": 9.689201726307655e-06,
+      "loss": 0.0318,
+      "step": 402
+    },
+    {
+      "epoch": 0.13997915943035777,
+      "grad_norm": 1.975870726288315,
+      "learning_rate": 9.687246138109888e-06,
+      "loss": 0.0322,
+      "step": 403
+    },
+    {
+      "epoch": 0.14032650225772839,
+      "grad_norm": 2.1898213167704648,
+      "learning_rate": 9.68528461537799e-06,
+      "loss": 0.0304,
+      "step": 404
+    },
+    {
+      "epoch": 0.140673845085099,
+      "grad_norm": 0.620886516690147,
+      "learning_rate": 9.683317160595457e-06,
+      "loss": 0.0178,
+      "step": 405
+    },
+    {
+      "epoch": 0.14102118791246962,
+      "grad_norm": 0.751397853860502,
+      "learning_rate": 9.681343776253284e-06,
+      "loss": 0.0227,
+      "step": 406
+    },
+    {
+      "epoch": 0.1413685307398402,
+      "grad_norm": 1.5360568776293804,
+      "learning_rate": 9.679364464849983e-06,
+      "loss": 0.024,
+      "step": 407
+    },
+    {
+      "epoch": 0.14171587356721083,
+      "grad_norm": 1.9507404317073442,
+      "learning_rate": 9.67737922889156e-06,
+      "loss": 0.0299,
+      "step": 408
+    },
+    {
+      "epoch": 0.14206321639458144,
+      "grad_norm": 0.3885448349712659,
+      "learning_rate": 9.675388070891527e-06,
+      "loss": 0.0169,
+      "step": 409
+    },
+    {
+      "epoch": 0.14241055922195206,
+      "grad_norm": 0.7487805008765015,
+      "learning_rate": 9.6733909933709e-06,
+      "loss": 0.0258,
+      "step": 410
+    },
+    {
+      "epoch": 0.14275790204932268,
+      "grad_norm": 0.8741306809093075,
+      "learning_rate": 9.671387998858178e-06,
+      "loss": 0.029,
+      "step": 411
+    },
+    {
+      "epoch": 0.1431052448766933,
+      "grad_norm": 0.7791251348925693,
+      "learning_rate": 9.669379089889361e-06,
+      "loss": 0.0287,
+      "step": 412
+    },
+    {
+      "epoch": 0.1434525877040639,
+      "grad_norm": 1.6127960765398552,
+      "learning_rate": 9.66736426900793e-06,
+      "loss": 0.0207,
+      "step": 413
+    },
+    {
+      "epoch": 0.14379993053143453,
+      "grad_norm": 2.360372239374618,
+      "learning_rate": 9.66534353876486e-06,
+      "loss": 0.0257,
+      "step": 414
+    },
+    {
+      "epoch": 0.14414727335880514,
+      "grad_norm": 1.0591550444954392,
+      "learning_rate": 9.663316901718599e-06,
+      "loss": 0.0252,
+      "step": 415
+    },
+    {
+      "epoch": 0.14449461618617576,
+      "grad_norm": 0.6279163360317281,
+      "learning_rate": 9.661284360435075e-06,
+      "loss": 0.0168,
+      "step": 416
+    },
+    {
+      "epoch": 0.14484195901354638,
+      "grad_norm": 0.9205829596021957,
+      "learning_rate": 9.659245917487698e-06,
+      "loss": 0.028,
+      "step": 417
+    },
+    {
+      "epoch": 0.145189301840917,
+      "grad_norm": 0.7205918553309815,
+      "learning_rate": 9.657201575457346e-06,
+      "loss": 0.0239,
+      "step": 418
+    },
+    {
+      "epoch": 0.1455366446682876,
+      "grad_norm": 0.4050107455693852,
+      "learning_rate": 9.655151336932362e-06,
+      "loss": 0.0152,
+      "step": 419
+    },
+    {
+      "epoch": 0.1458839874956582,
+      "grad_norm": 1.5415577326332006,
+      "learning_rate": 9.653095204508562e-06,
+      "loss": 0.0222,
+      "step": 420
+    },
+    {
+      "epoch": 0.14623133032302882,
+      "grad_norm": 0.4578806435756731,
+      "learning_rate": 9.651033180789216e-06,
+      "loss": 0.016,
+      "step": 421
+    },
+    {
+      "epoch": 0.14657867315039944,
+      "grad_norm": 0.8614555068468486,
+      "learning_rate": 9.648965268385062e-06,
+      "loss": 0.0257,
+      "step": 422
+    },
+    {
+      "epoch": 0.14692601597777005,
+      "grad_norm": 1.470299031527,
+      "learning_rate": 9.646891469914285e-06,
+      "loss": 0.0368,
+      "step": 423
+    },
+    {
+      "epoch": 0.14727335880514067,
+      "grad_norm": 1.0949073778594718,
+      "learning_rate": 9.644811788002531e-06,
+      "loss": 0.033,
+      "step": 424
+    },
+    {
+      "epoch": 0.1476207016325113,
+      "grad_norm": 0.7401125451520727,
+      "learning_rate": 9.642726225282886e-06,
+      "loss": 0.0288,
+      "step": 425
+    },
+    {
+      "epoch": 0.1479680444598819,
+      "grad_norm": 2.2131411209283565,
+      "learning_rate": 9.64063478439589e-06,
+      "loss": 0.0175,
+      "step": 426
+    },
+    {
+      "epoch": 0.14831538728725252,
+      "grad_norm": 2.06363585907779,
+      "learning_rate": 9.638537467989517e-06,
+      "loss": 0.0276,
+      "step": 427
+    },
+    {
+      "epoch": 0.14866273011462314,
+      "grad_norm": 1.467812889206939,
+      "learning_rate": 9.63643427871919e-06,
+      "loss": 0.0291,
+      "step": 428
+    },
+    {
+      "epoch": 0.14901007294199375,
+      "grad_norm": 0.4505213888594297,
+      "learning_rate": 9.634325219247758e-06,
+      "loss": 0.0153,
+      "step": 429
+    },
+    {
+      "epoch": 0.14935741576936437,
+      "grad_norm": 1.3450924566368194,
+      "learning_rate": 9.632210292245508e-06,
+      "loss": 0.0225,
+      "step": 430
+    },
+    {
+      "epoch": 0.149704758596735,
+      "grad_norm": 2.347870224615922,
+      "learning_rate": 9.630089500390154e-06,
+      "loss": 0.028,
+      "step": 431
+    },
+    {
+      "epoch": 0.1500521014241056,
+      "grad_norm": 1.6462241884720297,
+      "learning_rate": 9.627962846366838e-06,
+      "loss": 0.024,
+      "step": 432
+    },
+    {
+      "epoch": 0.1503994442514762,
+      "grad_norm": 0.39196556018543693,
+      "learning_rate": 9.62583033286812e-06,
+      "loss": 0.0201,
+      "step": 433
+    },
+    {
+      "epoch": 0.1507467870788468,
+      "grad_norm": 1.5059296209518138,
+      "learning_rate": 9.62369196259398e-06,
+      "loss": 0.02,
+      "step": 434
+    },
+    {
+      "epoch": 0.15109412990621743,
+      "grad_norm": 0.45555178968971705,
+      "learning_rate": 9.621547738251816e-06,
+      "loss": 0.0221,
+      "step": 435
+    },
+    {
+      "epoch": 0.15144147273358805,
+      "grad_norm": 0.2882865925387099,
+      "learning_rate": 9.619397662556434e-06,
+      "loss": 0.0128,
+      "step": 436
+    },
+    {
+      "epoch": 0.15178881556095866,
+      "grad_norm": 0.6918769536905994,
+      "learning_rate": 9.617241738230051e-06,
+      "loss": 0.0242,
+      "step": 437
+    },
+    {
+      "epoch": 0.15213615838832928,
+      "grad_norm": 1.6957528332450178,
+      "learning_rate": 9.61507996800229e-06,
+      "loss": 0.0313,
+      "step": 438
+    },
+    {
+      "epoch": 0.1524835012156999,
+      "grad_norm": 1.317146675116354,
+      "learning_rate": 9.61291235461017e-06,
+      "loss": 0.0215,
+      "step": 439
+    },
+    {
+      "epoch": 0.1528308440430705,
+      "grad_norm": 0.727706359709951,
+      "learning_rate": 9.610738900798116e-06,
+      "loss": 0.0271,
+      "step": 440
+    },
+    {
+      "epoch": 0.15317818687044113,
+      "grad_norm": 3.5123826215661365,
+      "learning_rate": 9.60855960931794e-06,
+      "loss": 0.0383,
+      "step": 441
+    },
+    {
+      "epoch": 0.15352552969781175,
+      "grad_norm": 3.728862284149442,
+      "learning_rate": 9.606374482928849e-06,
+      "loss": 0.0318,
+      "step": 442
+    },
+    {
+      "epoch": 0.15387287252518236,
+      "grad_norm": 2.760620871544741,
+      "learning_rate": 9.604183524397439e-06,
+      "loss": 0.041,
+      "step": 443
+    },
+    {
+      "epoch": 0.15422021535255298,
+      "grad_norm": 3.1069152542667604,
+      "learning_rate": 9.601986736497686e-06,
+      "loss": 0.0325,
+      "step": 444
+    },
+    {
+      "epoch": 0.1545675581799236,
+      "grad_norm": 1.8573780055969038,
+      "learning_rate": 9.59978412201095e-06,
+      "loss": 0.0317,
+      "step": 445
+    },
+    {
+      "epoch": 0.1549149010072942,
+      "grad_norm": 1.0027964210682174,
+      "learning_rate": 9.597575683725965e-06,
+      "loss": 0.0227,
+      "step": 446
+    },
+    {
+      "epoch": 0.1552622438346648,
+      "grad_norm": 1.4276269980988359,
+      "learning_rate": 9.595361424438841e-06,
+      "loss": 0.0362,
+      "step": 447
+    },
+    {
+      "epoch": 0.15560958666203542,
+      "grad_norm": 1.317331217220333,
+      "learning_rate": 9.593141346953059e-06,
+      "loss": 0.0226,
+      "step": 448
+    },
+    {
+      "epoch": 0.15595692948940604,
+      "grad_norm": 0.5933495170717192,
+      "learning_rate": 9.590915454079463e-06,
+      "loss": 0.0227,
+      "step": 449
+    },
+    {
+      "epoch": 0.15630427231677665,
+      "grad_norm": 0.4622800957314312,
+      "learning_rate": 9.588683748636262e-06,
+      "loss": 0.0267,
+      "step": 450
+    },
+    {
+      "epoch": 0.15665161514414727,
+      "grad_norm": 0.7911625488227271,
+      "learning_rate": 9.586446233449024e-06,
+      "loss": 0.0246,
+      "step": 451
+    },
+    {
+      "epoch": 0.1569989579715179,
+      "grad_norm": 0.9702252686758073,
+      "learning_rate": 9.584202911350672e-06,
+      "loss": 0.0236,
+      "step": 452
+    },
+    {
+      "epoch": 0.1573463007988885,
+      "grad_norm": 0.6033941783861664,
+      "learning_rate": 9.581953785181482e-06,
+      "loss": 0.0214,
+      "step": 453
+    },
+    {
+      "epoch": 0.15769364362625912,
+      "grad_norm": 0.4655527204441117,
+      "learning_rate": 9.579698857789078e-06,
+      "loss": 0.0283,
+      "step": 454
+    },
+    {
+      "epoch": 0.15804098645362974,
+      "grad_norm": 0.5742266010042084,
+      "learning_rate": 9.577438132028431e-06,
+      "loss": 0.0313,
+      "step": 455
+    },
+    {
+      "epoch": 0.15838832928100036,
+      "grad_norm": 0.5365426672112383,
+      "learning_rate": 9.575171610761848e-06,
+      "loss": 0.0281,
+      "step": 456
+    },
+    {
+      "epoch": 0.15873567210837097,
+      "grad_norm": 0.4233818161136484,
+      "learning_rate": 9.572899296858981e-06,
+      "loss": 0.0315,
+      "step": 457
+    },
+    {
+      "epoch": 0.1590830149357416,
+      "grad_norm": 0.5680963519098914,
+      "learning_rate": 9.570621193196811e-06,
+      "loss": 0.0269,
+      "step": 458
+    },
+    {
+      "epoch": 0.15943035776311218,
+      "grad_norm": 0.5125465753073855,
+      "learning_rate": 9.568337302659652e-06,
+      "loss": 0.0167,
+      "step": 459
+    },
+    {
+      "epoch": 0.1597777005904828,
+      "grad_norm": 0.3119364502263281,
+      "learning_rate": 9.566047628139142e-06,
+      "loss": 0.021,
+      "step": 460
+    },
+    {
+      "epoch": 0.1601250434178534,
+      "grad_norm": 0.9906415221821926,
+      "learning_rate": 9.563752172534242e-06,
+      "loss": 0.0235,
+      "step": 461
+    },
+    {
+      "epoch": 0.16047238624522403,
+      "grad_norm": 0.8066115586763946,
+      "learning_rate": 9.561450938751238e-06,
+      "loss": 0.0229,
+      "step": 462
+    },
+    {
+      "epoch": 0.16081972907259465,
+      "grad_norm": 1.005027835921377,
+      "learning_rate": 9.559143929703724e-06,
+      "loss": 0.0203,
+      "step": 463
+    },
+    {
+      "epoch": 0.16116707189996526,
+      "grad_norm": 0.32646704240882035,
+      "learning_rate": 9.556831148312612e-06,
+      "loss": 0.0183,
+      "step": 464
+    },
+    {
+      "epoch": 0.16151441472733588,
+      "grad_norm": 0.8970575237008778,
+      "learning_rate": 9.554512597506122e-06,
+      "loss": 0.0237,
+      "step": 465
+    },
+    {
+      "epoch": 0.1618617575547065,
+      "grad_norm": 1.9756358587933307,
+      "learning_rate": 9.552188280219773e-06,
+      "loss": 0.0232,
+      "step": 466
+    },
+    {
+      "epoch": 0.16220910038207711,
+      "grad_norm": 0.659327280239366,
+      "learning_rate": 9.549858199396394e-06,
+      "loss": 0.0154,
+      "step": 467
+    },
+    {
+      "epoch": 0.16255644320944773,
+      "grad_norm": 0.9285348605348003,
+      "learning_rate": 9.547522357986102e-06,
+      "loss": 0.0254,
+      "step": 468
+    },
+    {
+      "epoch": 0.16290378603681835,
+      "grad_norm": 0.8854039259165357,
+      "learning_rate": 9.545180758946312e-06,
+      "loss": 0.023,
+      "step": 469
+    },
+    {
+      "epoch": 0.16325112886418897,
+      "grad_norm": 0.7572168628873044,
+      "learning_rate": 9.542833405241729e-06,
+      "loss": 0.0229,
+      "step": 470
+    },
+    {
+      "epoch": 0.16359847169155958,
+      "grad_norm": 1.1833229445171547,
+      "learning_rate": 9.540480299844345e-06,
+      "loss": 0.0194,
+      "step": 471
+    },
+    {
+      "epoch": 0.16394581451893017,
+      "grad_norm": 1.128447523594609,
+      "learning_rate": 9.538121445733431e-06,
+      "loss": 0.0312,
+      "step": 472
+    },
+    {
+      "epoch": 0.1642931573463008,
+      "grad_norm": 0.8646345138050862,
+      "learning_rate": 9.53575684589554e-06,
+      "loss": 0.034,
+      "step": 473
+    },
+    {
+      "epoch": 0.1646405001736714,
+      "grad_norm": 2.142392363566201,
+      "learning_rate": 9.533386503324495e-06,
+      "loss": 0.0267,
+      "step": 474
+    },
+    {
+      "epoch": 0.16498784300104202,
+      "grad_norm": 0.31920633758411143,
+      "learning_rate": 9.531010421021396e-06,
+      "loss": 0.0264,
+      "step": 475
+    },
+    {
+      "epoch": 0.16533518582841264,
+      "grad_norm": 1.1720964352470096,
+      "learning_rate": 9.528628601994603e-06,
+      "loss": 0.0203,
+      "step": 476
+    },
+    {
+      "epoch": 0.16568252865578326,
+      "grad_norm": 0.3038247546428463,
+      "learning_rate": 9.526241049259746e-06,
+      "loss": 0.0178,
+      "step": 477
+    },
+    {
+      "epoch": 0.16602987148315387,
+      "grad_norm": 1.4917694833284774,
+      "learning_rate": 9.523847765839712e-06,
+      "loss": 0.0242,
+      "step": 478
+    },
+    {
+      "epoch": 0.1663772143105245,
+      "grad_norm": 0.818082889175988,
+      "learning_rate": 9.52144875476464e-06,
+      "loss": 0.0222,
+      "step": 479
+    },
+    {
+      "epoch": 0.1667245571378951,
+      "grad_norm": 0.3877555907825711,
+      "learning_rate": 9.519044019071926e-06,
+      "loss": 0.0243,
+      "step": 480
+    },
+    {
+      "epoch": 0.16707189996526572,
+      "grad_norm": 0.4639811409245256,
+      "learning_rate": 9.51663356180621e-06,
+      "loss": 0.0251,
+      "step": 481
+    },
+    {
+      "epoch": 0.16741924279263634,
+      "grad_norm": 0.8099085137002117,
+      "learning_rate": 9.514217386019381e-06,
+      "loss": 0.0303,
+      "step": 482
+    },
+    {
+      "epoch": 0.16776658562000696,
+      "grad_norm": 0.9835471538813636,
+      "learning_rate": 9.511795494770563e-06,
+      "loss": 0.0259,
+      "step": 483
+    },
+    {
+      "epoch": 0.16811392844737758,
+      "grad_norm": 0.6270381501097829,
+      "learning_rate": 9.509367891126122e-06,
+      "loss": 0.0159,
+      "step": 484
+    },
+    {
+      "epoch": 0.16846127127474816,
+      "grad_norm": 0.4636485528169751,
+      "learning_rate": 9.506934578159648e-06,
+      "loss": 0.0238,
+      "step": 485
+    },
+    {
+      "epoch": 0.16880861410211878,
+      "grad_norm": 0.6141677857665021,
+      "learning_rate": 9.50449555895197e-06,
+      "loss": 0.0204,
+      "step": 486
+    },
+    {
+      "epoch": 0.1691559569294894,
+      "grad_norm": 0.9292538654705289,
+      "learning_rate": 9.50205083659113e-06,
+      "loss": 0.0327,
+      "step": 487
+    },
+    {
+      "epoch": 0.16950329975686002,
+      "grad_norm": 1.1196044763325756,
+      "learning_rate": 9.499600414172402e-06,
+      "loss": 0.0222,
+      "step": 488
+    },
+    {
+      "epoch": 0.16985064258423063,
+      "grad_norm": 1.5976292188516854,
+      "learning_rate": 9.49714429479827e-06,
+      "loss": 0.0324,
+      "step": 489
+    },
+    {
+      "epoch": 0.17019798541160125,
+      "grad_norm": 1.2857006123730916,
+      "learning_rate": 9.494682481578436e-06,
+      "loss": 0.0205,
+      "step": 490
+    },
+    {
+      "epoch": 0.17054532823897187,
+      "grad_norm": 1.276460947257805,
+      "learning_rate": 9.492214977629804e-06,
+      "loss": 0.016,
+      "step": 491
+    },
+    {
+      "epoch": 0.17089267106634248,
+      "grad_norm": 1.252602259656139,
+      "learning_rate": 9.489741786076488e-06,
+      "loss": 0.0237,
+      "step": 492
+    },
+    {
+      "epoch": 0.1712400138937131,
+      "grad_norm": 0.5734672626180113,
+      "learning_rate": 9.487262910049804e-06,
+      "loss": 0.017,
+      "step": 493
+    },
+    {
+      "epoch": 0.17158735672108372,
+      "grad_norm": 1.2737930527140826,
+      "learning_rate": 9.48477835268826e-06,
+      "loss": 0.0231,
+      "step": 494
+    },
+    {
+      "epoch": 0.17193469954845433,
+      "grad_norm": 0.9557221623397816,
+      "learning_rate": 9.482288117137561e-06,
+      "loss": 0.0186,
+      "step": 495
+    },
+    {
+      "epoch": 0.17228204237582495,
+      "grad_norm": 1.7629581363677385,
+      "learning_rate": 9.479792206550604e-06,
+      "loss": 0.0333,
+      "step": 496
+    },
+    {
+      "epoch": 0.17262938520319557,
+      "grad_norm": 0.4160720670777094,
+      "learning_rate": 9.477290624087464e-06,
+      "loss": 0.0207,
+      "step": 497
+    },
+    {
+      "epoch": 0.17297672803056616,
+      "grad_norm": 1.657594987733875,
+      "learning_rate": 9.474783372915401e-06,
+      "loss": 0.0187,
+      "step": 498
+    },
+    {
+      "epoch": 0.17332407085793677,
+      "grad_norm": 0.6777171710639321,
+      "learning_rate": 9.472270456208856e-06,
+      "loss": 0.0311,
+      "step": 499
+    },
+    {
+      "epoch": 0.1736714136853074,
+      "grad_norm": 0.5649689115711107,
+      "learning_rate": 9.469751877149434e-06,
+      "loss": 0.0319,
+      "step": 500
+    },
+    {
+      "epoch": 0.174018756512678,
+      "grad_norm": 1.440887458216018,
+      "learning_rate": 9.467227638925917e-06,
+      "loss": 0.0314,
+      "step": 501
+    },
+    {
+      "epoch": 0.17436609934004862,
+      "grad_norm": 1.4716216093022831,
+      "learning_rate": 9.464697744734248e-06,
+      "loss": 0.0228,
+      "step": 502
+    },
+    {
+      "epoch": 0.17471344216741924,
+      "grad_norm": 0.8445232330484052,
+      "learning_rate": 9.462162197777533e-06,
+      "loss": 0.0298,
+      "step": 503
+    },
+    {
+      "epoch": 0.17506078499478986,
+      "grad_norm": 0.8274867059781084,
+      "learning_rate": 9.459621001266036e-06,
+      "loss": 0.0253,
+      "step": 504
+    },
+    {
+      "epoch": 0.17540812782216048,
+      "grad_norm": 1.3093883304254912,
+      "learning_rate": 9.45707415841717e-06,
+      "loss": 0.0284,
+      "step": 505
+    },
+    {
+      "epoch": 0.1757554706495311,
+      "grad_norm": 0.4071291730241176,
+      "learning_rate": 9.454521672455501e-06,
+      "loss": 0.0235,
+      "step": 506
+    },
+    {
+      "epoch": 0.1761028134769017,
+      "grad_norm": 1.1764547349254069,
+      "learning_rate": 9.451963546612737e-06,
+      "loss": 0.0195,
+      "step": 507
+    },
+    {
+      "epoch": 0.17645015630427233,
+      "grad_norm": 0.6569963887274749,
+      "learning_rate": 9.449399784127726e-06,
+      "loss": 0.0264,
+      "step": 508
+    },
+    {
+      "epoch": 0.17679749913164294,
+      "grad_norm": 0.5238198578448529,
+      "learning_rate": 9.446830388246457e-06,
+      "loss": 0.0194,
+      "step": 509
+    },
+    {
+      "epoch": 0.17714484195901353,
+      "grad_norm": 0.7417799180982843,
+      "learning_rate": 9.444255362222046e-06,
+      "loss": 0.0247,
+      "step": 510
+    },
+    {
+      "epoch": 0.17749218478638415,
+      "grad_norm": 1.3703456770005742,
+      "learning_rate": 9.441674709314743e-06,
+      "loss": 0.0274,
+      "step": 511
+    },
+    {
+      "epoch": 0.17783952761375477,
+      "grad_norm": 1.4652033192638536,
+      "learning_rate": 9.439088432791916e-06,
+      "loss": 0.0139,
+      "step": 512
+    },
+    {
+      "epoch": 0.17818687044112538,
+      "grad_norm": 1.5832446589993328,
+      "learning_rate": 9.436496535928057e-06,
+      "loss": 0.034,
+      "step": 513
+    },
+    {
+      "epoch": 0.178534213268496,
+      "grad_norm": 1.9119409177551294,
+      "learning_rate": 9.433899022004774e-06,
+      "loss": 0.0251,
+      "step": 514
+    },
+    {
+      "epoch": 0.17888155609586662,
+      "grad_norm": 0.788548038638029,
+      "learning_rate": 9.431295894310786e-06,
+      "loss": 0.0308,
+      "step": 515
+    },
+    {
+      "epoch": 0.17922889892323723,
+      "grad_norm": 0.8067843750520071,
+      "learning_rate": 9.428687156141919e-06,
+      "loss": 0.0234,
+      "step": 516
+    },
+    {
+      "epoch": 0.17957624175060785,
+      "grad_norm": 1.1709927168009244,
+      "learning_rate": 9.426072810801104e-06,
+      "loss": 0.0319,
+      "step": 517
+    },
+    {
+      "epoch": 0.17992358457797847,
+      "grad_norm": 0.6216918250812782,
+      "learning_rate": 9.423452861598367e-06,
+      "loss": 0.0243,
+      "step": 518
+    },
+    {
+      "epoch": 0.18027092740534909,
+      "grad_norm": 0.8872099116597142,
+      "learning_rate": 9.420827311850836e-06,
+      "loss": 0.0252,
+      "step": 519
+    },
+    {
+      "epoch": 0.1806182702327197,
+      "grad_norm": 0.586557224997833,
+      "learning_rate": 9.418196164882725e-06,
+      "loss": 0.0267,
+      "step": 520
+    },
+    {
+      "epoch": 0.18096561306009032,
+      "grad_norm": 0.5527086813309143,
+      "learning_rate": 9.415559424025335e-06,
+      "loss": 0.0212,
+      "step": 521
+    },
+    {
+      "epoch": 0.18131295588746094,
+      "grad_norm": 0.6211483589771798,
+      "learning_rate": 9.41291709261705e-06,
+      "loss": 0.018,
+      "step": 522
+    },
+    {
+      "epoch": 0.18166029871483153,
+      "grad_norm": 0.5607071996617911,
+      "learning_rate": 9.410269174003333e-06,
+      "loss": 0.0208,
+      "step": 523
+    },
+    {
+      "epoch": 0.18200764154220214,
+      "grad_norm": 1.8330990714819673,
+      "learning_rate": 9.407615671536723e-06,
+      "loss": 0.0306,
+      "step": 524
+    },
+    {
+      "epoch": 0.18235498436957276,
+      "grad_norm": 0.7432517423995653,
+      "learning_rate": 9.404956588576822e-06,
+      "loss": 0.0266,
+      "step": 525
+    },
+    {
+      "epoch": 0.18270232719694338,
+      "grad_norm": 0.5660339296085843,
+      "learning_rate": 9.402291928490302e-06,
+      "loss": 0.0184,
+      "step": 526
+    },
+    {
+      "epoch": 0.183049670024314,
+      "grad_norm": 0.7219272711113525,
+      "learning_rate": 9.399621694650898e-06,
+      "loss": 0.0285,
+      "step": 527
+    },
+    {
+      "epoch": 0.1833970128516846,
+      "grad_norm": 1.778600324374738,
+      "learning_rate": 9.3969458904394e-06,
+      "loss": 0.0345,
+      "step": 528
+    },
+    {
+      "epoch": 0.18374435567905523,
+      "grad_norm": 1.6387273409591696,
+      "learning_rate": 9.394264519243649e-06,
+      "loss": 0.0273,
+      "step": 529
+    },
+    {
+      "epoch": 0.18409169850642584,
+      "grad_norm": 0.8842030070616902,
+      "learning_rate": 9.391577584458536e-06,
+      "loss": 0.0203,
+      "step": 530
+    },
+    {
+      "epoch": 0.18443904133379646,
+      "grad_norm": 0.9269965797470187,
+      "learning_rate": 9.388885089485995e-06,
+      "loss": 0.018,
+      "step": 531
+    },
+    {
+      "epoch": 0.18478638416116708,
+      "grad_norm": 0.47186675698812003,
+      "learning_rate": 9.386187037735004e-06,
+      "loss": 0.0208,
+      "step": 532
+    },
+    {
+      "epoch": 0.1851337269885377,
+      "grad_norm": 0.6914305009354391,
+      "learning_rate": 9.383483432621569e-06,
+      "loss": 0.021,
+      "step": 533
+    },
+    {
+      "epoch": 0.1854810698159083,
+      "grad_norm": 0.5583142545579645,
+      "learning_rate": 9.380774277568733e-06,
+      "loss": 0.0169,
+      "step": 534
+    },
+    {
+      "epoch": 0.18582841264327893,
+      "grad_norm": 0.5064603142442761,
+      "learning_rate": 9.378059576006567e-06,
+      "loss": 0.0263,
+      "step": 535
+    },
+    {
+      "epoch": 0.18617575547064952,
+      "grad_norm": 0.5330806328958541,
+      "learning_rate": 9.375339331372155e-06,
+      "loss": 0.0191,
+      "step": 536
+    },
+    {
+      "epoch": 0.18652309829802013,
+      "grad_norm": 1.945098806666559,
+      "learning_rate": 9.37261354710961e-06,
+      "loss": 0.0291,
+      "step": 537
+    },
+    {
+      "epoch": 0.18687044112539075,
+      "grad_norm": 2.044217190353534,
+      "learning_rate": 9.369882226670054e-06,
+      "loss": 0.0258,
+      "step": 538
+    },
+    {
+      "epoch": 0.18721778395276137,
+      "grad_norm": 0.6771224467650478,
+      "learning_rate": 9.36714537351162e-06,
+      "loss": 0.0168,
+      "step": 539
+    },
+    {
+      "epoch": 0.18756512678013199,
+      "grad_norm": 0.695744615748055,
+      "learning_rate": 9.36440299109944e-06,
+      "loss": 0.015,
+      "step": 540
+    },
+    {
+      "epoch": 0.1879124696075026,
+      "grad_norm": 0.4503395885743224,
+      "learning_rate": 9.361655082905654e-06,
+      "loss": 0.013,
+      "step": 541
+    },
+    {
+      "epoch": 0.18825981243487322,
+      "grad_norm": 1.0753578393601884,
+      "learning_rate": 9.358901652409398e-06,
+      "loss": 0.018,
+      "step": 542
+    },
+    {
+      "epoch": 0.18860715526224384,
+      "grad_norm": 0.49649547384514064,
+      "learning_rate": 9.356142703096793e-06,
+      "loss": 0.0147,
+      "step": 543
+    },
+    {
+      "epoch": 0.18895449808961445,
+      "grad_norm": 1.5646023573061925,
+      "learning_rate": 9.353378238460955e-06,
+      "loss": 0.0237,
+      "step": 544
+    },
+    {
+      "epoch": 0.18930184091698507,
+      "grad_norm": 0.7887255494483407,
+      "learning_rate": 9.350608262001978e-06,
+      "loss": 0.0175,
+      "step": 545
+    },
+    {
+      "epoch": 0.1896491837443557,
+      "grad_norm": 1.0681506636597586,
+      "learning_rate": 9.347832777226936e-06,
+      "loss": 0.0186,
+      "step": 546
+    },
+    {
+      "epoch": 0.1899965265717263,
+      "grad_norm": 0.7785556397165594,
+      "learning_rate": 9.345051787649877e-06,
+      "loss": 0.0184,
+      "step": 547
+    },
+    {
+      "epoch": 0.19034386939909692,
+      "grad_norm": 0.509526910733021,
+      "learning_rate": 9.34226529679182e-06,
+      "loss": 0.0225,
+      "step": 548
+    },
+    {
+      "epoch": 0.1906912122264675,
+      "grad_norm": 1.319059024617917,
+      "learning_rate": 9.339473308180746e-06,
+      "loss": 0.0214,
+      "step": 549
+    },
+    {
+      "epoch": 0.19103855505383813,
+      "grad_norm": 0.5224211698738431,
+      "learning_rate": 9.336675825351602e-06,
+      "loss": 0.0257,
+      "step": 550
+    },
+    {
+      "epoch": 0.19138589788120874,
+      "grad_norm": 1.1460327083281117,
+      "learning_rate": 9.333872851846285e-06,
+      "loss": 0.0255,
+      "step": 551
+    },
+    {
+      "epoch": 0.19173324070857936,
+      "grad_norm": 0.7498296132522941,
+      "learning_rate": 9.33106439121365e-06,
+      "loss": 0.0262,
+      "step": 552
+    },
+    {
+      "epoch": 0.19208058353594998,
+      "grad_norm": 0.38935101782706116,
+      "learning_rate": 9.328250447009493e-06,
+      "loss": 0.0147,
+      "step": 553
+    },
+    {
+      "epoch": 0.1924279263633206,
+      "grad_norm": 0.8699070629651166,
+      "learning_rate": 9.325431022796559e-06,
+      "loss": 0.0213,
+      "step": 554
+    },
+    {
+      "epoch": 0.1927752691906912,
+      "grad_norm": 0.8664760328709794,
+      "learning_rate": 9.322606122144524e-06,
+      "loss": 0.0236,
+      "step": 555
+    },
+    {
+      "epoch": 0.19312261201806183,
+      "grad_norm": 2.250608635056857,
+      "learning_rate": 9.319775748630004e-06,
+      "loss": 0.0283,
+      "step": 556
+    },
+    {
+      "epoch": 0.19346995484543245,
+      "grad_norm": 1.7828983126970328,
+      "learning_rate": 9.316939905836543e-06,
+      "loss": 0.0182,
+      "step": 557
+    },
+    {
+      "epoch": 0.19381729767280306,
+      "grad_norm": 0.5137189754751468,
+      "learning_rate": 9.314098597354608e-06,
+      "loss": 0.0236,
+      "step": 558
+    },
+    {
+      "epoch": 0.19416464050017368,
+      "grad_norm": 0.6266424659785979,
+      "learning_rate": 9.311251826781587e-06,
+      "loss": 0.0228,
+      "step": 559
+    },
+    {
+      "epoch": 0.1945119833275443,
+      "grad_norm": 1.662923917212151,
+      "learning_rate": 9.308399597721782e-06,
+      "loss": 0.0232,
+      "step": 560
+    },
+    {
+      "epoch": 0.1948593261549149,
+      "grad_norm": 0.7492399109293421,
+      "learning_rate": 9.305541913786409e-06,
+      "loss": 0.0224,
+      "step": 561
+    },
+    {
+      "epoch": 0.1952066689822855,
+      "grad_norm": 0.5213910576317119,
+      "learning_rate": 9.302678778593586e-06,
+      "loss": 0.0232,
+      "step": 562
+    },
+    {
+      "epoch": 0.19555401180965612,
+      "grad_norm": 1.4562306345231064,
+      "learning_rate": 9.299810195768341e-06,
+      "loss": 0.0282,
+      "step": 563
+    },
+    {
+      "epoch": 0.19590135463702674,
+      "grad_norm": 0.9407546811393638,
+      "learning_rate": 9.296936168942589e-06,
+      "loss": 0.0199,
+      "step": 564
+    },
+    {
+      "epoch": 0.19624869746439735,
+      "grad_norm": 0.8454850863651415,
+      "learning_rate": 9.294056701755144e-06,
+      "loss": 0.0169,
+      "step": 565
+    },
+    {
+      "epoch": 0.19659604029176797,
+      "grad_norm": 0.27061382482871477,
+      "learning_rate": 9.291171797851708e-06,
+      "loss": 0.0131,
+      "step": 566
+    },
+    {
+      "epoch": 0.1969433831191386,
+      "grad_norm": 0.49862269994808495,
+      "learning_rate": 9.288281460884864e-06,
+      "loss": 0.0171,
+      "step": 567
+    },
+    {
+      "epoch": 0.1972907259465092,
+      "grad_norm": 0.8522107595930313,
+      "learning_rate": 9.285385694514075e-06,
+      "loss": 0.0218,
+      "step": 568
+    },
+    {
+      "epoch": 0.19763806877387982,
+      "grad_norm": 0.8148444399323233,
+      "learning_rate": 9.282484502405677e-06,
+      "loss": 0.021,
+      "step": 569
+    },
+    {
+      "epoch": 0.19798541160125044,
+      "grad_norm": 1.375134823060513,
+      "learning_rate": 9.27957788823288e-06,
+      "loss": 0.0298,
+      "step": 570
+    },
+    {
+      "epoch": 0.19833275442862106,
+      "grad_norm": 1.1166507276668756,
+      "learning_rate": 9.276665855675751e-06,
+      "loss": 0.0186,
+      "step": 571
+    },
+    {
+      "epoch": 0.19868009725599167,
+      "grad_norm": 1.3330794003539927,
+      "learning_rate": 9.273748408421224e-06,
+      "loss": 0.0254,
+      "step": 572
+    },
+    {
+      "epoch": 0.1990274400833623,
+      "grad_norm": 0.369557234187912,
+      "learning_rate": 9.270825550163088e-06,
+      "loss": 0.0144,
+      "step": 573
+    },
+    {
+      "epoch": 0.1993747829107329,
+      "grad_norm": 0.5934833218153897,
+      "learning_rate": 9.267897284601976e-06,
+      "loss": 0.0234,
+      "step": 574
+    },
+    {
+      "epoch": 0.1997221257381035,
+      "grad_norm": 0.5376211438611554,
+      "learning_rate": 9.264963615445378e-06,
+      "loss": 0.0159,
+      "step": 575
+    },
+    {
+      "epoch": 0.2000694685654741,
+      "grad_norm": 0.5613556695221377,
+      "learning_rate": 9.26202454640762e-06,
+      "loss": 0.0193,
+      "step": 576
+    },
+    {
+      "epoch": 0.20041681139284473,
+      "grad_norm": 0.8368239688441027,
+      "learning_rate": 9.259080081209861e-06,
+      "loss": 0.023,
+      "step": 577
+    },
+    {
+      "epoch": 0.20076415422021535,
+      "grad_norm": 1.2341577296539838,
+      "learning_rate": 9.256130223580096e-06,
+      "loss": 0.0198,
+      "step": 578
+    },
+    {
+      "epoch": 0.20111149704758596,
+      "grad_norm": 0.3531135473026086,
+      "learning_rate": 9.25317497725315e-06,
+      "loss": 0.0081,
+      "step": 579
+    },
+    {
+      "epoch": 0.20145883987495658,
+      "grad_norm": 0.7328484748799605,
+      "learning_rate": 9.250214345970665e-06,
+      "loss": 0.0184,
+      "step": 580
+    },
+    {
+      "epoch": 0.2018061827023272,
+      "grad_norm": 1.9853290246566473,
+      "learning_rate": 9.247248333481105e-06,
+      "loss": 0.0214,
+      "step": 581
+    },
+    {
+      "epoch": 0.20215352552969781,
+      "grad_norm": 1.4811646122579638,
+      "learning_rate": 9.244276943539746e-06,
+      "loss": 0.0241,
+      "step": 582
+    },
+    {
+      "epoch": 0.20250086835706843,
+      "grad_norm": 1.272707363559077,
+      "learning_rate": 9.241300179908672e-06,
+      "loss": 0.0339,
+      "step": 583
+    },
+    {
+      "epoch": 0.20284821118443905,
+      "grad_norm": 0.6314607624470274,
+      "learning_rate": 9.238318046356772e-06,
+      "loss": 0.0245,
+      "step": 584
+    },
+    {
+      "epoch": 0.20319555401180966,
+      "grad_norm": 1.0288562093872682,
+      "learning_rate": 9.235330546659731e-06,
+      "loss": 0.0232,
+      "step": 585
+    },
+    {
+      "epoch": 0.20354289683918028,
+      "grad_norm": 0.6216044232979469,
+      "learning_rate": 9.23233768460003e-06,
+      "loss": 0.0163,
+      "step": 586
+    },
+    {
+      "epoch": 0.2038902396665509,
+      "grad_norm": 0.7839133361622146,
+      "learning_rate": 9.229339463966942e-06,
+      "loss": 0.017,
+      "step": 587
+    },
+    {
+      "epoch": 0.2042375824939215,
+      "grad_norm": 1.065822018842104,
+      "learning_rate": 9.226335888556517e-06,
+      "loss": 0.0195,
+      "step": 588
+    },
+    {
+      "epoch": 0.2045849253212921,
+      "grad_norm": 1.0799416510268132,
+      "learning_rate": 9.223326962171594e-06,
+      "loss": 0.0329,
+      "step": 589
+    },
+    {
+      "epoch": 0.20493226814866272,
+      "grad_norm": 2.7643010329522597,
+      "learning_rate": 9.22031268862178e-06,
+      "loss": 0.0276,
+      "step": 590
+    },
+    {
+      "epoch": 0.20527961097603334,
+      "grad_norm": 2.7311705318356703,
+      "learning_rate": 9.217293071723455e-06,
+      "loss": 0.0267,
+      "step": 591
+    },
+    {
+      "epoch": 0.20562695380340396,
+      "grad_norm": 0.5493925419488683,
+      "learning_rate": 9.214268115299761e-06,
+      "loss": 0.0264,
+      "step": 592
+    },
+    {
+      "epoch": 0.20597429663077457,
+      "grad_norm": 1.2139756694249682,
+      "learning_rate": 9.211237823180605e-06,
+      "loss": 0.0216,
+      "step": 593
+    },
+    {
+      "epoch": 0.2063216394581452,
+      "grad_norm": 0.4237660059277888,
+      "learning_rate": 9.208202199202649e-06,
+      "loss": 0.0141,
+      "step": 594
+    },
+    {
+      "epoch": 0.2066689822855158,
+      "grad_norm": 0.8419781686821314,
+      "learning_rate": 9.205161247209303e-06,
+      "loss": 0.0166,
+      "step": 595
+    },
+    {
+      "epoch": 0.20701632511288642,
+      "grad_norm": 1.0055329342881736,
+      "learning_rate": 9.202114971050722e-06,
+      "loss": 0.0269,
+      "step": 596
+    },
+    {
+      "epoch": 0.20736366794025704,
+      "grad_norm": 0.58123379031455,
+      "learning_rate": 9.199063374583807e-06,
+      "loss": 0.0272,
+      "step": 597
+    },
+    {
+      "epoch": 0.20771101076762766,
+      "grad_norm": 1.258373507803076,
+      "learning_rate": 9.19600646167219e-06,
+      "loss": 0.0243,
+      "step": 598
+    },
+    {
+      "epoch": 0.20805835359499827,
+      "grad_norm": 0.8337160786397171,
+      "learning_rate": 9.192944236186237e-06,
+      "loss": 0.0175,
+      "step": 599
+    },
+    {
+      "epoch": 0.2084056964223689,
+      "grad_norm": 0.5397096011880417,
+      "learning_rate": 9.189876702003037e-06,
+      "loss": 0.0268,
+      "step": 600
+    },
+    {
+      "epoch": 0.20875303924973948,
+      "grad_norm": 2.0609037827991252,
+      "learning_rate": 9.186803863006408e-06,
+      "loss": 0.0288,
+      "step": 601
+    },
+    {
+      "epoch": 0.2091003820771101,
+      "grad_norm": 1.1561596599740218,
+      "learning_rate": 9.183725723086873e-06,
+      "loss": 0.0149,
+      "step": 602
+    },
+    {
+      "epoch": 0.20944772490448071,
+      "grad_norm": 1.7470410681014348,
+      "learning_rate": 9.180642286141678e-06,
+      "loss": 0.0284,
+      "step": 603
+    },
+    {
+      "epoch": 0.20979506773185133,
+      "grad_norm": 0.44157047622672,
+      "learning_rate": 9.177553556074766e-06,
+      "loss": 0.0204,
+      "step": 604
+    },
+    {
+      "epoch": 0.21014241055922195,
+      "grad_norm": 0.44448775934421736,
+      "learning_rate": 9.17445953679679e-06,
+      "loss": 0.0188,
+      "step": 605
+    },
+    {
+      "epoch": 0.21048975338659257,
+      "grad_norm": 1.4869913866249045,
+      "learning_rate": 9.171360232225091e-06,
+      "loss": 0.0216,
+      "step": 606
+    },
+    {
+      "epoch": 0.21083709621396318,
+      "grad_norm": 0.7283847092929767,
+      "learning_rate": 9.16825564628371e-06,
+      "loss": 0.0226,
+      "step": 607
+    },
+    {
+      "epoch": 0.2111844390413338,
+      "grad_norm": 0.44538635448331376,
+      "learning_rate": 9.165145782903369e-06,
+      "loss": 0.0112,
+      "step": 608
+    },
+    {
+      "epoch": 0.21153178186870442,
+      "grad_norm": 0.5848098598044253,
+      "learning_rate": 9.162030646021477e-06,
+      "loss": 0.0272,
+      "step": 609
+    },
+    {
+      "epoch": 0.21187912469607503,
+      "grad_norm": 0.6127021429282402,
+      "learning_rate": 9.15891023958211e-06,
+      "loss": 0.0183,
+      "step": 610
+    },
+    {
+      "epoch": 0.21222646752344565,
+      "grad_norm": 0.8170743093460934,
+      "learning_rate": 9.15578456753603e-06,
+      "loss": 0.0283,
+      "step": 611
+    },
+    {
+      "epoch": 0.21257381035081627,
+      "grad_norm": 1.1249613859894405,
+      "learning_rate": 9.152653633840654e-06,
+      "loss": 0.026,
+      "step": 612
+    },
+    {
+      "epoch": 0.21292115317818688,
+      "grad_norm": 0.985266358777264,
+      "learning_rate": 9.149517442460065e-06,
+      "loss": 0.0242,
+      "step": 613
+    },
+    {
+      "epoch": 0.21326849600555747,
+      "grad_norm": 0.5401136574602342,
+      "learning_rate": 9.146375997365006e-06,
+      "loss": 0.0236,
+      "step": 614
+    },
+    {
+      "epoch": 0.2136158388329281,
+      "grad_norm": 0.8086803852304764,
+      "learning_rate": 9.143229302532866e-06,
+      "loss": 0.0258,
+      "step": 615
+    },
+    {
+      "epoch": 0.2139631816602987,
+      "grad_norm": 0.8975085166076582,
+      "learning_rate": 9.140077361947681e-06,
+      "loss": 0.0194,
+      "step": 616
+    },
+    {
+      "epoch": 0.21431052448766932,
+      "grad_norm": 1.1788638290843472,
+      "learning_rate": 9.136920179600137e-06,
+      "loss": 0.0204,
+      "step": 617
+    },
+    {
+      "epoch": 0.21465786731503994,
+      "grad_norm": 0.5154845254827837,
+      "learning_rate": 9.133757759487545e-06,
+      "loss": 0.021,
+      "step": 618
+    },
+    {
+      "epoch": 0.21500521014241056,
+      "grad_norm": 0.8015539062724568,
+      "learning_rate": 9.130590105613854e-06,
+      "loss": 0.0227,
+      "step": 619
+    },
+    {
+      "epoch": 0.21535255296978117,
+      "grad_norm": 1.5385432558467305,
+      "learning_rate": 9.127417221989643e-06,
+      "loss": 0.0423,
+      "step": 620
+    },
+    {
+      "epoch": 0.2156998957971518,
+      "grad_norm": 0.9984997426393253,
+      "learning_rate": 9.1242391126321e-06,
+      "loss": 0.0217,
+      "step": 621
+    },
+    {
+      "epoch": 0.2160472386245224,
+      "grad_norm": 0.28392172266267085,
+      "learning_rate": 9.121055781565044e-06,
+      "loss": 0.0238,
+      "step": 622
+    },
+    {
+      "epoch": 0.21639458145189303,
+      "grad_norm": 1.264151293918839,
+      "learning_rate": 9.117867232818897e-06,
+      "loss": 0.0209,
+      "step": 623
+    },
+    {
+      "epoch": 0.21674192427926364,
+      "grad_norm": 1.0368184665287132,
+      "learning_rate": 9.114673470430688e-06,
+      "loss": 0.021,
+      "step": 624
+    },
+    {
+      "epoch": 0.21708926710663426,
+      "grad_norm": 0.6918438110380035,
+      "learning_rate": 9.111474498444046e-06,
+      "loss": 0.0263,
+      "step": 625
+    },
+    {
+      "epoch": 0.21743660993400488,
+      "grad_norm": 2.1367175315409477,
+      "learning_rate": 9.1082703209092e-06,
+      "loss": 0.0295,
+      "step": 626
+    },
+    {
+      "epoch": 0.21778395276137547,
+      "grad_norm": 0.7119870045258382,
+      "learning_rate": 9.105060941882966e-06,
+      "loss": 0.023,
+      "step": 627
+    },
+    {
+      "epoch": 0.21813129558874608,
+      "grad_norm": 1.14033999216725,
+      "learning_rate": 9.101846365428747e-06,
+      "loss": 0.0164,
+      "step": 628
+    },
+    {
+      "epoch": 0.2184786384161167,
+      "grad_norm": 0.4345990776226327,
+      "learning_rate": 9.098626595616527e-06,
+      "loss": 0.0226,
+      "step": 629
+    },
+    {
+      "epoch": 0.21882598124348732,
+      "grad_norm": 1.4156487567989067,
+      "learning_rate": 9.095401636522863e-06,
+      "loss": 0.0254,
+      "step": 630
+    },
+    {
+      "epoch": 0.21917332407085793,
+      "grad_norm": 1.7788155380270627,
+      "learning_rate": 9.092171492230883e-06,
+      "loss": 0.0287,
+      "step": 631
+    },
+    {
+      "epoch": 0.21952066689822855,
+      "grad_norm": 0.9127292614192289,
+      "learning_rate": 9.088936166830285e-06,
+      "loss": 0.0266,
+      "step": 632
+    },
+    {
+      "epoch": 0.21986800972559917,
+      "grad_norm": 0.847423011464306,
+      "learning_rate": 9.08569566441732e-06,
+      "loss": 0.0232,
+      "step": 633
+    },
+    {
+      "epoch": 0.22021535255296978,
+      "grad_norm": 0.43930165735953547,
+      "learning_rate": 9.082449989094798e-06,
+      "loss": 0.0183,
+      "step": 634
+    },
+    {
+      "epoch": 0.2205626953803404,
+      "grad_norm": 0.7809095217627077,
+      "learning_rate": 9.079199144972072e-06,
+      "loss": 0.014,
+      "step": 635
+    },
+    {
+      "epoch": 0.22091003820771102,
+      "grad_norm": 1.0253606245776024,
+      "learning_rate": 9.075943136165049e-06,
+      "loss": 0.0156,
+      "step": 636
+    },
+    {
+      "epoch": 0.22125738103508164,
+      "grad_norm": 1.5700732329185856,
+      "learning_rate": 9.072681966796169e-06,
+      "loss": 0.0255,
+      "step": 637
+    },
+    {
+      "epoch": 0.22160472386245225,
+      "grad_norm": 0.49771019117253024,
+      "learning_rate": 9.069415640994403e-06,
+      "loss": 0.0243,
+      "step": 638
+    },
+    {
+      "epoch": 0.22195206668982287,
+      "grad_norm": 0.46212016511074455,
+      "learning_rate": 9.066144162895259e-06,
+      "loss": 0.023,
+      "step": 639
+    },
+    {
+      "epoch": 0.22229940951719346,
+      "grad_norm": 0.7782758261073233,
+      "learning_rate": 9.062867536640762e-06,
+      "loss": 0.017,
+      "step": 640
+    },
+    {
+      "epoch": 0.22264675234456408,
+      "grad_norm": 0.7838906985841836,
+      "learning_rate": 9.059585766379455e-06,
+      "loss": 0.0165,
+      "step": 641
+    },
+    {
+      "epoch": 0.2229940951719347,
+      "grad_norm": 1.011196077425446,
+      "learning_rate": 9.056298856266399e-06,
+      "loss": 0.0218,
+      "step": 642
+    },
+    {
+      "epoch": 0.2233414379993053,
+      "grad_norm": 1.518375606641246,
+      "learning_rate": 9.053006810463156e-06,
+      "loss": 0.0254,
+      "step": 643
+    },
+    {
+      "epoch": 0.22368878082667593,
+      "grad_norm": 0.5838655335928358,
+      "learning_rate": 9.049709633137796e-06,
+      "loss": 0.0255,
+      "step": 644
+    },
+    {
+      "epoch": 0.22403612365404654,
+      "grad_norm": 0.7210470798591194,
+      "learning_rate": 9.04640732846488e-06,
+      "loss": 0.0238,
+      "step": 645
+    },
+    {
+      "epoch": 0.22438346648141716,
+      "grad_norm": 0.6437322596951384,
+      "learning_rate": 9.043099900625468e-06,
+      "loss": 0.0237,
+      "step": 646
+    },
+    {
+      "epoch": 0.22473080930878778,
+      "grad_norm": 1.0308720116120658,
+      "learning_rate": 9.039787353807101e-06,
+      "loss": 0.0305,
+      "step": 647
+    },
+    {
+      "epoch": 0.2250781521361584,
+      "grad_norm": 1.6497385641943867,
+      "learning_rate": 9.036469692203804e-06,
+      "loss": 0.0296,
+      "step": 648
+    },
+    {
+      "epoch": 0.225425494963529,
+      "grad_norm": 0.3313837148743059,
+      "learning_rate": 9.033146920016073e-06,
+      "loss": 0.0224,
+      "step": 649
+    },
+    {
+      "epoch": 0.22577283779089963,
+      "grad_norm": 0.5293106608754529,
+      "learning_rate": 9.029819041450884e-06,
+      "loss": 0.0203,
+      "step": 650
+    },
+    {
+      "epoch": 0.22612018061827024,
+      "grad_norm": 1.1616316561406497,
+      "learning_rate": 9.026486060721668e-06,
+      "loss": 0.0288,
+      "step": 651
+    },
+    {
+      "epoch": 0.22646752344564083,
+      "grad_norm": 1.3686370104380872,
+      "learning_rate": 9.023147982048322e-06,
+      "loss": 0.0247,
+      "step": 652
+    },
+    {
+      "epoch": 0.22681486627301145,
+      "grad_norm": 0.9530520238325823,
+      "learning_rate": 9.019804809657195e-06,
+      "loss": 0.0222,
+      "step": 653
+    },
+    {
+      "epoch": 0.22716220910038207,
+      "grad_norm": 0.4302751935992573,
+      "learning_rate": 9.016456547781088e-06,
+      "loss": 0.0222,
+      "step": 654
+    },
+    {
+      "epoch": 0.22750955192775268,
+      "grad_norm": 0.4268832341371163,
+      "learning_rate": 9.01310320065924e-06,
+      "loss": 0.0189,
+      "step": 655
+    },
+    {
+      "epoch": 0.2278568947551233,
+      "grad_norm": 0.3637254931806751,
+      "learning_rate": 9.009744772537336e-06,
+      "loss": 0.0208,
+      "step": 656
+    },
+    {
+      "epoch": 0.22820423758249392,
+      "grad_norm": 0.9443986440525276,
+      "learning_rate": 9.006381267667489e-06,
+      "loss": 0.0157,
+      "step": 657
+    },
+    {
+      "epoch": 0.22855158040986454,
+      "grad_norm": 1.033293442849867,
+      "learning_rate": 9.00301269030824e-06,
+      "loss": 0.0235,
+      "step": 658
+    },
+    {
+      "epoch": 0.22889892323723515,
+      "grad_norm": 0.7206031262994533,
+      "learning_rate": 8.999639044724555e-06,
+      "loss": 0.0144,
+      "step": 659
+    },
+    {
+      "epoch": 0.22924626606460577,
+      "grad_norm": 0.5569325989347864,
+      "learning_rate": 8.996260335187813e-06,
+      "loss": 0.0273,
+      "step": 660
+    },
+    {
+      "epoch": 0.2295936088919764,
+      "grad_norm": 0.6950933663288503,
+      "learning_rate": 8.992876565975809e-06,
+      "loss": 0.0351,
+      "step": 661
+    },
+    {
+      "epoch": 0.229940951719347,
+      "grad_norm": 0.961443300736354,
+      "learning_rate": 8.98948774137274e-06,
+      "loss": 0.0249,
+      "step": 662
+    },
+    {
+      "epoch": 0.23028829454671762,
+      "grad_norm": 1.9014588737941847,
+      "learning_rate": 8.986093865669205e-06,
+      "loss": 0.0313,
+      "step": 663
+    },
+    {
+      "epoch": 0.23063563737408824,
+      "grad_norm": 0.9492079615362812,
+      "learning_rate": 8.9826949431622e-06,
+      "loss": 0.0248,
+      "step": 664
+    },
+    {
+      "epoch": 0.23098298020145883,
+      "grad_norm": 0.5262876532553753,
+      "learning_rate": 8.97929097815511e-06,
+      "loss": 0.0271,
+      "step": 665
+    },
+    {
+      "epoch": 0.23133032302882944,
+      "grad_norm": 1.4774684801073679,
+      "learning_rate": 8.9758819749577e-06,
+      "loss": 0.0259,
+      "step": 666
+    },
+    {
+      "epoch": 0.23167766585620006,
+      "grad_norm": 0.47912155013309676,
+      "learning_rate": 8.972467937886122e-06,
+      "loss": 0.0198,
+      "step": 667
+    },
+    {
+      "epoch": 0.23202500868357068,
+      "grad_norm": 0.4857405075161966,
+      "learning_rate": 8.969048871262895e-06,
+      "loss": 0.021,
+      "step": 668
+    },
+    {
+      "epoch": 0.2323723515109413,
+      "grad_norm": 0.5501028852401879,
+      "learning_rate": 8.965624779416907e-06,
+      "loss": 0.0229,
+      "step": 669
+    },
+    {
+      "epoch": 0.2327196943383119,
+      "grad_norm": 0.37333376724281975,
+      "learning_rate": 8.96219566668341e-06,
+      "loss": 0.0216,
+      "step": 670
+    },
+    {
+      "epoch": 0.23306703716568253,
+      "grad_norm": 0.48066801556311134,
+      "learning_rate": 8.958761537404012e-06,
+      "loss": 0.0198,
+      "step": 671
+    },
+    {
+      "epoch": 0.23341437999305314,
+      "grad_norm": 0.813514292061828,
+      "learning_rate": 8.955322395926673e-06,
+      "loss": 0.0139,
+      "step": 672
+    },
+    {
+      "epoch": 0.23376172282042376,
+      "grad_norm": 0.7125740685535722,
+      "learning_rate": 8.9518782466057e-06,
+      "loss": 0.0158,
+      "step": 673
+    },
+    {
+      "epoch": 0.23410906564779438,
+      "grad_norm": 0.3825152111739057,
+      "learning_rate": 8.948429093801738e-06,
+      "loss": 0.0215,
+      "step": 674
+    },
+    {
+      "epoch": 0.234456408475165,
+      "grad_norm": 1.3795608714085683,
+      "learning_rate": 8.944974941881766e-06,
+      "loss": 0.0274,
+      "step": 675
+    },
+    {
+      "epoch": 0.2348037513025356,
+      "grad_norm": 0.5497870254567753,
+      "learning_rate": 8.941515795219098e-06,
+      "loss": 0.0197,
+      "step": 676
+    },
+    {
+      "epoch": 0.23515109412990623,
+      "grad_norm": 0.8354225349175881,
+      "learning_rate": 8.938051658193365e-06,
+      "loss": 0.0227,
+      "step": 677
+    },
+    {
+      "epoch": 0.23549843695727682,
+      "grad_norm": 0.5221880740171788,
+      "learning_rate": 8.934582535190522e-06,
+      "loss": 0.0212,
+      "step": 678
+    },
+    {
+      "epoch": 0.23584577978464744,
+      "grad_norm": 0.5930924226753035,
+      "learning_rate": 8.931108430602834e-06,
+      "loss": 0.0176,
+      "step": 679
+    },
+    {
+      "epoch": 0.23619312261201805,
+      "grad_norm": 1.4128463606001023,
+      "learning_rate": 8.927629348828874e-06,
+      "loss": 0.0171,
+      "step": 680
+    },
+    {
+      "epoch": 0.23654046543938867,
+      "grad_norm": 0.8753822994129902,
+      "learning_rate": 8.924145294273515e-06,
+      "loss": 0.0211,
+      "step": 681
+    },
+    {
+      "epoch": 0.2368878082667593,
+      "grad_norm": 0.7822190197357625,
+      "learning_rate": 8.920656271347925e-06,
+      "loss": 0.0251,
+      "step": 682
+    },
+    {
+      "epoch": 0.2372351510941299,
+      "grad_norm": 1.4919418777559417,
+      "learning_rate": 8.917162284469569e-06,
+      "loss": 0.0224,
+      "step": 683
+    },
+    {
+      "epoch": 0.23758249392150052,
+      "grad_norm": 1.5473820670531835,
+      "learning_rate": 8.91366333806219e-06,
+      "loss": 0.0277,
+      "step": 684
+    },
+    {
+      "epoch": 0.23792983674887114,
+      "grad_norm": 1.1016478393828795,
+      "learning_rate": 8.910159436555813e-06,
+      "loss": 0.0252,
+      "step": 685
+    },
+    {
+      "epoch": 0.23827717957624175,
+      "grad_norm": 1.5287564476717936,
+      "learning_rate": 8.90665058438674e-06,
+      "loss": 0.0188,
+      "step": 686
+    },
+    {
+      "epoch": 0.23862452240361237,
+      "grad_norm": 0.7366291157632144,
+      "learning_rate": 8.903136785997533e-06,
+      "loss": 0.0295,
+      "step": 687
+    },
+    {
+      "epoch": 0.238971865230983,
+      "grad_norm": 0.7896406244087025,
+      "learning_rate": 8.899618045837025e-06,
+      "loss": 0.0237,
+      "step": 688
+    },
+    {
+      "epoch": 0.2393192080583536,
+      "grad_norm": 1.1875606974598063,
+      "learning_rate": 8.896094368360297e-06,
+      "loss": 0.0162,
+      "step": 689
+    },
+    {
+      "epoch": 0.23966655088572422,
+      "grad_norm": 0.47042040685656056,
+      "learning_rate": 8.892565758028688e-06,
+      "loss": 0.0282,
+      "step": 690
+    },
+    {
+      "epoch": 0.2400138937130948,
+      "grad_norm": 0.5817301391235052,
+      "learning_rate": 8.889032219309781e-06,
+      "loss": 0.0291,
+      "step": 691
+    },
+    {
+      "epoch": 0.24036123654046543,
+      "grad_norm": 0.6761086942386368,
+      "learning_rate": 8.885493756677399e-06,
+      "loss": 0.0248,
+      "step": 692
+    },
+    {
+      "epoch": 0.24070857936783605,
+      "grad_norm": 0.8209577079998156,
+      "learning_rate": 8.881950374611597e-06,
+      "loss": 0.0196,
+      "step": 693
+    },
+    {
+      "epoch": 0.24105592219520666,
+      "grad_norm": 1.314377357321471,
+      "learning_rate": 8.878402077598662e-06,
+      "loss": 0.023,
+      "step": 694
+    },
+    {
+      "epoch": 0.24140326502257728,
+      "grad_norm": 1.1012244390968478,
+      "learning_rate": 8.874848870131098e-06,
+      "loss": 0.0255,
+      "step": 695
+    },
+    {
+      "epoch": 0.2417506078499479,
+      "grad_norm": 0.7845674765254901,
+      "learning_rate": 8.871290756707634e-06,
+      "loss": 0.0221,
+      "step": 696
+    },
+    {
+      "epoch": 0.2420979506773185,
+      "grad_norm": 0.59132264309778,
+      "learning_rate": 8.867727741833204e-06,
+      "loss": 0.0179,
+      "step": 697
+    },
+    {
+      "epoch": 0.24244529350468913,
+      "grad_norm": 1.2702082566630686,
+      "learning_rate": 8.86415983001895e-06,
+      "loss": 0.0284,
+      "step": 698
+    },
+    {
+      "epoch": 0.24279263633205975,
+      "grad_norm": 0.7795467424110684,
+      "learning_rate": 8.860587025782215e-06,
+      "loss": 0.0171,
+      "step": 699
+    },
+    {
+      "epoch": 0.24313997915943036,
+      "grad_norm": 1.84470369765152,
+      "learning_rate": 8.857009333646535e-06,
+      "loss": 0.0214,
+      "step": 700
+    },
+    {
+      "epoch": 0.24348732198680098,
+      "grad_norm": 1.9319346606245589,
+      "learning_rate": 8.853426758141635e-06,
+      "loss": 0.0242,
+      "step": 701
+    },
+    {
+      "epoch": 0.2438346648141716,
+      "grad_norm": 1.3940656814702743,
+      "learning_rate": 8.849839303803425e-06,
+      "loss": 0.0239,
+      "step": 702
+    },
+    {
+      "epoch": 0.24418200764154221,
+      "grad_norm": 0.6685433203227141,
+      "learning_rate": 8.846246975173985e-06,
+      "loss": 0.0224,
+      "step": 703
+    },
+    {
+      "epoch": 0.2445293504689128,
+      "grad_norm": 0.8276105742660492,
+      "learning_rate": 8.842649776801576e-06,
+      "loss": 0.031,
+      "step": 704
+    },
+    {
+      "epoch": 0.24487669329628342,
+      "grad_norm": 0.9152658175517937,
+      "learning_rate": 8.839047713240619e-06,
+      "loss": 0.0157,
+      "step": 705
+    },
+    {
+      "epoch": 0.24522403612365404,
+      "grad_norm": 1.0704740818766338,
+      "learning_rate": 8.835440789051692e-06,
+      "loss": 0.0143,
+      "step": 706
+    },
+    {
+      "epoch": 0.24557137895102465,
+      "grad_norm": 1.0779902900776792,
+      "learning_rate": 8.831829008801536e-06,
+      "loss": 0.0175,
+      "step": 707
+    },
+    {
+      "epoch": 0.24591872177839527,
+      "grad_norm": 0.6030988446891595,
+      "learning_rate": 8.828212377063033e-06,
+      "loss": 0.0194,
+      "step": 708
+    },
+    {
+      "epoch": 0.2462660646057659,
+      "grad_norm": 0.8610359328267158,
+      "learning_rate": 8.824590898415209e-06,
+      "loss": 0.0258,
+      "step": 709
+    },
+    {
+      "epoch": 0.2466134074331365,
+      "grad_norm": 0.5357367765825686,
+      "learning_rate": 8.820964577443227e-06,
+      "loss": 0.0278,
+      "step": 710
+    },
+    {
+      "epoch": 0.24696075026050712,
+      "grad_norm": 1.5065093568870387,
+      "learning_rate": 8.817333418738382e-06,
+      "loss": 0.0195,
+      "step": 711
+    },
+    {
+      "epoch": 0.24730809308787774,
+      "grad_norm": 1.4013893590690267,
+      "learning_rate": 8.813697426898094e-06,
+      "loss": 0.0276,
+      "step": 712
+    },
+    {
+      "epoch": 0.24765543591524836,
+      "grad_norm": 1.4557088461414078,
+      "learning_rate": 8.810056606525899e-06,
+      "loss": 0.0232,
+      "step": 713
+    },
+    {
+      "epoch": 0.24800277874261897,
+      "grad_norm": 1.1924673869323386,
+      "learning_rate": 8.80641096223145e-06,
+      "loss": 0.0163,
+      "step": 714
+    },
+    {
+      "epoch": 0.2483501215699896,
+      "grad_norm": 0.42056499190089386,
+      "learning_rate": 8.802760498630507e-06,
+      "loss": 0.0249,
+      "step": 715
+    },
+    {
+      "epoch": 0.2486974643973602,
+      "grad_norm": 0.3585210385651358,
+      "learning_rate": 8.79910522034493e-06,
+      "loss": 0.019,
+      "step": 716
+    },
+    {
+      "epoch": 0.2490448072247308,
+      "grad_norm": 0.46610691242048347,
+      "learning_rate": 8.795445132002679e-06,
+      "loss": 0.0154,
+      "step": 717
+    },
+    {
+      "epoch": 0.2493921500521014,
+      "grad_norm": 1.3759596235421945,
+      "learning_rate": 8.791780238237794e-06,
+      "loss": 0.0249,
+      "step": 718
+    },
+    {
+      "epoch": 0.24973949287947203,
+      "grad_norm": 10.041742445276327,
+      "learning_rate": 8.788110543690415e-06,
+      "loss": 0.0274,
+      "step": 719
+    },
+    {
+      "epoch": 0.25008683570684265,
+      "grad_norm": 0.9018639927916149,
+      "learning_rate": 8.784436053006746e-06,
+      "loss": 0.026,
+      "step": 720
+    },
+    {
+      "epoch": 0.2504341785342133,
+      "grad_norm": 0.6316727580755663,
+      "learning_rate": 8.780756770839071e-06,
+      "loss": 0.0117,
+      "step": 721
+    },
+    {
+      "epoch": 0.2507815213615839,
+      "grad_norm": 0.8380724086190908,
+      "learning_rate": 8.777072701845738e-06,
+      "loss": 0.0286,
+      "step": 722
+    },
+    {
+      "epoch": 0.25112886418895447,
+      "grad_norm": 0.7088681744129693,
+      "learning_rate": 8.773383850691155e-06,
+      "loss": 0.0264,
+      "step": 723
+    },
+    {
+      "epoch": 0.2514762070163251,
+      "grad_norm": 1.6040597806062724,
+      "learning_rate": 8.769690222045787e-06,
+      "loss": 0.0213,
+      "step": 724
+    },
+    {
+      "epoch": 0.2518235498436957,
+      "grad_norm": 0.59234889125779,
+      "learning_rate": 8.765991820586147e-06,
+      "loss": 0.0193,
+      "step": 725
+    },
+    {
+      "epoch": 0.25217089267106635,
+      "grad_norm": 1.3987574754257743,
+      "learning_rate": 8.762288650994786e-06,
+      "loss": 0.0302,
+      "step": 726
+    },
+    {
+      "epoch": 0.25251823549843694,
+      "grad_norm": 0.7554420842981747,
+      "learning_rate": 8.758580717960303e-06,
+      "loss": 0.02,
+      "step": 727
+    },
+    {
+      "epoch": 0.2528655783258076,
+      "grad_norm": 0.5299359962965386,
+      "learning_rate": 8.754868026177317e-06,
+      "loss": 0.0228,
+      "step": 728
+    },
+    {
+      "epoch": 0.25321292115317817,
+      "grad_norm": 0.5066652476216912,
+      "learning_rate": 8.751150580346477e-06,
+      "loss": 0.0152,
+      "step": 729
+    },
+    {
+      "epoch": 0.2535602639805488,
+      "grad_norm": 0.8528730067781101,
+      "learning_rate": 8.747428385174452e-06,
+      "loss": 0.0207,
+      "step": 730
+    },
+    {
+      "epoch": 0.2539076068079194,
+      "grad_norm": 0.6236339038844014,
+      "learning_rate": 8.743701445373922e-06,
+      "loss": 0.0217,
+      "step": 731
+    },
+    {
+      "epoch": 0.25425494963529005,
+      "grad_norm": 1.1276577230096638,
+      "learning_rate": 8.739969765663574e-06,
+      "loss": 0.018,
+      "step": 732
+    },
+    {
+      "epoch": 0.25460229246266064,
+      "grad_norm": 1.0286105613817527,
+      "learning_rate": 8.736233350768097e-06,
+      "loss": 0.0166,
+      "step": 733
+    },
+    {
+      "epoch": 0.2549496352900313,
+      "grad_norm": 1.4581045926440899,
+      "learning_rate": 8.732492205418176e-06,
+      "loss": 0.031,
+      "step": 734
+    },
+    {
+      "epoch": 0.2552969781174019,
+      "grad_norm": 1.8876644886015665,
+      "learning_rate": 8.728746334350483e-06,
+      "loss": 0.027,
+      "step": 735
+    },
+    {
+      "epoch": 0.25564432094477246,
+      "grad_norm": 1.0450409512942507,
+      "learning_rate": 8.72499574230768e-06,
+      "loss": 0.0207,
+      "step": 736
+    },
+    {
+      "epoch": 0.2559916637721431,
+      "grad_norm": 0.6743560512268864,
+      "learning_rate": 8.721240434038395e-06,
+      "loss": 0.0244,
+      "step": 737
+    },
+    {
+      "epoch": 0.2563390065995137,
+      "grad_norm": 0.8447479142593041,
+      "learning_rate": 8.717480414297236e-06,
+      "loss": 0.0288,
+      "step": 738
+    },
+    {
+      "epoch": 0.25668634942688434,
+      "grad_norm": 0.9145315370795962,
+      "learning_rate": 8.713715687844772e-06,
+      "loss": 0.0185,
+      "step": 739
+    },
+    {
+      "epoch": 0.25703369225425493,
+      "grad_norm": 1.5301247728960465,
+      "learning_rate": 8.709946259447535e-06,
+      "loss": 0.0275,
+      "step": 740
+    },
+    {
+      "epoch": 0.2573810350816256,
+      "grad_norm": 1.901274677450291,
+      "learning_rate": 8.706172133878006e-06,
+      "loss": 0.015,
+      "step": 741
+    },
+    {
+      "epoch": 0.25772837790899616,
+      "grad_norm": 0.49087360665043905,
+      "learning_rate": 8.702393315914615e-06,
+      "loss": 0.022,
+      "step": 742
+    },
+    {
+      "epoch": 0.2580757207363668,
+      "grad_norm": 1.0922750212562884,
+      "learning_rate": 8.698609810341733e-06,
+      "loss": 0.0218,
+      "step": 743
+    },
+    {
+      "epoch": 0.2584230635637374,
+      "grad_norm": 0.7102928208875048,
+      "learning_rate": 8.694821621949667e-06,
+      "loss": 0.0322,
+      "step": 744
+    },
+    {
+      "epoch": 0.25877040639110804,
+      "grad_norm": 0.886964613687143,
+      "learning_rate": 8.69102875553465e-06,
+      "loss": 0.0159,
+      "step": 745
+    },
+    {
+      "epoch": 0.25911774921847863,
+      "grad_norm": 0.7318076441241718,
+      "learning_rate": 8.68723121589884e-06,
+      "loss": 0.0233,
+      "step": 746
+    },
+    {
+      "epoch": 0.2594650920458493,
+      "grad_norm": 0.6058986578436707,
+      "learning_rate": 8.683429007850313e-06,
+      "loss": 0.0251,
+      "step": 747
+    },
+    {
+      "epoch": 0.25981243487321987,
+      "grad_norm": 0.9052930522075193,
+      "learning_rate": 8.679622136203055e-06,
+      "loss": 0.0149,
+      "step": 748
+    },
+    {
+      "epoch": 0.26015977770059046,
+      "grad_norm": 1.2032045295627478,
+      "learning_rate": 8.67581060577695e-06,
+      "loss": 0.0224,
+      "step": 749
+    },
+    {
+      "epoch": 0.2605071205279611,
+      "grad_norm": 0.539015089299102,
+      "learning_rate": 8.671994421397793e-06,
+      "loss": 0.0246,
+      "step": 750
+    },
+    {
+      "epoch": 0.2608544633553317,
+      "grad_norm": 1.5525307354093696,
+      "learning_rate": 8.668173587897261e-06,
+      "loss": 0.0191,
+      "step": 751
+    },
+    {
+      "epoch": 0.26120180618270233,
+      "grad_norm": 0.8769437334779882,
+      "learning_rate": 8.664348110112923e-06,
+      "loss": 0.0283,
+      "step": 752
+    },
+    {
+      "epoch": 0.2615491490100729,
+      "grad_norm": 0.8682417310780673,
+      "learning_rate": 8.660517992888225e-06,
+      "loss": 0.0158,
+      "step": 753
+    },
+    {
+      "epoch": 0.26189649183744357,
+      "grad_norm": 0.6663404002194353,
+      "learning_rate": 8.656683241072488e-06,
+      "loss": 0.0267,
+      "step": 754
+    },
+    {
+      "epoch": 0.26224383466481416,
+      "grad_norm": 0.5892947695489575,
+      "learning_rate": 8.6528438595209e-06,
+      "loss": 0.0278,
+      "step": 755
+    },
+    {
+      "epoch": 0.2625911774921848,
+      "grad_norm": 1.202175698552294,
+      "learning_rate": 8.648999853094514e-06,
+      "loss": 0.0266,
+      "step": 756
+    },
+    {
+      "epoch": 0.2629385203195554,
+      "grad_norm": 0.7025664951173237,
+      "learning_rate": 8.645151226660234e-06,
+      "loss": 0.0204,
+      "step": 757
+    },
+    {
+      "epoch": 0.26328586314692604,
+      "grad_norm": 0.4207213062287094,
+      "learning_rate": 8.641297985090815e-06,
+      "loss": 0.0231,
+      "step": 758
+    },
+    {
+      "epoch": 0.2636332059742966,
+      "grad_norm": 0.5565332142739162,
+      "learning_rate": 8.637440133264858e-06,
+      "loss": 0.0201,
+      "step": 759
+    },
+    {
+      "epoch": 0.26398054880166727,
+      "grad_norm": 0.98247646350947,
+      "learning_rate": 8.6335776760668e-06,
+      "loss": 0.03,
+      "step": 760
+    },
+    {
+      "epoch": 0.26432789162903786,
+      "grad_norm": 1.331753783533657,
+      "learning_rate": 8.629710618386903e-06,
+      "loss": 0.0237,
+      "step": 761
+    },
+    {
+      "epoch": 0.26467523445640845,
+      "grad_norm": 0.8360917594131502,
+      "learning_rate": 8.625838965121263e-06,
+      "loss": 0.0214,
+      "step": 762
+    },
+    {
+      "epoch": 0.2650225772837791,
+      "grad_norm": 0.6406382478000888,
+      "learning_rate": 8.621962721171789e-06,
+      "loss": 0.0252,
+      "step": 763
+    },
+    {
+      "epoch": 0.2653699201111497,
+      "grad_norm": 0.9906478193068576,
+      "learning_rate": 8.618081891446201e-06,
+      "loss": 0.0167,
+      "step": 764
+    },
+    {
+      "epoch": 0.2657172629385203,
+      "grad_norm": 0.5398729124754951,
+      "learning_rate": 8.61419648085803e-06,
+      "loss": 0.02,
+      "step": 765
+    },
+    {
+      "epoch": 0.2660646057658909,
+      "grad_norm": 1.1959531243334165,
+      "learning_rate": 8.610306494326601e-06,
+      "loss": 0.0253,
+      "step": 766
+    },
+    {
+      "epoch": 0.26641194859326156,
+      "grad_norm": 1.1692706174579726,
+      "learning_rate": 8.60641193677704e-06,
+      "loss": 0.0297,
+      "step": 767
+    },
+    {
+      "epoch": 0.26675929142063215,
+      "grad_norm": 1.0701239722359508,
+      "learning_rate": 8.602512813140251e-06,
+      "loss": 0.0214,
+      "step": 768
+    },
+    {
+      "epoch": 0.2671066342480028,
+      "grad_norm": 0.4828984533321399,
+      "learning_rate": 8.59860912835293e-06,
+      "loss": 0.0253,
+      "step": 769
+    },
+    {
+      "epoch": 0.2674539770753734,
+      "grad_norm": 1.1545942236415125,
+      "learning_rate": 8.594700887357537e-06,
+      "loss": 0.022,
+      "step": 770
+    },
+    {
+      "epoch": 0.26780131990274403,
+      "grad_norm": 1.5002177682572735,
+      "learning_rate": 8.59078809510231e-06,
+      "loss": 0.0247,
+      "step": 771
+    },
+    {
+      "epoch": 0.2681486627301146,
+      "grad_norm": 2.1953312403285024,
+      "learning_rate": 8.58687075654124e-06,
+      "loss": 0.0263,
+      "step": 772
+    },
+    {
+      "epoch": 0.26849600555748526,
+      "grad_norm": 1.2121243717081425,
+      "learning_rate": 8.582948876634084e-06,
+      "loss": 0.0167,
+      "step": 773
+    },
+    {
+      "epoch": 0.26884334838485585,
+      "grad_norm": 0.7901330703897526,
+      "learning_rate": 8.579022460346343e-06,
+      "loss": 0.0207,
+      "step": 774
+    },
+    {
+      "epoch": 0.26919069121222644,
+      "grad_norm": 0.44307666800843953,
+      "learning_rate": 8.57509151264926e-06,
+      "loss": 0.0142,
+      "step": 775
+    },
+    {
+      "epoch": 0.2695380340395971,
+      "grad_norm": 1.1332463643190542,
+      "learning_rate": 8.57115603851982e-06,
+      "loss": 0.0248,
+      "step": 776
+    },
+    {
+      "epoch": 0.2698853768669677,
+      "grad_norm": 1.2837269229641983,
+      "learning_rate": 8.567216042940735e-06,
+      "loss": 0.0159,
+      "step": 777
+    },
+    {
+      "epoch": 0.2702327196943383,
+      "grad_norm": 2.1182651419994714,
+      "learning_rate": 8.563271530900448e-06,
+      "loss": 0.0345,
+      "step": 778
+    },
+    {
+      "epoch": 0.2705800625217089,
+      "grad_norm": 1.0165246753047423,
+      "learning_rate": 8.55932250739311e-06,
+      "loss": 0.0254,
+      "step": 779
+    },
+    {
+      "epoch": 0.27092740534907955,
+      "grad_norm": 2.105285605392031,
+      "learning_rate": 8.555368977418593e-06,
+      "loss": 0.0306,
+      "step": 780
+    },
+    {
+      "epoch": 0.27127474817645014,
+      "grad_norm": 1.9706982563429236,
+      "learning_rate": 8.551410945982469e-06,
+      "loss": 0.0388,
+      "step": 781
+    },
+    {
+      "epoch": 0.2716220910038208,
+      "grad_norm": 1.2642900443762466,
+      "learning_rate": 8.547448418096012e-06,
+      "loss": 0.0169,
+      "step": 782
+    },
+    {
+      "epoch": 0.2719694338311914,
+      "grad_norm": 0.6266295783177325,
+      "learning_rate": 8.543481398776188e-06,
+      "loss": 0.0218,
+      "step": 783
+    },
+    {
+      "epoch": 0.272316776658562,
+      "grad_norm": 0.5973986334152628,
+      "learning_rate": 8.539509893045654e-06,
+      "loss": 0.0338,
+      "step": 784
+    },
+    {
+      "epoch": 0.2726641194859326,
+      "grad_norm": 2.4245373478500687,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.0231,
+      "step": 785
+    },
+    {
+      "epoch": 0.27301146231330325,
+      "grad_norm": 2.516583165841234,
+      "learning_rate": 8.531553442471453e-06,
+      "loss": 0.0262,
+      "step": 786
+    },
+    {
+      "epoch": 0.27335880514067384,
+      "grad_norm": 3.32894137710812,
+      "learning_rate": 8.527568507701467e-06,
+      "loss": 0.0319,
+      "step": 787
+    },
+    {
+      "epoch": 0.27370614796804443,
+      "grad_norm": 2.3072141104017625,
+      "learning_rate": 8.523579106668121e-06,
+      "loss": 0.0308,
+      "step": 788
+    },
+    {
+      "epoch": 0.2740534907954151,
+      "grad_norm": 1.1713281727561171,
+      "learning_rate": 8.519585244422405e-06,
+      "loss": 0.0184,
+      "step": 789
+    },
+    {
+      "epoch": 0.27440083362278567,
+      "grad_norm": 1.0031858926420991,
+      "learning_rate": 8.515586926020959e-06,
+      "loss": 0.0235,
+      "step": 790
+    },
+    {
+      "epoch": 0.2747481764501563,
+      "grad_norm": 0.36533679027936755,
+      "learning_rate": 8.511584156526059e-06,
+      "loss": 0.0183,
+      "step": 791
+    },
+    {
+      "epoch": 0.2750955192775269,
+      "grad_norm": 1.1097620564934818,
+      "learning_rate": 8.507576941005626e-06,
+      "loss": 0.0265,
+      "step": 792
+    },
+    {
+      "epoch": 0.27544286210489755,
+      "grad_norm": 0.8446308489408966,
+      "learning_rate": 8.503565284533206e-06,
+      "loss": 0.0201,
+      "step": 793
+    },
+    {
+      "epoch": 0.27579020493226813,
+      "grad_norm": 1.5612638631332916,
+      "learning_rate": 8.499549192187965e-06,
+      "loss": 0.022,
+      "step": 794
+    },
+    {
+      "epoch": 0.2761375477596388,
+      "grad_norm": 1.0436423254930283,
+      "learning_rate": 8.495528669054688e-06,
+      "loss": 0.0188,
+      "step": 795
+    },
+    {
+      "epoch": 0.27648489058700937,
+      "grad_norm": 0.7033478660193114,
+      "learning_rate": 8.49150372022377e-06,
+      "loss": 0.016,
+      "step": 796
+    },
+    {
+      "epoch": 0.27683223341438,
+      "grad_norm": 0.7998512004622108,
+      "learning_rate": 8.48747435079121e-06,
+      "loss": 0.0297,
+      "step": 797
+    },
+    {
+      "epoch": 0.2771795762417506,
+      "grad_norm": 0.5199486841483459,
+      "learning_rate": 8.483440565858599e-06,
+      "loss": 0.0216,
+      "step": 798
+    },
+    {
+      "epoch": 0.27752691906912125,
+      "grad_norm": 1.2032922365716692,
+      "learning_rate": 8.479402370533127e-06,
+      "loss": 0.03,
+      "step": 799
+    },
+    {
+      "epoch": 0.27787426189649184,
+      "grad_norm": 1.9065992211352163,
+      "learning_rate": 8.47535976992756e-06,
+      "loss": 0.0242,
+      "step": 800
+    },
+    {
+      "epoch": 0.2782216047238624,
+      "grad_norm": 0.7910306233638731,
+      "learning_rate": 8.471312769160247e-06,
+      "loss": 0.014,
+      "step": 801
+    },
+    {
+      "epoch": 0.27856894755123307,
+      "grad_norm": 0.9658546663370163,
+      "learning_rate": 8.467261373355104e-06,
+      "loss": 0.0234,
+      "step": 802
+    },
+    {
+      "epoch": 0.27891629037860366,
+      "grad_norm": 0.6546478423971199,
+      "learning_rate": 8.463205587641614e-06,
+      "loss": 0.0189,
+      "step": 803
+    },
+    {
+      "epoch": 0.2792636332059743,
+      "grad_norm": 0.6930129832525015,
+      "learning_rate": 8.459145417154817e-06,
+      "loss": 0.0258,
+      "step": 804
+    },
+    {
+      "epoch": 0.2796109760333449,
+      "grad_norm": 0.4441678217661302,
+      "learning_rate": 8.455080867035307e-06,
+      "loss": 0.0125,
+      "step": 805
+    },
+    {
+      "epoch": 0.27995831886071554,
+      "grad_norm": 1.1591532386678776,
+      "learning_rate": 8.451011942429219e-06,
+      "loss": 0.0283,
+      "step": 806
+    },
+    {
+      "epoch": 0.2803056616880861,
+      "grad_norm": 1.2540462658684661,
+      "learning_rate": 8.44693864848823e-06,
+      "loss": 0.0144,
+      "step": 807
+    },
+    {
+      "epoch": 0.28065300451545677,
+      "grad_norm": 0.8407468838106213,
+      "learning_rate": 8.442860990369545e-06,
+      "loss": 0.0194,
+      "step": 808
+    },
+    {
+      "epoch": 0.28100034734282736,
+      "grad_norm": 1.02240345065518,
+      "learning_rate": 8.438778973235904e-06,
+      "loss": 0.0372,
+      "step": 809
+    },
+    {
+      "epoch": 0.281347690170198,
+      "grad_norm": 0.711512705654793,
+      "learning_rate": 8.43469260225555e-06,
+      "loss": 0.0136,
+      "step": 810
+    },
+    {
+      "epoch": 0.2816950329975686,
+      "grad_norm": 0.5328105749063056,
+      "learning_rate": 8.430601882602256e-06,
+      "loss": 0.0228,
+      "step": 811
+    },
+    {
+      "epoch": 0.28204237582493924,
+      "grad_norm": 0.4977620396488783,
+      "learning_rate": 8.426506819455285e-06,
+      "loss": 0.0173,
+      "step": 812
+    },
+    {
+      "epoch": 0.28238971865230983,
+      "grad_norm": 0.7517007103690861,
+      "learning_rate": 8.422407417999413e-06,
+      "loss": 0.0215,
+      "step": 813
+    },
+    {
+      "epoch": 0.2827370614796804,
+      "grad_norm": 1.988030057566032,
+      "learning_rate": 8.418303683424898e-06,
+      "loss": 0.0236,
+      "step": 814
+    },
+    {
+      "epoch": 0.28308440430705106,
+      "grad_norm": 0.564222331342316,
+      "learning_rate": 8.414195620927491e-06,
+      "loss": 0.0168,
+      "step": 815
+    },
+    {
+      "epoch": 0.28343174713442165,
+      "grad_norm": 0.6853557880647918,
+      "learning_rate": 8.410083235708422e-06,
+      "loss": 0.0154,
+      "step": 816
+    },
+    {
+      "epoch": 0.2837790899617923,
+      "grad_norm": 2.11338878622302,
+      "learning_rate": 8.405966532974388e-06,
+      "loss": 0.0264,
+      "step": 817
+    },
+    {
+      "epoch": 0.2841264327891629,
+      "grad_norm": 0.8210234110109191,
+      "learning_rate": 8.401845517937558e-06,
+      "loss": 0.0291,
+      "step": 818
+    },
+    {
+      "epoch": 0.28447377561653353,
+      "grad_norm": 0.889999073550466,
+      "learning_rate": 8.397720195815561e-06,
+      "loss": 0.0295,
+      "step": 819
+    },
+    {
+      "epoch": 0.2848211184439041,
+      "grad_norm": 0.631173907871462,
+      "learning_rate": 8.393590571831478e-06,
+      "loss": 0.0217,
+      "step": 820
+    },
+    {
+      "epoch": 0.28516846127127476,
+      "grad_norm": 0.7731774878641796,
+      "learning_rate": 8.389456651213834e-06,
+      "loss": 0.0155,
+      "step": 821
+    },
+    {
+      "epoch": 0.28551580409864535,
+      "grad_norm": 1.5606323479981803,
+      "learning_rate": 8.385318439196597e-06,
+      "loss": 0.0249,
+      "step": 822
+    },
+    {
+      "epoch": 0.285863146926016,
+      "grad_norm": 0.596618016585276,
+      "learning_rate": 8.381175941019171e-06,
+      "loss": 0.0325,
+      "step": 823
+    },
+    {
+      "epoch": 0.2862104897533866,
+      "grad_norm": 1.0647983506484207,
+      "learning_rate": 8.377029161926378e-06,
+      "loss": 0.0268,
+      "step": 824
+    },
+    {
+      "epoch": 0.28655783258075723,
+      "grad_norm": 0.5119216226794532,
+      "learning_rate": 8.372878107168469e-06,
+      "loss": 0.0238,
+      "step": 825
+    },
+    {
+      "epoch": 0.2869051754081278,
+      "grad_norm": 0.46410901637663077,
+      "learning_rate": 8.368722782001104e-06,
+      "loss": 0.0169,
+      "step": 826
+    },
+    {
+      "epoch": 0.2872525182354984,
+      "grad_norm": 1.3248050562513085,
+      "learning_rate": 8.364563191685348e-06,
+      "loss": 0.0224,
+      "step": 827
+    },
+    {
+      "epoch": 0.28759986106286906,
+      "grad_norm": 0.9361071547212366,
+      "learning_rate": 8.360399341487675e-06,
+      "loss": 0.0259,
+      "step": 828
+    },
+    {
+      "epoch": 0.28794720389023964,
+      "grad_norm": 1.1870448572485746,
+      "learning_rate": 8.35623123667994e-06,
+      "loss": 0.0193,
+      "step": 829
+    },
+    {
+      "epoch": 0.2882945467176103,
+      "grad_norm": 0.9586109702853123,
+      "learning_rate": 8.352058882539394e-06,
+      "loss": 0.0206,
+      "step": 830
+    },
+    {
+      "epoch": 0.2886418895449809,
+      "grad_norm": 0.5309149834858535,
+      "learning_rate": 8.347882284348665e-06,
+      "loss": 0.0208,
+      "step": 831
+    },
+    {
+      "epoch": 0.2889892323723515,
+      "grad_norm": 1.3733779916187276,
+      "learning_rate": 8.343701447395754e-06,
+      "loss": 0.0183,
+      "step": 832
+    },
+    {
+      "epoch": 0.2893365751997221,
+      "grad_norm": 1.3507083935119284,
+      "learning_rate": 8.339516376974028e-06,
+      "loss": 0.0228,
+      "step": 833
+    },
+    {
+      "epoch": 0.28968391802709276,
+      "grad_norm": 0.9249455391192148,
+      "learning_rate": 8.33532707838222e-06,
+      "loss": 0.02,
+      "step": 834
+    },
+    {
+      "epoch": 0.29003126085446335,
+      "grad_norm": 1.6816616489709881,
+      "learning_rate": 8.331133556924404e-06,
+      "loss": 0.0134,
+      "step": 835
+    },
+    {
+      "epoch": 0.290378603681834,
+      "grad_norm": 1.0063521448061856,
+      "learning_rate": 8.326935817910014e-06,
+      "loss": 0.0188,
+      "step": 836
+    },
+    {
+      "epoch": 0.2907259465092046,
+      "grad_norm": 0.8258758559049553,
+      "learning_rate": 8.322733866653814e-06,
+      "loss": 0.0263,
+      "step": 837
+    },
+    {
+      "epoch": 0.2910732893365752,
+      "grad_norm": 1.173146724262853,
+      "learning_rate": 8.31852770847591e-06,
+      "loss": 0.0249,
+      "step": 838
+    },
+    {
+      "epoch": 0.2914206321639458,
+      "grad_norm": 6.319556253427923,
+      "learning_rate": 8.314317348701724e-06,
+      "loss": 0.0334,
+      "step": 839
+    },
+    {
+      "epoch": 0.2917679749913164,
+      "grad_norm": 3.142508712422825,
+      "learning_rate": 8.310102792662006e-06,
+      "loss": 0.0402,
+      "step": 840
+    },
+    {
+      "epoch": 0.29211531781868705,
+      "grad_norm": 2.293229808909149,
+      "learning_rate": 8.305884045692815e-06,
+      "loss": 0.0401,
+      "step": 841
+    },
+    {
+      "epoch": 0.29246266064605764,
+      "grad_norm": 1.021991216996574,
+      "learning_rate": 8.30166111313552e-06,
+      "loss": 0.0162,
+      "step": 842
+    },
+    {
+      "epoch": 0.2928100034734283,
+      "grad_norm": 1.1326524259624917,
+      "learning_rate": 8.297434000336781e-06,
+      "loss": 0.0184,
+      "step": 843
+    },
+    {
+      "epoch": 0.29315734630079887,
+      "grad_norm": 0.6967485472695889,
+      "learning_rate": 8.29320271264856e-06,
+      "loss": 0.0228,
+      "step": 844
+    },
+    {
+      "epoch": 0.2935046891281695,
+      "grad_norm": 1.5690012404474116,
+      "learning_rate": 8.288967255428102e-06,
+      "loss": 0.0356,
+      "step": 845
+    },
+    {
+      "epoch": 0.2938520319555401,
+      "grad_norm": 1.6508824584996764,
+      "learning_rate": 8.284727634037928e-06,
+      "loss": 0.022,
+      "step": 846
+    },
+    {
+      "epoch": 0.29419937478291075,
+      "grad_norm": 1.7189879458010435,
+      "learning_rate": 8.280483853845831e-06,
+      "loss": 0.0176,
+      "step": 847
+    },
+    {
+      "epoch": 0.29454671761028134,
+      "grad_norm": 1.3338043589868147,
+      "learning_rate": 8.276235920224877e-06,
+      "loss": 0.0139,
+      "step": 848
+    },
+    {
+      "epoch": 0.294894060437652,
+      "grad_norm": 0.48731311842832525,
+      "learning_rate": 8.271983838553383e-06,
+      "loss": 0.0344,
+      "step": 849
+    },
+    {
+      "epoch": 0.2952414032650226,
+      "grad_norm": 0.6636965122047247,
+      "learning_rate": 8.26772761421492e-06,
+      "loss": 0.0213,
+      "step": 850
+    },
+    {
+      "epoch": 0.2955887460923932,
+      "grad_norm": 0.7449395913052544,
+      "learning_rate": 8.263467252598303e-06,
+      "loss": 0.0283,
+      "step": 851
+    },
+    {
+      "epoch": 0.2959360889197638,
+      "grad_norm": 0.5866063388294399,
+      "learning_rate": 8.25920275909759e-06,
+      "loss": 0.0237,
+      "step": 852
+    },
+    {
+      "epoch": 0.2962834317471344,
+      "grad_norm": 0.5762702547689298,
+      "learning_rate": 8.254934139112062e-06,
+      "loss": 0.0197,
+      "step": 853
+    },
+    {
+      "epoch": 0.29663077457450504,
+      "grad_norm": 0.4188655050848168,
+      "learning_rate": 8.250661398046236e-06,
+      "loss": 0.0201,
+      "step": 854
+    },
+    {
+      "epoch": 0.29697811740187563,
+      "grad_norm": 0.4948572771836202,
+      "learning_rate": 8.246384541309835e-06,
+      "loss": 0.0228,
+      "step": 855
+    },
+    {
+      "epoch": 0.2973254602292463,
+      "grad_norm": 0.8130265751045145,
+      "learning_rate": 8.242103574317802e-06,
+      "loss": 0.0192,
+      "step": 856
+    },
+    {
+      "epoch": 0.29767280305661686,
+      "grad_norm": 0.7538997021511098,
+      "learning_rate": 8.237818502490273e-06,
+      "loss": 0.0207,
+      "step": 857
+    },
+    {
+      "epoch": 0.2980201458839875,
+      "grad_norm": 0.7989130279170902,
+      "learning_rate": 8.233529331252598e-06,
+      "loss": 0.0203,
+      "step": 858
+    },
+    {
+      "epoch": 0.2983674887113581,
+      "grad_norm": 1.7902955280949013,
+      "learning_rate": 8.2292360660353e-06,
+      "loss": 0.0287,
+      "step": 859
+    },
+    {
+      "epoch": 0.29871483153872874,
+      "grad_norm": 0.6270804680604328,
+      "learning_rate": 8.224938712274097e-06,
+      "loss": 0.0241,
+      "step": 860
+    },
+    {
+      "epoch": 0.29906217436609933,
+      "grad_norm": 0.8032825382756686,
+      "learning_rate": 8.220637275409878e-06,
+      "loss": 0.0194,
+      "step": 861
+    },
+    {
+      "epoch": 0.29940951719347,
+      "grad_norm": 1.1586121906804554,
+      "learning_rate": 8.2163317608887e-06,
+      "loss": 0.0207,
+      "step": 862
+    },
+    {
+      "epoch": 0.29975686002084057,
+      "grad_norm": 0.9100243599349078,
+      "learning_rate": 8.21202217416179e-06,
+      "loss": 0.0156,
+      "step": 863
+    },
+    {
+      "epoch": 0.3001042028482112,
+      "grad_norm": 0.6532828480664288,
+      "learning_rate": 8.207708520685526e-06,
+      "loss": 0.0187,
+      "step": 864
+    },
+    {
+      "epoch": 0.3004515456755818,
+      "grad_norm": 0.5964555383406674,
+      "learning_rate": 8.203390805921437e-06,
+      "loss": 0.0176,
+      "step": 865
+    },
+    {
+      "epoch": 0.3007988885029524,
+      "grad_norm": 0.6421563749310066,
+      "learning_rate": 8.199069035336186e-06,
+      "loss": 0.017,
+      "step": 866
+    },
+    {
+      "epoch": 0.30114623133032303,
+      "grad_norm": 0.5207710470189775,
+      "learning_rate": 8.194743214401587e-06,
+      "loss": 0.0169,
+      "step": 867
+    },
+    {
+      "epoch": 0.3014935741576936,
+      "grad_norm": 0.6923218916041749,
+      "learning_rate": 8.190413348594564e-06,
+      "loss": 0.027,
+      "step": 868
+    },
+    {
+      "epoch": 0.30184091698506427,
+      "grad_norm": 0.5670711050522923,
+      "learning_rate": 8.186079443397174e-06,
+      "loss": 0.0278,
+      "step": 869
+    },
+    {
+      "epoch": 0.30218825981243486,
+      "grad_norm": 0.7702120413403443,
+      "learning_rate": 8.181741504296588e-06,
+      "loss": 0.0251,
+      "step": 870
+    },
+    {
+      "epoch": 0.3025356026398055,
+      "grad_norm": 0.6513641907856846,
+      "learning_rate": 8.17739953678508e-06,
+      "loss": 0.0193,
+      "step": 871
+    },
+    {
+      "epoch": 0.3028829454671761,
+      "grad_norm": 1.2819392100935951,
+      "learning_rate": 8.173053546360025e-06,
+      "loss": 0.0239,
+      "step": 872
+    },
+    {
+      "epoch": 0.30323028829454673,
+      "grad_norm": 0.8392899596493151,
+      "learning_rate": 8.168703538523892e-06,
+      "loss": 0.0281,
+      "step": 873
+    },
+    {
+      "epoch": 0.3035776311219173,
+      "grad_norm": 0.5848723579885045,
+      "learning_rate": 8.16434951878424e-06,
+      "loss": 0.0191,
+      "step": 874
+    },
+    {
+      "epoch": 0.30392497394928797,
+      "grad_norm": 1.1086246012363608,
+      "learning_rate": 8.1599914926537e-06,
+      "loss": 0.0167,
+      "step": 875
+    },
+    {
+      "epoch": 0.30427231677665856,
+      "grad_norm": 0.7106156523093721,
+      "learning_rate": 8.155629465649983e-06,
+      "loss": 0.0231,
+      "step": 876
+    },
+    {
+      "epoch": 0.3046196596040292,
+      "grad_norm": 0.770138606469383,
+      "learning_rate": 8.15126344329586e-06,
+      "loss": 0.0157,
+      "step": 877
+    },
+    {
+      "epoch": 0.3049670024313998,
+      "grad_norm": 0.47477270596349086,
+      "learning_rate": 8.146893431119166e-06,
+      "loss": 0.0281,
+      "step": 878
+    },
+    {
+      "epoch": 0.3053143452587704,
+      "grad_norm": 0.6077495292480687,
+      "learning_rate": 8.142519434652782e-06,
+      "loss": 0.0203,
+      "step": 879
+    },
+    {
+      "epoch": 0.305661688086141,
+      "grad_norm": 0.8375044162407927,
+      "learning_rate": 8.138141459434638e-06,
+      "loss": 0.0207,
+      "step": 880
+    },
+    {
+      "epoch": 0.3060090309135116,
+      "grad_norm": 0.7961579743921388,
+      "learning_rate": 8.133759511007697e-06,
+      "loss": 0.018,
+      "step": 881
+    },
+    {
+      "epoch": 0.30635637374088226,
+      "grad_norm": 0.9954610793959501,
+      "learning_rate": 8.129373594919957e-06,
+      "loss": 0.0177,
+      "step": 882
+    },
+    {
+      "epoch": 0.30670371656825285,
+      "grad_norm": 0.2668325158805243,
+      "learning_rate": 8.124983716724434e-06,
+      "loss": 0.0165,
+      "step": 883
+    },
+    {
+      "epoch": 0.3070510593956235,
+      "grad_norm": 0.7195212219421603,
+      "learning_rate": 8.120589881979167e-06,
+      "loss": 0.0174,
+      "step": 884
+    },
+    {
+      "epoch": 0.3073984022229941,
+      "grad_norm": 0.5395482390763875,
+      "learning_rate": 8.116192096247202e-06,
+      "loss": 0.0271,
+      "step": 885
+    },
+    {
+      "epoch": 0.3077457450503647,
+      "grad_norm": 1.297519296316108,
+      "learning_rate": 8.111790365096584e-06,
+      "loss": 0.0211,
+      "step": 886
+    },
+    {
+      "epoch": 0.3080930878777353,
+      "grad_norm": 1.110719079206034,
+      "learning_rate": 8.107384694100355e-06,
+      "loss": 0.0305,
+      "step": 887
+    },
+    {
+      "epoch": 0.30844043070510596,
+      "grad_norm": 0.4497669645124215,
+      "learning_rate": 8.102975088836551e-06,
+      "loss": 0.0152,
+      "step": 888
+    },
+    {
+      "epoch": 0.30878777353247655,
+      "grad_norm": 0.6803690138422418,
+      "learning_rate": 8.098561554888181e-06,
+      "loss": 0.036,
+      "step": 889
+    },
+    {
+      "epoch": 0.3091351163598472,
+      "grad_norm": 0.5046240454353891,
+      "learning_rate": 8.09414409784323e-06,
+      "loss": 0.019,
+      "step": 890
+    },
+    {
+      "epoch": 0.3094824591872178,
+      "grad_norm": 0.7226803158978641,
+      "learning_rate": 8.089722723294654e-06,
+      "loss": 0.0221,
+      "step": 891
+    },
+    {
+      "epoch": 0.3098298020145884,
+      "grad_norm": 1.0803635312237747,
+      "learning_rate": 8.085297436840365e-06,
+      "loss": 0.0224,
+      "step": 892
+    },
+    {
+      "epoch": 0.310177144841959,
+      "grad_norm": 0.41686248676010906,
+      "learning_rate": 8.080868244083232e-06,
+      "loss": 0.0257,
+      "step": 893
+    },
+    {
+      "epoch": 0.3105244876693296,
+      "grad_norm": 1.0314297720832435,
+      "learning_rate": 8.076435150631064e-06,
+      "loss": 0.0182,
+      "step": 894
+    },
+    {
+      "epoch": 0.31087183049670025,
+      "grad_norm": 2.59167850646609,
+      "learning_rate": 8.071998162096613e-06,
+      "loss": 0.0291,
+      "step": 895
+    },
+    {
+      "epoch": 0.31121917332407084,
+      "grad_norm": 1.7037505969911395,
+      "learning_rate": 8.06755728409756e-06,
+      "loss": 0.0255,
+      "step": 896
+    },
+    {
+      "epoch": 0.3115665161514415,
+      "grad_norm": 0.5736683932010955,
+      "learning_rate": 8.063112522256516e-06,
+      "loss": 0.0352,
+      "step": 897
+    },
+    {
+      "epoch": 0.3119138589788121,
+      "grad_norm": 0.6074990637984911,
+      "learning_rate": 8.058663882200998e-06,
+      "loss": 0.0221,
+      "step": 898
+    },
+    {
+      "epoch": 0.3122612018061827,
+      "grad_norm": 1.047458516096098,
+      "learning_rate": 8.054211369563448e-06,
+      "loss": 0.0239,
+      "step": 899
+    },
+    {
+      "epoch": 0.3126085446335533,
+      "grad_norm": 1.3336377703933007,
+      "learning_rate": 8.049754989981198e-06,
+      "loss": 0.0258,
+      "step": 900
+    },
+    {
+      "epoch": 0.31295588746092395,
+      "grad_norm": 0.5743775466749255,
+      "learning_rate": 8.045294749096485e-06,
+      "loss": 0.031,
+      "step": 901
+    },
+    {
+      "epoch": 0.31330323028829454,
+      "grad_norm": 0.6348423077667337,
+      "learning_rate": 8.040830652556429e-06,
+      "loss": 0.0225,
+      "step": 902
+    },
+    {
+      "epoch": 0.3136505731156652,
+      "grad_norm": 0.4690059985847731,
+      "learning_rate": 8.036362706013033e-06,
+      "loss": 0.0219,
+      "step": 903
+    },
+    {
+      "epoch": 0.3139979159430358,
+      "grad_norm": 0.6063025129771344,
+      "learning_rate": 8.031890915123178e-06,
+      "loss": 0.0306,
+      "step": 904
+    },
+    {
+      "epoch": 0.31434525877040637,
+      "grad_norm": 0.8197671720949243,
+      "learning_rate": 8.02741528554861e-06,
+      "loss": 0.023,
+      "step": 905
+    },
+    {
+      "epoch": 0.314692601597777,
+      "grad_norm": 0.6251152385832162,
+      "learning_rate": 8.02293582295593e-06,
+      "loss": 0.0215,
+      "step": 906
+    },
+    {
+      "epoch": 0.3150399444251476,
+      "grad_norm": 0.6072902912466627,
+      "learning_rate": 8.018452533016604e-06,
+      "loss": 0.0231,
+      "step": 907
+    },
+    {
+      "epoch": 0.31538728725251824,
+      "grad_norm": 0.31262167331309026,
+      "learning_rate": 8.01396542140693e-06,
+      "loss": 0.0179,
+      "step": 908
+    },
+    {
+      "epoch": 0.31573463007988883,
+      "grad_norm": 1.1201275890016398,
+      "learning_rate": 8.009474493808054e-06,
+      "loss": 0.0173,
+      "step": 909
+    },
+    {
+      "epoch": 0.3160819729072595,
+      "grad_norm": 0.614103933391006,
+      "learning_rate": 8.004979755905953e-06,
+      "loss": 0.0204,
+      "step": 910
+    },
+    {
+      "epoch": 0.31642931573463007,
+      "grad_norm": 1.2968687606304459,
+      "learning_rate": 8.000481213391422e-06,
+      "loss": 0.0304,
+      "step": 911
+    },
+    {
+      "epoch": 0.3167766585620007,
+      "grad_norm": 0.5937663477943244,
+      "learning_rate": 7.995978871960079e-06,
+      "loss": 0.0205,
+      "step": 912
+    },
+    {
+      "epoch": 0.3171240013893713,
+      "grad_norm": 0.5499626397370164,
+      "learning_rate": 7.991472737312351e-06,
+      "loss": 0.0185,
+      "step": 913
+    },
+    {
+      "epoch": 0.31747134421674195,
+      "grad_norm": 1.101916609969889,
+      "learning_rate": 7.986962815153466e-06,
+      "loss": 0.0301,
+      "step": 914
+    },
+    {
+      "epoch": 0.31781868704411254,
+      "grad_norm": 1.091262784145427,
+      "learning_rate": 7.982449111193445e-06,
+      "loss": 0.0208,
+      "step": 915
+    },
+    {
+      "epoch": 0.3181660298714832,
+      "grad_norm": 0.6932386123391473,
+      "learning_rate": 7.977931631147102e-06,
+      "loss": 0.019,
+      "step": 916
+    },
+    {
+      "epoch": 0.31851337269885377,
+      "grad_norm": 1.697367211349032,
+      "learning_rate": 7.97341038073403e-06,
+      "loss": 0.0225,
+      "step": 917
+    },
+    {
+      "epoch": 0.31886071552622436,
+      "grad_norm": 0.48030348066887196,
+      "learning_rate": 7.968885365678596e-06,
+      "loss": 0.0198,
+      "step": 918
+    },
+    {
+      "epoch": 0.319208058353595,
+      "grad_norm": 0.7911731814497744,
+      "learning_rate": 7.96435659170993e-06,
+      "loss": 0.0179,
+      "step": 919
+    },
+    {
+      "epoch": 0.3195554011809656,
+      "grad_norm": 0.507749020308773,
+      "learning_rate": 7.959824064561927e-06,
+      "loss": 0.01,
+      "step": 920
+    },
+    {
+      "epoch": 0.31990274400833624,
+      "grad_norm": 1.4662590811027465,
+      "learning_rate": 7.955287789973231e-06,
+      "loss": 0.0183,
+      "step": 921
+    },
+    {
+      "epoch": 0.3202500868357068,
+      "grad_norm": 1.2953903150942983,
+      "learning_rate": 7.950747773687231e-06,
+      "loss": 0.0191,
+      "step": 922
+    },
+    {
+      "epoch": 0.32059742966307747,
+      "grad_norm": 0.5523865660394646,
+      "learning_rate": 7.946204021452049e-06,
+      "loss": 0.0194,
+      "step": 923
+    },
+    {
+      "epoch": 0.32094477249044806,
+      "grad_norm": 0.6034008798581597,
+      "learning_rate": 7.941656539020546e-06,
+      "loss": 0.0207,
+      "step": 924
+    },
+    {
+      "epoch": 0.3212921153178187,
+      "grad_norm": 0.5589263526644,
+      "learning_rate": 7.9371053321503e-06,
+      "loss": 0.0185,
+      "step": 925
+    },
+    {
+      "epoch": 0.3216394581451893,
+      "grad_norm": 0.5269322519646701,
+      "learning_rate": 7.932550406603603e-06,
+      "loss": 0.0196,
+      "step": 926
+    },
+    {
+      "epoch": 0.32198680097255994,
+      "grad_norm": 1.2367553916721554,
+      "learning_rate": 7.92799176814746e-06,
+      "loss": 0.0224,
+      "step": 927
+    },
+    {
+      "epoch": 0.32233414379993053,
+      "grad_norm": 0.5712602991453356,
+      "learning_rate": 7.923429422553574e-06,
+      "loss": 0.0239,
+      "step": 928
+    },
+    {
+      "epoch": 0.3226814866273012,
+      "grad_norm": 0.7155877194834468,
+      "learning_rate": 7.91886337559834e-06,
+      "loss": 0.0178,
+      "step": 929
+    },
+    {
+      "epoch": 0.32302882945467176,
+      "grad_norm": 0.9800497349713431,
+      "learning_rate": 7.914293633062845e-06,
+      "loss": 0.0189,
+      "step": 930
+    },
+    {
+      "epoch": 0.32337617228204235,
+      "grad_norm": 1.6132143422048213,
+      "learning_rate": 7.90972020073285e-06,
+      "loss": 0.0229,
+      "step": 931
+    },
+    {
+      "epoch": 0.323723515109413,
+      "grad_norm": 0.7257100719458756,
+      "learning_rate": 7.905143084398792e-06,
+      "loss": 0.0172,
+      "step": 932
+    },
+    {
+      "epoch": 0.3240708579367836,
+      "grad_norm": 1.2683287542082773,
+      "learning_rate": 7.900562289855763e-06,
+      "loss": 0.017,
+      "step": 933
+    },
+    {
+      "epoch": 0.32441820076415423,
+      "grad_norm": 0.7872928535413164,
+      "learning_rate": 7.895977822903524e-06,
+      "loss": 0.0218,
+      "step": 934
+    },
+    {
+      "epoch": 0.3247655435915248,
+      "grad_norm": 0.9240714912057278,
+      "learning_rate": 7.891389689346479e-06,
+      "loss": 0.03,
+      "step": 935
+    },
+    {
+      "epoch": 0.32511288641889546,
+      "grad_norm": 1.2754448071139646,
+      "learning_rate": 7.886797894993674e-06,
+      "loss": 0.0295,
+      "step": 936
+    },
+    {
+      "epoch": 0.32546022924626605,
+      "grad_norm": 0.7875701058115853,
+      "learning_rate": 7.882202445658792e-06,
+      "loss": 0.0191,
+      "step": 937
+    },
+    {
+      "epoch": 0.3258075720736367,
+      "grad_norm": 1.1225423591432484,
+      "learning_rate": 7.877603347160144e-06,
+      "loss": 0.0174,
+      "step": 938
+    },
+    {
+      "epoch": 0.3261549149010073,
+      "grad_norm": 0.7416127576619395,
+      "learning_rate": 7.873000605320658e-06,
+      "loss": 0.0163,
+      "step": 939
+    },
+    {
+      "epoch": 0.32650225772837793,
+      "grad_norm": 0.8331153940172414,
+      "learning_rate": 7.868394225967881e-06,
+      "loss": 0.0247,
+      "step": 940
+    },
+    {
+      "epoch": 0.3268496005557485,
+      "grad_norm": 2.008150547125355,
+      "learning_rate": 7.863784214933957e-06,
+      "loss": 0.031,
+      "step": 941
+    },
+    {
+      "epoch": 0.32719694338311917,
+      "grad_norm": 0.7135068543742416,
+      "learning_rate": 7.859170578055633e-06,
+      "loss": 0.0239,
+      "step": 942
+    },
+    {
+      "epoch": 0.32754428621048975,
+      "grad_norm": 0.7525548520537905,
+      "learning_rate": 7.85455332117425e-06,
+      "loss": 0.0157,
+      "step": 943
+    },
+    {
+      "epoch": 0.32789162903786034,
+      "grad_norm": 0.7034077910253422,
+      "learning_rate": 7.849932450135726e-06,
+      "loss": 0.027,
+      "step": 944
+    },
+    {
+      "epoch": 0.328238971865231,
+      "grad_norm": 0.5421480780633295,
+      "learning_rate": 7.84530797079056e-06,
+      "loss": 0.0206,
+      "step": 945
+    },
+    {
+      "epoch": 0.3285863146926016,
+      "grad_norm": 0.5588408113827876,
+      "learning_rate": 7.84067988899381e-06,
+      "loss": 0.0274,
+      "step": 946
+    },
+    {
+      "epoch": 0.3289336575199722,
+      "grad_norm": 0.43480112336904514,
+      "learning_rate": 7.836048210605109e-06,
+      "loss": 0.0226,
+      "step": 947
+    },
+    {
+      "epoch": 0.3292810003473428,
+      "grad_norm": 0.769442840903224,
+      "learning_rate": 7.831412941488634e-06,
+      "loss": 0.0148,
+      "step": 948
+    },
+    {
+      "epoch": 0.32962834317471346,
+      "grad_norm": 1.3172824990527807,
+      "learning_rate": 7.826774087513113e-06,
+      "loss": 0.0276,
+      "step": 949
+    },
+    {
+      "epoch": 0.32997568600208405,
+      "grad_norm": 2.0103119905473577,
+      "learning_rate": 7.822131654551807e-06,
+      "loss": 0.0295,
+      "step": 950
+    },
+    {
+      "epoch": 0.3303230288294547,
+      "grad_norm": 0.41890011549619927,
+      "learning_rate": 7.817485648482514e-06,
+      "loss": 0.0203,
+      "step": 951
+    },
+    {
+      "epoch": 0.3306703716568253,
+      "grad_norm": 0.34546681427916026,
+      "learning_rate": 7.812836075187555e-06,
+      "loss": 0.0159,
+      "step": 952
+    },
+    {
+      "epoch": 0.3310177144841959,
+      "grad_norm": 1.3371854680462747,
+      "learning_rate": 7.808182940553765e-06,
+      "loss": 0.0306,
+      "step": 953
+    },
+    {
+      "epoch": 0.3313650573115665,
+      "grad_norm": 1.8044729130465431,
+      "learning_rate": 7.803526250472488e-06,
+      "loss": 0.0338,
+      "step": 954
+    },
+    {
+      "epoch": 0.33171240013893716,
+      "grad_norm": 1.0737141556403031,
+      "learning_rate": 7.798866010839577e-06,
+      "loss": 0.016,
+      "step": 955
+    },
+    {
+      "epoch": 0.33205974296630775,
+      "grad_norm": 0.9299379110949423,
+      "learning_rate": 7.794202227555365e-06,
+      "loss": 0.0242,
+      "step": 956
+    },
+    {
+      "epoch": 0.33240708579367834,
+      "grad_norm": 0.6512501538111303,
+      "learning_rate": 7.789534906524684e-06,
+      "loss": 0.0259,
+      "step": 957
+    },
+    {
+      "epoch": 0.332754428621049,
+      "grad_norm": 0.7546601828240426,
+      "learning_rate": 7.784864053656842e-06,
+      "loss": 0.0251,
+      "step": 958
+    },
+    {
+      "epoch": 0.33310177144841957,
+      "grad_norm": 0.4817810206017834,
+      "learning_rate": 7.780189674865617e-06,
+      "loss": 0.0198,
+      "step": 959
+    },
+    {
+      "epoch": 0.3334491142757902,
+      "grad_norm": 1.3719190919474487,
+      "learning_rate": 7.77551177606925e-06,
+      "loss": 0.0219,
+      "step": 960
+    },
+    {
+      "epoch": 0.3337964571031608,
+      "grad_norm": 1.3853549674094732,
+      "learning_rate": 7.770830363190442e-06,
+      "loss": 0.0184,
+      "step": 961
+    },
+    {
+      "epoch": 0.33414379993053145,
+      "grad_norm": 0.9212792032039464,
+      "learning_rate": 7.76614544215634e-06,
+      "loss": 0.0238,
+      "step": 962
+    },
+    {
+      "epoch": 0.33449114275790204,
+      "grad_norm": 0.5577408235128498,
+      "learning_rate": 7.761457018898536e-06,
+      "loss": 0.0222,
+      "step": 963
+    },
+    {
+      "epoch": 0.3348384855852727,
+      "grad_norm": 1.9775404910250387,
+      "learning_rate": 7.756765099353052e-06,
+      "loss": 0.0236,
+      "step": 964
+    },
+    {
+      "epoch": 0.33518582841264327,
+      "grad_norm": 1.3737059966757152,
+      "learning_rate": 7.752069689460345e-06,
+      "loss": 0.025,
+      "step": 965
+    },
+    {
+      "epoch": 0.3355331712400139,
+      "grad_norm": 0.5574547827963353,
+      "learning_rate": 7.747370795165277e-06,
+      "loss": 0.0161,
+      "step": 966
+    },
+    {
+      "epoch": 0.3358805140673845,
+      "grad_norm": 0.9820016637326211,
+      "learning_rate": 7.742668422417137e-06,
+      "loss": 0.0133,
+      "step": 967
+    },
+    {
+      "epoch": 0.33622785689475515,
+      "grad_norm": 0.4953092620978607,
+      "learning_rate": 7.737962577169606e-06,
+      "loss": 0.0134,
+      "step": 968
+    },
+    {
+      "epoch": 0.33657519972212574,
+      "grad_norm": 0.8358656568579974,
+      "learning_rate": 7.73325326538077e-06,
+      "loss": 0.0209,
+      "step": 969
+    },
+    {
+      "epoch": 0.33692254254949633,
+      "grad_norm": 1.3762297024347394,
+      "learning_rate": 7.728540493013098e-06,
+      "loss": 0.0221,
+      "step": 970
+    },
+    {
+      "epoch": 0.337269885376867,
+      "grad_norm": 1.8553914564827385,
+      "learning_rate": 7.723824266033444e-06,
+      "loss": 0.0247,
+      "step": 971
+    },
+    {
+      "epoch": 0.33761722820423756,
+      "grad_norm": 1.3117680989693183,
+      "learning_rate": 7.719104590413036e-06,
+      "loss": 0.0241,
+      "step": 972
+    },
+    {
+      "epoch": 0.3379645710316082,
+      "grad_norm": 1.1924550600193087,
+      "learning_rate": 7.714381472127466e-06,
+      "loss": 0.0251,
+      "step": 973
+    },
+    {
+      "epoch": 0.3383119138589788,
+      "grad_norm": 0.7787134219349535,
+      "learning_rate": 7.709654917156683e-06,
+      "loss": 0.0213,
+      "step": 974
+    },
+    {
+      "epoch": 0.33865925668634944,
+      "grad_norm": 0.5251026227118649,
+      "learning_rate": 7.704924931484997e-06,
+      "loss": 0.0135,
+      "step": 975
+    },
+    {
+      "epoch": 0.33900659951372003,
+      "grad_norm": 0.6878735825950388,
+      "learning_rate": 7.700191521101047e-06,
+      "loss": 0.0193,
+      "step": 976
+    },
+    {
+      "epoch": 0.3393539423410907,
+      "grad_norm": 1.2530620449693388,
+      "learning_rate": 7.695454691997824e-06,
+      "loss": 0.0234,
+      "step": 977
+    },
+    {
+      "epoch": 0.33970128516846126,
+      "grad_norm": 1.9903113138666264,
+      "learning_rate": 7.690714450172633e-06,
+      "loss": 0.027,
+      "step": 978
+    },
+    {
+      "epoch": 0.3400486279958319,
+      "grad_norm": 0.9816472108674068,
+      "learning_rate": 7.685970801627108e-06,
+      "loss": 0.0153,
+      "step": 979
+    },
+    {
+      "epoch": 0.3403959708232025,
+      "grad_norm": 1.218057722488078,
+      "learning_rate": 7.681223752367195e-06,
+      "loss": 0.0261,
+      "step": 980
+    },
+    {
+      "epoch": 0.34074331365057314,
+      "grad_norm": 0.5081769560208316,
+      "learning_rate": 7.676473308403142e-06,
+      "loss": 0.0149,
+      "step": 981
+    },
+    {
+      "epoch": 0.34109065647794373,
+      "grad_norm": 0.541037545259369,
+      "learning_rate": 7.671719475749502e-06,
+      "loss": 0.0179,
+      "step": 982
+    },
+    {
+      "epoch": 0.3414379993053143,
+      "grad_norm": 1.0709500749793777,
+      "learning_rate": 7.666962260425113e-06,
+      "loss": 0.0296,
+      "step": 983
+    },
+    {
+      "epoch": 0.34178534213268497,
+      "grad_norm": 1.0183627382474816,
+      "learning_rate": 7.662201668453098e-06,
+      "loss": 0.0241,
+      "step": 984
+    },
+    {
+      "epoch": 0.34213268496005556,
+      "grad_norm": 0.490710306987966,
+      "learning_rate": 7.657437705860853e-06,
+      "loss": 0.0083,
+      "step": 985
+    },
+    {
+      "epoch": 0.3424800277874262,
+      "grad_norm": 0.9120936689637056,
+      "learning_rate": 7.652670378680043e-06,
+      "loss": 0.0183,
+      "step": 986
+    },
+    {
+      "epoch": 0.3428273706147968,
+      "grad_norm": 0.8239513692459495,
+      "learning_rate": 7.647899692946594e-06,
+      "loss": 0.0199,
+      "step": 987
+    },
+    {
+      "epoch": 0.34317471344216743,
+      "grad_norm": 0.6785851737249843,
+      "learning_rate": 7.643125654700684e-06,
+      "loss": 0.0203,
+      "step": 988
+    },
+    {
+      "epoch": 0.343522056269538,
+      "grad_norm": 0.9952364860290294,
+      "learning_rate": 7.638348269986733e-06,
+      "loss": 0.0233,
+      "step": 989
+    },
+    {
+      "epoch": 0.34386939909690867,
+      "grad_norm": 0.7687287712488833,
+      "learning_rate": 7.6335675448534e-06,
+      "loss": 0.0143,
+      "step": 990
+    },
+    {
+      "epoch": 0.34421674192427926,
+      "grad_norm": 1.0499353749760423,
+      "learning_rate": 7.628783485353573e-06,
+      "loss": 0.0259,
+      "step": 991
+    },
+    {
+      "epoch": 0.3445640847516499,
+      "grad_norm": 1.474581388095261,
+      "learning_rate": 7.623996097544364e-06,
+      "loss": 0.0258,
+      "step": 992
+    },
+    {
+      "epoch": 0.3449114275790205,
+      "grad_norm": 1.8524514914939911,
+      "learning_rate": 7.619205387487094e-06,
+      "loss": 0.0263,
+      "step": 993
+    },
+    {
+      "epoch": 0.34525877040639114,
+      "grad_norm": 1.9787996414155324,
+      "learning_rate": 7.614411361247296e-06,
+      "loss": 0.0241,
+      "step": 994
+    },
+    {
+      "epoch": 0.3456061132337617,
+      "grad_norm": 0.9020458030038447,
+      "learning_rate": 7.609614024894694e-06,
+      "loss": 0.0282,
+      "step": 995
+    },
+    {
+      "epoch": 0.3459534560611323,
+      "grad_norm": 1.6705983295073557,
+      "learning_rate": 7.604813384503212e-06,
+      "loss": 0.0297,
+      "step": 996
+    },
+    {
+      "epoch": 0.34630079888850296,
+      "grad_norm": 0.6271112596802708,
+      "learning_rate": 7.600009446150951e-06,
+      "loss": 0.0202,
+      "step": 997
+    },
+    {
+      "epoch": 0.34664814171587355,
+      "grad_norm": 0.43730341616505064,
+      "learning_rate": 7.59520221592019e-06,
+      "loss": 0.0165,
+      "step": 998
+    },
+    {
+      "epoch": 0.3469954845432442,
+      "grad_norm": 0.6300030347921803,
+      "learning_rate": 7.5903916998973745e-06,
+      "loss": 0.0226,
+      "step": 999
+    },
+    {
+      "epoch": 0.3473428273706148,
+      "grad_norm": 0.6996825232281534,
+      "learning_rate": 7.585577904173113e-06,
+      "loss": 0.0172,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3476901701979854,
+      "grad_norm": 0.6380166925486069,
+      "learning_rate": 7.580760834842162e-06,
+      "loss": 0.0201,
+      "step": 1001
+    },
+    {
+      "epoch": 0.348037513025356,
+      "grad_norm": 1.3568481449306653,
+      "learning_rate": 7.575940498003425e-06,
+      "loss": 0.0212,
+      "step": 1002
+    },
+    {
+      "epoch": 0.34838485585272666,
+      "grad_norm": 1.2549578263294605,
+      "learning_rate": 7.571116899759945e-06,
+      "loss": 0.0238,
+      "step": 1003
+    },
+    {
+      "epoch": 0.34873219868009725,
+      "grad_norm": 1.0886856632970126,
+      "learning_rate": 7.566290046218889e-06,
+      "loss": 0.0243,
+      "step": 1004
+    },
+    {
+      "epoch": 0.3490795415074679,
+      "grad_norm": 0.4043124260071018,
+      "learning_rate": 7.5614599434915514e-06,
+      "loss": 0.0149,
+      "step": 1005
+    },
+    {
+      "epoch": 0.3494268843348385,
+      "grad_norm": 0.4143945807665345,
+      "learning_rate": 7.556626597693335e-06,
+      "loss": 0.0172,
+      "step": 1006
+    },
+    {
+      "epoch": 0.34977422716220913,
+      "grad_norm": 0.47662985164661864,
+      "learning_rate": 7.551790014943752e-06,
+      "loss": 0.0194,
+      "step": 1007
+    },
+    {
+      "epoch": 0.3501215699895797,
+      "grad_norm": 0.9261090190118699,
+      "learning_rate": 7.546950201366412e-06,
+      "loss": 0.0261,
+      "step": 1008
+    },
+    {
+      "epoch": 0.3504689128169503,
+      "grad_norm": 1.2825084254203554,
+      "learning_rate": 7.542107163089016e-06,
+      "loss": 0.0151,
+      "step": 1009
+    },
+    {
+      "epoch": 0.35081625564432095,
+      "grad_norm": 0.8371175718056099,
+      "learning_rate": 7.537260906243344e-06,
+      "loss": 0.0267,
+      "step": 1010
+    },
+    {
+      "epoch": 0.35116359847169154,
+      "grad_norm": 0.6892235320884719,
+      "learning_rate": 7.532411436965258e-06,
+      "loss": 0.0253,
+      "step": 1011
+    },
+    {
+      "epoch": 0.3515109412990622,
+      "grad_norm": 0.731192690685414,
+      "learning_rate": 7.52755876139468e-06,
+      "loss": 0.0139,
+      "step": 1012
+    },
+    {
+      "epoch": 0.3518582841264328,
+      "grad_norm": 0.6312927924236015,
+      "learning_rate": 7.522702885675597e-06,
+      "loss": 0.0228,
+      "step": 1013
+    },
+    {
+      "epoch": 0.3522056269538034,
+      "grad_norm": 0.9551305586895227,
+      "learning_rate": 7.517843815956045e-06,
+      "loss": 0.0178,
+      "step": 1014
+    },
+    {
+      "epoch": 0.352552969781174,
+      "grad_norm": 0.5677401120338555,
+      "learning_rate": 7.512981558388101e-06,
+      "loss": 0.0225,
+      "step": 1015
+    },
+    {
+      "epoch": 0.35290031260854465,
+      "grad_norm": 0.7040980402544591,
+      "learning_rate": 7.5081161191278874e-06,
+      "loss": 0.0176,
+      "step": 1016
+    },
+    {
+      "epoch": 0.35324765543591524,
+      "grad_norm": 0.9674600691403624,
+      "learning_rate": 7.5032475043355444e-06,
+      "loss": 0.0243,
+      "step": 1017
+    },
+    {
+      "epoch": 0.3535949982632859,
+      "grad_norm": 0.6023226292917273,
+      "learning_rate": 7.498375720175239e-06,
+      "loss": 0.0183,
+      "step": 1018
+    },
+    {
+      "epoch": 0.3539423410906565,
+      "grad_norm": 0.9673022657430923,
+      "learning_rate": 7.49350077281515e-06,
+      "loss": 0.0287,
+      "step": 1019
+    },
+    {
+      "epoch": 0.35428968391802707,
+      "grad_norm": 0.606921696103078,
+      "learning_rate": 7.48862266842746e-06,
+      "loss": 0.0226,
+      "step": 1020
+    },
+    {
+      "epoch": 0.3546370267453977,
+      "grad_norm": 0.9175466939892832,
+      "learning_rate": 7.483741413188349e-06,
+      "loss": 0.0205,
+      "step": 1021
+    },
+    {
+      "epoch": 0.3549843695727683,
+      "grad_norm": 0.5474768689731467,
+      "learning_rate": 7.478857013277987e-06,
+      "loss": 0.0253,
+      "step": 1022
+    },
+    {
+      "epoch": 0.35533171240013894,
+      "grad_norm": 0.7005547254579504,
+      "learning_rate": 7.473969474880527e-06,
+      "loss": 0.0239,
+      "step": 1023
+    },
+    {
+      "epoch": 0.35567905522750953,
+      "grad_norm": 0.6566673110253187,
+      "learning_rate": 7.469078804184088e-06,
+      "loss": 0.0153,
+      "step": 1024
+    },
+    {
+      "epoch": 0.3560263980548802,
+      "grad_norm": 0.44731824465158737,
+      "learning_rate": 7.464185007380767e-06,
+      "loss": 0.0226,
+      "step": 1025
+    },
+    {
+      "epoch": 0.35637374088225077,
+      "grad_norm": 0.929371757186016,
+      "learning_rate": 7.459288090666605e-06,
+      "loss": 0.0215,
+      "step": 1026
+    },
+    {
+      "epoch": 0.3567210837096214,
+      "grad_norm": 1.754520347098642,
+      "learning_rate": 7.45438806024161e-06,
+      "loss": 0.0245,
+      "step": 1027
+    },
+    {
+      "epoch": 0.357068426536992,
+      "grad_norm": 1.154684337826048,
+      "learning_rate": 7.449484922309713e-06,
+      "loss": 0.0183,
+      "step": 1028
+    },
+    {
+      "epoch": 0.35741576936436265,
+      "grad_norm": 0.7743738340491029,
+      "learning_rate": 7.444578683078798e-06,
+      "loss": 0.0265,
+      "step": 1029
+    },
+    {
+      "epoch": 0.35776311219173323,
+      "grad_norm": 0.39037178440555265,
+      "learning_rate": 7.4396693487606605e-06,
+      "loss": 0.0183,
+      "step": 1030
+    },
+    {
+      "epoch": 0.3581104550191039,
+      "grad_norm": 0.6375860488882318,
+      "learning_rate": 7.4347569255710254e-06,
+      "loss": 0.0309,
+      "step": 1031
+    },
+    {
+      "epoch": 0.35845779784647447,
+      "grad_norm": 0.6020351665546125,
+      "learning_rate": 7.429841419729521e-06,
+      "loss": 0.025,
+      "step": 1032
+    },
+    {
+      "epoch": 0.35880514067384506,
+      "grad_norm": 1.3286872104930505,
+      "learning_rate": 7.424922837459683e-06,
+      "loss": 0.0224,
+      "step": 1033
+    },
+    {
+      "epoch": 0.3591524835012157,
+      "grad_norm": 0.5744413112024466,
+      "learning_rate": 7.42000118498894e-06,
+      "loss": 0.0184,
+      "step": 1034
+    },
+    {
+      "epoch": 0.3594998263285863,
+      "grad_norm": 1.1943187896682033,
+      "learning_rate": 7.41507646854861e-06,
+      "loss": 0.0262,
+      "step": 1035
+    },
+    {
+      "epoch": 0.35984716915595694,
+      "grad_norm": 0.5497582747698295,
+      "learning_rate": 7.4101486943738865e-06,
+      "loss": 0.0211,
+      "step": 1036
+    },
+    {
+      "epoch": 0.3601945119833275,
+      "grad_norm": 0.4778102052106275,
+      "learning_rate": 7.405217868703839e-06,
+      "loss": 0.0165,
+      "step": 1037
+    },
+    {
+      "epoch": 0.36054185481069817,
+      "grad_norm": 0.7214250751175056,
+      "learning_rate": 7.400283997781397e-06,
+      "loss": 0.0286,
+      "step": 1038
+    },
+    {
+      "epoch": 0.36088919763806876,
+      "grad_norm": 0.7687435033252493,
+      "learning_rate": 7.395347087853349e-06,
+      "loss": 0.0214,
+      "step": 1039
+    },
+    {
+      "epoch": 0.3612365404654394,
+      "grad_norm": 1.0790158493374187,
+      "learning_rate": 7.390407145170325e-06,
+      "loss": 0.027,
+      "step": 1040
+    },
+    {
+      "epoch": 0.36158388329281,
+      "grad_norm": 0.37118096549279656,
+      "learning_rate": 7.385464175986803e-06,
+      "loss": 0.0124,
+      "step": 1041
+    },
+    {
+      "epoch": 0.36193122612018064,
+      "grad_norm": 1.4499606402413763,
+      "learning_rate": 7.380518186561086e-06,
+      "loss": 0.0224,
+      "step": 1042
+    },
+    {
+      "epoch": 0.3622785689475512,
+      "grad_norm": 1.7916464514957589,
+      "learning_rate": 7.375569183155306e-06,
+      "loss": 0.0269,
+      "step": 1043
+    },
+    {
+      "epoch": 0.36262591177492187,
+      "grad_norm": 0.6417320309447412,
+      "learning_rate": 7.370617172035406e-06,
+      "loss": 0.0187,
+      "step": 1044
+    },
+    {
+      "epoch": 0.36297325460229246,
+      "grad_norm": 0.9602465969222208,
+      "learning_rate": 7.365662159471142e-06,
+      "loss": 0.0171,
+      "step": 1045
+    },
+    {
+      "epoch": 0.36332059742966305,
+      "grad_norm": 0.5700015399771756,
+      "learning_rate": 7.3607041517360666e-06,
+      "loss": 0.023,
+      "step": 1046
+    },
+    {
+      "epoch": 0.3636679402570337,
+      "grad_norm": 0.42616867425402144,
+      "learning_rate": 7.355743155107526e-06,
+      "loss": 0.0185,
+      "step": 1047
+    },
+    {
+      "epoch": 0.3640152830844043,
+      "grad_norm": 0.5254170573018199,
+      "learning_rate": 7.3507791758666514e-06,
+      "loss": 0.0125,
+      "step": 1048
+    },
+    {
+      "epoch": 0.36436262591177493,
+      "grad_norm": 0.8659700459185796,
+      "learning_rate": 7.3458122202983495e-06,
+      "loss": 0.0184,
+      "step": 1049
+    },
+    {
+      "epoch": 0.3647099687391455,
+      "grad_norm": 0.78273420283934,
+      "learning_rate": 7.340842294691292e-06,
+      "loss": 0.0134,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36505731156651616,
+      "grad_norm": 1.2893737814650572,
+      "learning_rate": 7.335869405337919e-06,
+      "loss": 0.0188,
+      "step": 1051
+    },
+    {
+      "epoch": 0.36540465439388675,
+      "grad_norm": 0.901363581151114,
+      "learning_rate": 7.3308935585344135e-06,
+      "loss": 0.0217,
+      "step": 1052
+    },
+    {
+      "epoch": 0.3657519972212574,
+      "grad_norm": 0.481669992576634,
+      "learning_rate": 7.325914760580712e-06,
+      "loss": 0.0102,
+      "step": 1053
+    },
+    {
+      "epoch": 0.366099340048628,
+      "grad_norm": 1.0260877050410049,
+      "learning_rate": 7.32093301778048e-06,
+      "loss": 0.0263,
+      "step": 1054
+    },
+    {
+      "epoch": 0.36644668287599863,
+      "grad_norm": 0.8353120616059849,
+      "learning_rate": 7.3159483364411175e-06,
+      "loss": 0.022,
+      "step": 1055
+    },
+    {
+      "epoch": 0.3667940257033692,
+      "grad_norm": 0.7273624057486092,
+      "learning_rate": 7.310960722873739e-06,
+      "loss": 0.0308,
+      "step": 1056
+    },
+    {
+      "epoch": 0.36714136853073986,
+      "grad_norm": 1.13748674597797,
+      "learning_rate": 7.3059701833931766e-06,
+      "loss": 0.0135,
+      "step": 1057
+    },
+    {
+      "epoch": 0.36748871135811045,
+      "grad_norm": 0.6859340482408613,
+      "learning_rate": 7.300976724317964e-06,
+      "loss": 0.016,
+      "step": 1058
+    },
+    {
+      "epoch": 0.36783605418548104,
+      "grad_norm": 1.2719149842779836,
+      "learning_rate": 7.295980351970331e-06,
+      "loss": 0.0178,
+      "step": 1059
+    },
+    {
+      "epoch": 0.3681833970128517,
+      "grad_norm": 1.0679079371197404,
+      "learning_rate": 7.290981072676202e-06,
+      "loss": 0.0333,
+      "step": 1060
+    },
+    {
+      "epoch": 0.3685307398402223,
+      "grad_norm": 0.5473217513456246,
+      "learning_rate": 7.285978892765171e-06,
+      "loss": 0.0196,
+      "step": 1061
+    },
+    {
+      "epoch": 0.3688780826675929,
+      "grad_norm": 0.6477963035607689,
+      "learning_rate": 7.280973818570515e-06,
+      "loss": 0.0229,
+      "step": 1062
+    },
+    {
+      "epoch": 0.3692254254949635,
+      "grad_norm": 0.9779272803664357,
+      "learning_rate": 7.275965856429167e-06,
+      "loss": 0.0148,
+      "step": 1063
+    },
+    {
+      "epoch": 0.36957276832233416,
+      "grad_norm": 0.6549296775726913,
+      "learning_rate": 7.270955012681726e-06,
+      "loss": 0.0219,
+      "step": 1064
+    },
+    {
+      "epoch": 0.36992011114970474,
+      "grad_norm": 0.8692466378138357,
+      "learning_rate": 7.26594129367243e-06,
+      "loss": 0.0202,
+      "step": 1065
+    },
+    {
+      "epoch": 0.3702674539770754,
+      "grad_norm": 0.6315864672971158,
+      "learning_rate": 7.26092470574916e-06,
+      "loss": 0.0181,
+      "step": 1066
+    },
+    {
+      "epoch": 0.370614796804446,
+      "grad_norm": 0.8066486900818035,
+      "learning_rate": 7.255905255263434e-06,
+      "loss": 0.0234,
+      "step": 1067
+    },
+    {
+      "epoch": 0.3709621396318166,
+      "grad_norm": 0.8450924908715876,
+      "learning_rate": 7.25088294857039e-06,
+      "loss": 0.0212,
+      "step": 1068
+    },
+    {
+      "epoch": 0.3713094824591872,
+      "grad_norm": 0.7148838980127935,
+      "learning_rate": 7.245857792028781e-06,
+      "loss": 0.0204,
+      "step": 1069
+    },
+    {
+      "epoch": 0.37165682528655786,
+      "grad_norm": 0.7769253902283783,
+      "learning_rate": 7.240829792000974e-06,
+      "loss": 0.0184,
+      "step": 1070
+    },
+    {
+      "epoch": 0.37200416811392845,
+      "grad_norm": 0.5818321434788865,
+      "learning_rate": 7.235798954852929e-06,
+      "loss": 0.0218,
+      "step": 1071
+    },
+    {
+      "epoch": 0.37235151094129904,
+      "grad_norm": 0.7181155976463919,
+      "learning_rate": 7.230765286954204e-06,
+      "loss": 0.0229,
+      "step": 1072
+    },
+    {
+      "epoch": 0.3726988537686697,
+      "grad_norm": 0.5311889445893856,
+      "learning_rate": 7.2257287946779365e-06,
+      "loss": 0.0208,
+      "step": 1073
+    },
+    {
+      "epoch": 0.37304619659604027,
+      "grad_norm": 0.6470157370215145,
+      "learning_rate": 7.220689484400844e-06,
+      "loss": 0.0286,
+      "step": 1074
+    },
+    {
+      "epoch": 0.3733935394234109,
+      "grad_norm": 0.5839896666697149,
+      "learning_rate": 7.2156473625032075e-06,
+      "loss": 0.0208,
+      "step": 1075
+    },
+    {
+      "epoch": 0.3737408822507815,
+      "grad_norm": 0.5003859905535379,
+      "learning_rate": 7.210602435368873e-06,
+      "loss": 0.0277,
+      "step": 1076
+    },
+    {
+      "epoch": 0.37408822507815215,
+      "grad_norm": 0.938126994232941,
+      "learning_rate": 7.205554709385234e-06,
+      "loss": 0.0193,
+      "step": 1077
+    },
+    {
+      "epoch": 0.37443556790552274,
+      "grad_norm": 0.9825447432197518,
+      "learning_rate": 7.20050419094323e-06,
+      "loss": 0.0327,
+      "step": 1078
+    },
+    {
+      "epoch": 0.3747829107328934,
+      "grad_norm": 0.5808094753024735,
+      "learning_rate": 7.195450886437334e-06,
+      "loss": 0.0233,
+      "step": 1079
+    },
+    {
+      "epoch": 0.37513025356026397,
+      "grad_norm": 0.729519930229865,
+      "learning_rate": 7.190394802265548e-06,
+      "loss": 0.0254,
+      "step": 1080
+    },
+    {
+      "epoch": 0.3754775963876346,
+      "grad_norm": 0.32897954890179926,
+      "learning_rate": 7.185335944829391e-06,
+      "loss": 0.0193,
+      "step": 1081
+    },
+    {
+      "epoch": 0.3758249392150052,
+      "grad_norm": 0.4730691255515786,
+      "learning_rate": 7.1802743205339e-06,
+      "loss": 0.0255,
+      "step": 1082
+    },
+    {
+      "epoch": 0.37617228204237585,
+      "grad_norm": 0.8306626500245984,
+      "learning_rate": 7.175209935787605e-06,
+      "loss": 0.0261,
+      "step": 1083
+    },
+    {
+      "epoch": 0.37651962486974644,
+      "grad_norm": 0.5382124730055174,
+      "learning_rate": 7.17014279700254e-06,
+      "loss": 0.025,
+      "step": 1084
+    },
+    {
+      "epoch": 0.37686696769711703,
+      "grad_norm": 0.6717725725696359,
+      "learning_rate": 7.165072910594219e-06,
+      "loss": 0.0271,
+      "step": 1085
+    },
+    {
+      "epoch": 0.3772143105244877,
+      "grad_norm": 0.7370235521381453,
+      "learning_rate": 7.160000282981641e-06,
+      "loss": 0.0273,
+      "step": 1086
+    },
+    {
+      "epoch": 0.37756165335185826,
+      "grad_norm": 0.4891684754047165,
+      "learning_rate": 7.154924920587269e-06,
+      "loss": 0.0264,
+      "step": 1087
+    },
+    {
+      "epoch": 0.3779089961792289,
+      "grad_norm": 0.3231257861866142,
+      "learning_rate": 7.149846829837036e-06,
+      "loss": 0.0206,
+      "step": 1088
+    },
+    {
+      "epoch": 0.3782563390065995,
+      "grad_norm": 0.4101311080573536,
+      "learning_rate": 7.144766017160324e-06,
+      "loss": 0.0205,
+      "step": 1089
+    },
+    {
+      "epoch": 0.37860368183397014,
+      "grad_norm": 0.8294757845956588,
+      "learning_rate": 7.139682488989961e-06,
+      "loss": 0.0265,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37895102466134073,
+      "grad_norm": 0.5156785295009514,
+      "learning_rate": 7.134596251762217e-06,
+      "loss": 0.0209,
+      "step": 1091
+    },
+    {
+      "epoch": 0.3792983674887114,
+      "grad_norm": 0.6068386255398981,
+      "learning_rate": 7.129507311916789e-06,
+      "loss": 0.0191,
+      "step": 1092
+    },
+    {
+      "epoch": 0.37964571031608196,
+      "grad_norm": 0.42053223931776523,
+      "learning_rate": 7.124415675896796e-06,
+      "loss": 0.018,
+      "step": 1093
+    },
+    {
+      "epoch": 0.3799930531434526,
+      "grad_norm": 1.529702363413615,
+      "learning_rate": 7.119321350148772e-06,
+      "loss": 0.0305,
+      "step": 1094
+    },
+    {
+      "epoch": 0.3803403959708232,
+      "grad_norm": 0.9603413062647468,
+      "learning_rate": 7.114224341122655e-06,
+      "loss": 0.0211,
+      "step": 1095
+    },
+    {
+      "epoch": 0.38068773879819384,
+      "grad_norm": 0.46028100214920925,
+      "learning_rate": 7.109124655271782e-06,
+      "loss": 0.0121,
+      "step": 1096
+    },
+    {
+      "epoch": 0.38103508162556443,
+      "grad_norm": 0.38425802665843994,
+      "learning_rate": 7.1040222990528775e-06,
+      "loss": 0.0165,
+      "step": 1097
+    },
+    {
+      "epoch": 0.381382424452935,
+      "grad_norm": 0.34490477943437864,
+      "learning_rate": 7.098917278926046e-06,
+      "loss": 0.0215,
+      "step": 1098
+    },
+    {
+      "epoch": 0.38172976728030567,
+      "grad_norm": 0.4917657059195774,
+      "learning_rate": 7.093809601354769e-06,
+      "loss": 0.0193,
+      "step": 1099
+    },
+    {
+      "epoch": 0.38207711010767625,
+      "grad_norm": 0.3304192390082568,
+      "learning_rate": 7.088699272805888e-06,
+      "loss": 0.0161,
+      "step": 1100
+    },
+    {
+      "epoch": 0.3824244529350469,
+      "grad_norm": 0.8538379691183342,
+      "learning_rate": 7.0835862997496045e-06,
+      "loss": 0.0174,
+      "step": 1101
+    },
+    {
+      "epoch": 0.3827717957624175,
+      "grad_norm": 1.066827953382512,
+      "learning_rate": 7.078470688659465e-06,
+      "loss": 0.0257,
+      "step": 1102
+    },
+    {
+      "epoch": 0.38311913858978813,
+      "grad_norm": 0.7479008835369523,
+      "learning_rate": 7.073352446012357e-06,
+      "loss": 0.0213,
+      "step": 1103
+    },
+    {
+      "epoch": 0.3834664814171587,
+      "grad_norm": 1.0510022725835733,
+      "learning_rate": 7.068231578288502e-06,
+      "loss": 0.0259,
+      "step": 1104
+    },
+    {
+      "epoch": 0.38381382424452937,
+      "grad_norm": 0.6431698987590786,
+      "learning_rate": 7.063108091971444e-06,
+      "loss": 0.0161,
+      "step": 1105
+    },
+    {
+      "epoch": 0.38416116707189996,
+      "grad_norm": 0.626180547232632,
+      "learning_rate": 7.05798199354804e-06,
+      "loss": 0.015,
+      "step": 1106
+    },
+    {
+      "epoch": 0.3845085098992706,
+      "grad_norm": 0.7399567466553991,
+      "learning_rate": 7.052853289508458e-06,
+      "loss": 0.0229,
+      "step": 1107
+    },
+    {
+      "epoch": 0.3848558527266412,
+      "grad_norm": 0.5500014449728272,
+      "learning_rate": 7.047721986346161e-06,
+      "loss": 0.0232,
+      "step": 1108
+    },
+    {
+      "epoch": 0.38520319555401183,
+      "grad_norm": 1.4534975737472637,
+      "learning_rate": 7.042588090557906e-06,
+      "loss": 0.0228,
+      "step": 1109
+    },
+    {
+      "epoch": 0.3855505383813824,
+      "grad_norm": 1.3334339962937882,
+      "learning_rate": 7.037451608643732e-06,
+      "loss": 0.0201,
+      "step": 1110
+    },
+    {
+      "epoch": 0.385897881208753,
+      "grad_norm": 1.0089775574518236,
+      "learning_rate": 7.03231254710695e-06,
+      "loss": 0.0184,
+      "step": 1111
+    },
+    {
+      "epoch": 0.38624522403612366,
+      "grad_norm": 0.439074094269593,
+      "learning_rate": 7.027170912454141e-06,
+      "loss": 0.0162,
+      "step": 1112
+    },
+    {
+      "epoch": 0.38659256686349425,
+      "grad_norm": 0.8348064008657952,
+      "learning_rate": 7.02202671119514e-06,
+      "loss": 0.017,
+      "step": 1113
+    },
+    {
+      "epoch": 0.3869399096908649,
+      "grad_norm": 0.7951186556024389,
+      "learning_rate": 7.016879949843032e-06,
+      "loss": 0.0163,
+      "step": 1114
+    },
+    {
+      "epoch": 0.3872872525182355,
+      "grad_norm": 0.7523417760723944,
+      "learning_rate": 7.0117306349141485e-06,
+      "loss": 0.0236,
+      "step": 1115
+    },
+    {
+      "epoch": 0.3876345953456061,
+      "grad_norm": 1.3735341548254056,
+      "learning_rate": 7.006578772928045e-06,
+      "loss": 0.0218,
+      "step": 1116
+    },
+    {
+      "epoch": 0.3879819381729767,
+      "grad_norm": 0.5385452791468442,
+      "learning_rate": 7.0014243704075115e-06,
+      "loss": 0.0205,
+      "step": 1117
+    },
+    {
+      "epoch": 0.38832928100034736,
+      "grad_norm": 0.8605606189346207,
+      "learning_rate": 6.996267433878545e-06,
+      "loss": 0.0222,
+      "step": 1118
+    },
+    {
+      "epoch": 0.38867662382771795,
+      "grad_norm": 0.823981661806264,
+      "learning_rate": 6.991107969870363e-06,
+      "loss": 0.0264,
+      "step": 1119
+    },
+    {
+      "epoch": 0.3890239666550886,
+      "grad_norm": 0.7575159321093746,
+      "learning_rate": 6.985945984915368e-06,
+      "loss": 0.0229,
+      "step": 1120
+    },
+    {
+      "epoch": 0.3893713094824592,
+      "grad_norm": 0.45426836006848514,
+      "learning_rate": 6.9807814855491685e-06,
+      "loss": 0.0162,
+      "step": 1121
+    },
+    {
+      "epoch": 0.3897186523098298,
+      "grad_norm": 1.2049213408577124,
+      "learning_rate": 6.975614478310546e-06,
+      "loss": 0.0322,
+      "step": 1122
+    },
+    {
+      "epoch": 0.3900659951372004,
+      "grad_norm": 0.7306802680657365,
+      "learning_rate": 6.970444969741462e-06,
+      "loss": 0.0198,
+      "step": 1123
+    },
+    {
+      "epoch": 0.390413337964571,
+      "grad_norm": 0.7391729308842311,
+      "learning_rate": 6.965272966387046e-06,
+      "loss": 0.0184,
+      "step": 1124
+    },
+    {
+      "epoch": 0.39076068079194165,
+      "grad_norm": 0.6086940453113556,
+      "learning_rate": 6.960098474795583e-06,
+      "loss": 0.0243,
+      "step": 1125
+    },
+    {
+      "epoch": 0.39110802361931224,
+      "grad_norm": 0.4851910309801865,
+      "learning_rate": 6.954921501518511e-06,
+      "loss": 0.012,
+      "step": 1126
+    },
+    {
+      "epoch": 0.3914553664466829,
+      "grad_norm": 0.45385633767002753,
+      "learning_rate": 6.949742053110408e-06,
+      "loss": 0.0198,
+      "step": 1127
+    },
+    {
+      "epoch": 0.3918027092740535,
+      "grad_norm": 1.1129782341928363,
+      "learning_rate": 6.944560136128986e-06,
+      "loss": 0.0287,
+      "step": 1128
+    },
+    {
+      "epoch": 0.3921500521014241,
+      "grad_norm": 0.7720433053769181,
+      "learning_rate": 6.939375757135085e-06,
+      "loss": 0.0238,
+      "step": 1129
+    },
+    {
+      "epoch": 0.3924973949287947,
+      "grad_norm": 1.5172156671202623,
+      "learning_rate": 6.934188922692659e-06,
+      "loss": 0.0254,
+      "step": 1130
+    },
+    {
+      "epoch": 0.39284473775616535,
+      "grad_norm": 0.5132675364055385,
+      "learning_rate": 6.928999639368773e-06,
+      "loss": 0.0176,
+      "step": 1131
+    },
+    {
+      "epoch": 0.39319208058353594,
+      "grad_norm": 0.9917189063692697,
+      "learning_rate": 6.923807913733591e-06,
+      "loss": 0.0212,
+      "step": 1132
+    },
+    {
+      "epoch": 0.3935394234109066,
+      "grad_norm": 1.0346614661684557,
+      "learning_rate": 6.918613752360369e-06,
+      "loss": 0.0237,
+      "step": 1133
+    },
+    {
+      "epoch": 0.3938867662382772,
+      "grad_norm": 0.7774794514027609,
+      "learning_rate": 6.913417161825449e-06,
+      "loss": 0.021,
+      "step": 1134
+    },
+    {
+      "epoch": 0.3942341090656478,
+      "grad_norm": 0.8284608514835325,
+      "learning_rate": 6.908218148708248e-06,
+      "loss": 0.0196,
+      "step": 1135
+    },
+    {
+      "epoch": 0.3945814518930184,
+      "grad_norm": 0.5659423206737776,
+      "learning_rate": 6.903016719591249e-06,
+      "loss": 0.0221,
+      "step": 1136
+    },
+    {
+      "epoch": 0.394928794720389,
+      "grad_norm": 1.0709171868083134,
+      "learning_rate": 6.8978128810599935e-06,
+      "loss": 0.035,
+      "step": 1137
+    },
+    {
+      "epoch": 0.39527613754775964,
+      "grad_norm": 0.6527457095839493,
+      "learning_rate": 6.8926066397030745e-06,
+      "loss": 0.0204,
+      "step": 1138
+    },
+    {
+      "epoch": 0.39562348037513023,
+      "grad_norm": 0.5573043903274787,
+      "learning_rate": 6.887398002112129e-06,
+      "loss": 0.0211,
+      "step": 1139
+    },
+    {
+      "epoch": 0.3959708232025009,
+      "grad_norm": 0.8397780950700933,
+      "learning_rate": 6.8821869748818235e-06,
+      "loss": 0.018,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39631816602987147,
+      "grad_norm": 0.39026678264711556,
+      "learning_rate": 6.876973564609857e-06,
+      "loss": 0.0175,
+      "step": 1141
+    },
+    {
+      "epoch": 0.3966655088572421,
+      "grad_norm": 0.5775231347276166,
+      "learning_rate": 6.871757777896937e-06,
+      "loss": 0.0176,
+      "step": 1142
+    },
+    {
+      "epoch": 0.3970128516846127,
+      "grad_norm": 0.3504293355564611,
+      "learning_rate": 6.866539621346786e-06,
+      "loss": 0.0136,
+      "step": 1143
+    },
+    {
+      "epoch": 0.39736019451198334,
+      "grad_norm": 0.3267695926847211,
+      "learning_rate": 6.861319101566126e-06,
+      "loss": 0.0152,
+      "step": 1144
+    },
+    {
+      "epoch": 0.39770753733935393,
+      "grad_norm": 1.0251884908691054,
+      "learning_rate": 6.856096225164669e-06,
+      "loss": 0.017,
+      "step": 1145
+    },
+    {
+      "epoch": 0.3980548801667246,
+      "grad_norm": 0.484956168822129,
+      "learning_rate": 6.850870998755111e-06,
+      "loss": 0.0192,
+      "step": 1146
+    },
+    {
+      "epoch": 0.39840222299409517,
+      "grad_norm": 0.6237901046105527,
+      "learning_rate": 6.845643428953127e-06,
+      "loss": 0.0322,
+      "step": 1147
+    },
+    {
+      "epoch": 0.3987495658214658,
+      "grad_norm": 1.2212183860294297,
+      "learning_rate": 6.840413522377355e-06,
+      "loss": 0.0211,
+      "step": 1148
+    },
+    {
+      "epoch": 0.3990969086488364,
+      "grad_norm": 0.5405735574477039,
+      "learning_rate": 6.8351812856493905e-06,
+      "loss": 0.0227,
+      "step": 1149
+    },
+    {
+      "epoch": 0.399444251476207,
+      "grad_norm": 1.7602429727127982,
+      "learning_rate": 6.829946725393787e-06,
+      "loss": 0.0264,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39979159430357764,
+      "grad_norm": 0.9253181842069885,
+      "learning_rate": 6.824709848238028e-06,
+      "loss": 0.0167,
+      "step": 1151
+    },
+    {
+      "epoch": 0.4001389371309482,
+      "grad_norm": 0.39829378649964736,
+      "learning_rate": 6.819470660812543e-06,
+      "loss": 0.0153,
+      "step": 1152
+    },
+    {
+      "epoch": 0.40048627995831887,
+      "grad_norm": 1.4005769801829058,
+      "learning_rate": 6.814229169750675e-06,
+      "loss": 0.0239,
+      "step": 1153
+    },
+    {
+      "epoch": 0.40083362278568946,
+      "grad_norm": 0.599123382279925,
+      "learning_rate": 6.808985381688692e-06,
+      "loss": 0.0239,
+      "step": 1154
+    },
+    {
+      "epoch": 0.4011809656130601,
+      "grad_norm": 0.502770976659655,
+      "learning_rate": 6.8037393032657665e-06,
+      "loss": 0.0161,
+      "step": 1155
+    },
+    {
+      "epoch": 0.4015283084404307,
+      "grad_norm": 0.715985711190083,
+      "learning_rate": 6.798490941123972e-06,
+      "loss": 0.0321,
+      "step": 1156
+    },
+    {
+      "epoch": 0.40187565126780134,
+      "grad_norm": 1.3666815307149158,
+      "learning_rate": 6.793240301908273e-06,
+      "loss": 0.0217,
+      "step": 1157
+    },
+    {
+      "epoch": 0.4022229940951719,
+      "grad_norm": 1.3893597610770267,
+      "learning_rate": 6.7879873922665164e-06,
+      "loss": 0.0255,
+      "step": 1158
+    },
+    {
+      "epoch": 0.40257033692254257,
+      "grad_norm": 1.699863532332705,
+      "learning_rate": 6.782732218849425e-06,
+      "loss": 0.0191,
+      "step": 1159
+    },
+    {
+      "epoch": 0.40291767974991316,
+      "grad_norm": 0.5862064856816661,
+      "learning_rate": 6.777474788310586e-06,
+      "loss": 0.0278,
+      "step": 1160
+    },
+    {
+      "epoch": 0.4032650225772838,
+      "grad_norm": 0.5792277534533721,
+      "learning_rate": 6.772215107306448e-06,
+      "loss": 0.0244,
+      "step": 1161
+    },
+    {
+      "epoch": 0.4036123654046544,
+      "grad_norm": 0.7441381287676478,
+      "learning_rate": 6.766953182496303e-06,
+      "loss": 0.0164,
+      "step": 1162
+    },
+    {
+      "epoch": 0.403959708232025,
+      "grad_norm": 0.874933616008428,
+      "learning_rate": 6.761689020542288e-06,
+      "loss": 0.024,
+      "step": 1163
+    },
+    {
+      "epoch": 0.40430705105939563,
+      "grad_norm": 0.7249602510280612,
+      "learning_rate": 6.756422628109374e-06,
+      "loss": 0.0291,
+      "step": 1164
+    },
+    {
+      "epoch": 0.4046543938867662,
+      "grad_norm": 0.38558307647762535,
+      "learning_rate": 6.751154011865352e-06,
+      "loss": 0.0193,
+      "step": 1165
+    },
+    {
+      "epoch": 0.40500173671413686,
+      "grad_norm": 0.3569257075132759,
+      "learning_rate": 6.74588317848083e-06,
+      "loss": 0.0224,
+      "step": 1166
+    },
+    {
+      "epoch": 0.40534907954150745,
+      "grad_norm": 0.7027161271352107,
+      "learning_rate": 6.740610134629224e-06,
+      "loss": 0.0236,
+      "step": 1167
+    },
+    {
+      "epoch": 0.4056964223688781,
+      "grad_norm": 0.5055672476223064,
+      "learning_rate": 6.735334886986749e-06,
+      "loss": 0.0169,
+      "step": 1168
+    },
+    {
+      "epoch": 0.4060437651962487,
+      "grad_norm": 1.1794513716622257,
+      "learning_rate": 6.730057442232407e-06,
+      "loss": 0.0156,
+      "step": 1169
+    },
+    {
+      "epoch": 0.40639110802361933,
+      "grad_norm": 1.210535705797956,
+      "learning_rate": 6.724777807047985e-06,
+      "loss": 0.0195,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4067384508509899,
+      "grad_norm": 0.48600002401232995,
+      "learning_rate": 6.719495988118043e-06,
+      "loss": 0.0256,
+      "step": 1171
+    },
+    {
+      "epoch": 0.40708579367836056,
+      "grad_norm": 0.30168099824930594,
+      "learning_rate": 6.714211992129906e-06,
+      "loss": 0.0151,
+      "step": 1172
+    },
+    {
+      "epoch": 0.40743313650573115,
+      "grad_norm": 0.41890959033819075,
+      "learning_rate": 6.708925825773653e-06,
+      "loss": 0.0214,
+      "step": 1173
+    },
+    {
+      "epoch": 0.4077804793331018,
+      "grad_norm": 0.4456883093941609,
+      "learning_rate": 6.7036374957421125e-06,
+      "loss": 0.0164,
+      "step": 1174
+    },
+    {
+      "epoch": 0.4081278221604724,
+      "grad_norm": 0.8050022922465697,
+      "learning_rate": 6.698347008730854e-06,
+      "loss": 0.0177,
+      "step": 1175
+    },
+    {
+      "epoch": 0.408475164987843,
+      "grad_norm": 0.7402939338287975,
+      "learning_rate": 6.6930543714381745e-06,
+      "loss": 0.0217,
+      "step": 1176
+    },
+    {
+      "epoch": 0.4088225078152136,
+      "grad_norm": 0.7580203605323533,
+      "learning_rate": 6.687759590565097e-06,
+      "loss": 0.0217,
+      "step": 1177
+    },
+    {
+      "epoch": 0.4091698506425842,
+      "grad_norm": 0.5276697193580777,
+      "learning_rate": 6.6824626728153565e-06,
+      "loss": 0.0177,
+      "step": 1178
+    },
+    {
+      "epoch": 0.40951719346995485,
+      "grad_norm": 1.3525910498294889,
+      "learning_rate": 6.677163624895393e-06,
+      "loss": 0.0262,
+      "step": 1179
+    },
+    {
+      "epoch": 0.40986453629732544,
+      "grad_norm": 0.7784159092511207,
+      "learning_rate": 6.671862453514346e-06,
+      "loss": 0.0224,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4102118791246961,
+      "grad_norm": 0.5263360435847655,
+      "learning_rate": 6.666559165384041e-06,
+      "loss": 0.018,
+      "step": 1181
+    },
+    {
+      "epoch": 0.4105592219520667,
+      "grad_norm": 0.9366706697301209,
+      "learning_rate": 6.661253767218982e-06,
+      "loss": 0.0232,
+      "step": 1182
+    },
+    {
+      "epoch": 0.4109065647794373,
+      "grad_norm": 0.38867181577970566,
+      "learning_rate": 6.6559462657363525e-06,
+      "loss": 0.0182,
+      "step": 1183
+    },
+    {
+      "epoch": 0.4112539076068079,
+      "grad_norm": 1.6435992793264294,
+      "learning_rate": 6.6506366676559885e-06,
+      "loss": 0.0253,
+      "step": 1184
+    },
+    {
+      "epoch": 0.41160125043417856,
+      "grad_norm": 1.0752706014174525,
+      "learning_rate": 6.6453249797003885e-06,
+      "loss": 0.0266,
+      "step": 1185
+    },
+    {
+      "epoch": 0.41194859326154915,
+      "grad_norm": 0.8205362924547063,
+      "learning_rate": 6.640011208594691e-06,
+      "loss": 0.0228,
+      "step": 1186
+    },
+    {
+      "epoch": 0.4122959360889198,
+      "grad_norm": 0.7528049121740515,
+      "learning_rate": 6.634695361066679e-06,
+      "loss": 0.0205,
+      "step": 1187
+    },
+    {
+      "epoch": 0.4126432789162904,
+      "grad_norm": 0.5476931400142733,
+      "learning_rate": 6.629377443846756e-06,
+      "loss": 0.0256,
+      "step": 1188
+    },
+    {
+      "epoch": 0.41299062174366097,
+      "grad_norm": 1.964533528727283,
+      "learning_rate": 6.624057463667954e-06,
+      "loss": 0.0249,
+      "step": 1189
+    },
+    {
+      "epoch": 0.4133379645710316,
+      "grad_norm": 0.9431680832361578,
+      "learning_rate": 6.618735427265912e-06,
+      "loss": 0.0127,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4136853073984022,
+      "grad_norm": 0.31081854506575984,
+      "learning_rate": 6.613411341378872e-06,
+      "loss": 0.0168,
+      "step": 1191
+    },
+    {
+      "epoch": 0.41403265022577285,
+      "grad_norm": 0.7037510968761942,
+      "learning_rate": 6.608085212747676e-06,
+      "loss": 0.0158,
+      "step": 1192
+    },
+    {
+      "epoch": 0.41437999305314344,
+      "grad_norm": 0.7279786999120493,
+      "learning_rate": 6.602757048115745e-06,
+      "loss": 0.0208,
+      "step": 1193
+    },
+    {
+      "epoch": 0.4147273358805141,
+      "grad_norm": 1.5397066182596297,
+      "learning_rate": 6.597426854229085e-06,
+      "loss": 0.0178,
+      "step": 1194
+    },
+    {
+      "epoch": 0.41507467870788467,
+      "grad_norm": 0.8672894314292771,
+      "learning_rate": 6.592094637836266e-06,
+      "loss": 0.0207,
+      "step": 1195
+    },
+    {
+      "epoch": 0.4154220215352553,
+      "grad_norm": 0.5979467691977439,
+      "learning_rate": 6.586760405688421e-06,
+      "loss": 0.0184,
+      "step": 1196
+    },
+    {
+      "epoch": 0.4157693643626259,
+      "grad_norm": 1.304178348576468,
+      "learning_rate": 6.581424164539235e-06,
+      "loss": 0.0222,
+      "step": 1197
+    },
+    {
+      "epoch": 0.41611670718999655,
+      "grad_norm": 0.744170926834296,
+      "learning_rate": 6.5760859211449355e-06,
+      "loss": 0.0186,
+      "step": 1198
+    },
+    {
+      "epoch": 0.41646405001736714,
+      "grad_norm": 0.43247222642972794,
+      "learning_rate": 6.570745682264288e-06,
+      "loss": 0.0241,
+      "step": 1199
+    },
+    {
+      "epoch": 0.4168113928447378,
+      "grad_norm": 0.7241231215446736,
+      "learning_rate": 6.565403454658579e-06,
+      "loss": 0.0115,
+      "step": 1200
+    },
+    {
+      "epoch": 0.41715873567210837,
+      "grad_norm": 1.5981944791092622,
+      "learning_rate": 6.560059245091619e-06,
+      "loss": 0.0215,
+      "step": 1201
+    },
+    {
+      "epoch": 0.41750607849947896,
+      "grad_norm": 0.6073741226445987,
+      "learning_rate": 6.554713060329725e-06,
+      "loss": 0.0159,
+      "step": 1202
+    },
+    {
+      "epoch": 0.4178534213268496,
+      "grad_norm": 0.35103622430566045,
+      "learning_rate": 6.549364907141713e-06,
+      "loss": 0.0165,
+      "step": 1203
+    },
+    {
+      "epoch": 0.4182007641542202,
+      "grad_norm": 0.3060822356877603,
+      "learning_rate": 6.544014792298896e-06,
+      "loss": 0.0136,
+      "step": 1204
+    },
+    {
+      "epoch": 0.41854810698159084,
+      "grad_norm": 1.2509560719304376,
+      "learning_rate": 6.538662722575067e-06,
+      "loss": 0.0232,
+      "step": 1205
+    },
+    {
+      "epoch": 0.41889544980896143,
+      "grad_norm": 1.1252212083506805,
+      "learning_rate": 6.533308704746492e-06,
+      "loss": 0.0261,
+      "step": 1206
+    },
+    {
+      "epoch": 0.4192427926363321,
+      "grad_norm": 1.0051855699617804,
+      "learning_rate": 6.527952745591911e-06,
+      "loss": 0.0176,
+      "step": 1207
+    },
+    {
+      "epoch": 0.41959013546370266,
+      "grad_norm": 0.4225505750014969,
+      "learning_rate": 6.522594851892513e-06,
+      "loss": 0.0157,
+      "step": 1208
+    },
+    {
+      "epoch": 0.4199374782910733,
+      "grad_norm": 0.675487063845996,
+      "learning_rate": 6.5172350304319456e-06,
+      "loss": 0.0187,
+      "step": 1209
+    },
+    {
+      "epoch": 0.4202848211184439,
+      "grad_norm": 0.5556340570184015,
+      "learning_rate": 6.5118732879962866e-06,
+      "loss": 0.021,
+      "step": 1210
+    },
+    {
+      "epoch": 0.42063216394581454,
+      "grad_norm": 1.1461752497953057,
+      "learning_rate": 6.506509631374056e-06,
+      "loss": 0.024,
+      "step": 1211
+    },
+    {
+      "epoch": 0.42097950677318513,
+      "grad_norm": 0.8587515989365709,
+      "learning_rate": 6.501144067356191e-06,
+      "loss": 0.0171,
+      "step": 1212
+    },
+    {
+      "epoch": 0.4213268496005558,
+      "grad_norm": 0.6554268552282353,
+      "learning_rate": 6.4957766027360455e-06,
+      "loss": 0.0162,
+      "step": 1213
+    },
+    {
+      "epoch": 0.42167419242792636,
+      "grad_norm": 0.9717352789041409,
+      "learning_rate": 6.490407244309382e-06,
+      "loss": 0.0177,
+      "step": 1214
+    },
+    {
+      "epoch": 0.42202153525529695,
+      "grad_norm": 0.4666185404754438,
+      "learning_rate": 6.485035998874356e-06,
+      "loss": 0.0179,
+      "step": 1215
+    },
+    {
+      "epoch": 0.4223688780826676,
+      "grad_norm": 0.6687573112666673,
+      "learning_rate": 6.479662873231518e-06,
+      "loss": 0.0278,
+      "step": 1216
+    },
+    {
+      "epoch": 0.4227162209100382,
+      "grad_norm": 0.39201919773447125,
+      "learning_rate": 6.4742878741837924e-06,
+      "loss": 0.021,
+      "step": 1217
+    },
+    {
+      "epoch": 0.42306356373740883,
+      "grad_norm": 1.9376472430652754,
+      "learning_rate": 6.468911008536483e-06,
+      "loss": 0.0315,
+      "step": 1218
+    },
+    {
+      "epoch": 0.4234109065647794,
+      "grad_norm": 0.5349195261746512,
+      "learning_rate": 6.4635322830972465e-06,
+      "loss": 0.022,
+      "step": 1219
+    },
+    {
+      "epoch": 0.42375824939215007,
+      "grad_norm": 0.5994260198762377,
+      "learning_rate": 6.458151704676108e-06,
+      "loss": 0.0249,
+      "step": 1220
+    },
+    {
+      "epoch": 0.42410559221952066,
+      "grad_norm": 0.45829461719718667,
+      "learning_rate": 6.452769280085427e-06,
+      "loss": 0.0172,
+      "step": 1221
+    },
+    {
+      "epoch": 0.4244529350468913,
+      "grad_norm": 0.4015948567021689,
+      "learning_rate": 6.447385016139906e-06,
+      "loss": 0.0159,
+      "step": 1222
+    },
+    {
+      "epoch": 0.4248002778742619,
+      "grad_norm": 0.41359006368663803,
+      "learning_rate": 6.441998919656575e-06,
+      "loss": 0.0217,
+      "step": 1223
+    },
+    {
+      "epoch": 0.42514762070163253,
+      "grad_norm": 0.784104999649454,
+      "learning_rate": 6.436610997454785e-06,
+      "loss": 0.0153,
+      "step": 1224
+    },
+    {
+      "epoch": 0.4254949635290031,
+      "grad_norm": 0.4486969339475269,
+      "learning_rate": 6.431221256356197e-06,
+      "loss": 0.0191,
+      "step": 1225
+    },
+    {
+      "epoch": 0.42584230635637377,
+      "grad_norm": 0.807568173718478,
+      "learning_rate": 6.425829703184776e-06,
+      "loss": 0.0191,
+      "step": 1226
+    },
+    {
+      "epoch": 0.42618964918374436,
+      "grad_norm": 1.214899649576311,
+      "learning_rate": 6.420436344766781e-06,
+      "loss": 0.0242,
+      "step": 1227
+    },
+    {
+      "epoch": 0.42653699201111495,
+      "grad_norm": 0.41836170664521255,
+      "learning_rate": 6.415041187930757e-06,
+      "loss": 0.0195,
+      "step": 1228
+    },
+    {
+      "epoch": 0.4268843348384856,
+      "grad_norm": 0.8701313451372041,
+      "learning_rate": 6.409644239507524e-06,
+      "loss": 0.0199,
+      "step": 1229
+    },
+    {
+      "epoch": 0.4272316776658562,
+      "grad_norm": 0.46742950813151507,
+      "learning_rate": 6.404245506330175e-06,
+      "loss": 0.013,
+      "step": 1230
+    },
+    {
+      "epoch": 0.4275790204932268,
+      "grad_norm": 0.5597197002970087,
+      "learning_rate": 6.398844995234057e-06,
+      "loss": 0.0246,
+      "step": 1231
+    },
+    {
+      "epoch": 0.4279263633205974,
+      "grad_norm": 1.3414468877460513,
+      "learning_rate": 6.393442713056772e-06,
+      "loss": 0.021,
+      "step": 1232
+    },
+    {
+      "epoch": 0.42827370614796806,
+      "grad_norm": 0.6723941756786661,
+      "learning_rate": 6.388038666638163e-06,
+      "loss": 0.0228,
+      "step": 1233
+    },
+    {
+      "epoch": 0.42862104897533865,
+      "grad_norm": 0.5282980798460091,
+      "learning_rate": 6.382632862820306e-06,
+      "loss": 0.019,
+      "step": 1234
+    },
+    {
+      "epoch": 0.4289683918027093,
+      "grad_norm": 0.8434630536972346,
+      "learning_rate": 6.377225308447503e-06,
+      "loss": 0.0279,
+      "step": 1235
+    },
+    {
+      "epoch": 0.4293157346300799,
+      "grad_norm": 1.2873804248352099,
+      "learning_rate": 6.371816010366274e-06,
+      "loss": 0.0242,
+      "step": 1236
+    },
+    {
+      "epoch": 0.4296630774574505,
+      "grad_norm": 1.0288152920759024,
+      "learning_rate": 6.366404975425342e-06,
+      "loss": 0.0136,
+      "step": 1237
+    },
+    {
+      "epoch": 0.4300104202848211,
+      "grad_norm": 0.9931125336676734,
+      "learning_rate": 6.360992210475635e-06,
+      "loss": 0.0238,
+      "step": 1238
+    },
+    {
+      "epoch": 0.43035776311219176,
+      "grad_norm": 1.136404288127991,
+      "learning_rate": 6.355577722370264e-06,
+      "loss": 0.0171,
+      "step": 1239
+    },
+    {
+      "epoch": 0.43070510593956235,
+      "grad_norm": 0.6828220416786258,
+      "learning_rate": 6.3501615179645315e-06,
+      "loss": 0.0195,
+      "step": 1240
+    },
+    {
+      "epoch": 0.43105244876693294,
+      "grad_norm": 1.3364697049657948,
+      "learning_rate": 6.344743604115903e-06,
+      "loss": 0.0293,
+      "step": 1241
+    },
+    {
+      "epoch": 0.4313997915943036,
+      "grad_norm": 1.2606216983253666,
+      "learning_rate": 6.339323987684015e-06,
+      "loss": 0.0197,
+      "step": 1242
+    },
+    {
+      "epoch": 0.4317471344216742,
+      "grad_norm": 0.5059234102433385,
+      "learning_rate": 6.333902675530657e-06,
+      "loss": 0.0216,
+      "step": 1243
+    },
+    {
+      "epoch": 0.4320944772490448,
+      "grad_norm": 0.6807195324342239,
+      "learning_rate": 6.328479674519766e-06,
+      "loss": 0.0225,
+      "step": 1244
+    },
+    {
+      "epoch": 0.4324418200764154,
+      "grad_norm": 0.5504876720573679,
+      "learning_rate": 6.323054991517416e-06,
+      "loss": 0.0128,
+      "step": 1245
+    },
+    {
+      "epoch": 0.43278916290378605,
+      "grad_norm": 0.8748832096236718,
+      "learning_rate": 6.317628633391816e-06,
+      "loss": 0.0208,
+      "step": 1246
+    },
+    {
+      "epoch": 0.43313650573115664,
+      "grad_norm": 1.3807423200829632,
+      "learning_rate": 6.312200607013287e-06,
+      "loss": 0.0192,
+      "step": 1247
+    },
+    {
+      "epoch": 0.4334838485585273,
+      "grad_norm": 0.4591330354830137,
+      "learning_rate": 6.306770919254268e-06,
+      "loss": 0.0182,
+      "step": 1248
+    },
+    {
+      "epoch": 0.4338311913858979,
+      "grad_norm": 1.1115703829370096,
+      "learning_rate": 6.301339576989301e-06,
+      "loss": 0.0257,
+      "step": 1249
+    },
+    {
+      "epoch": 0.4341785342132685,
+      "grad_norm": 0.9601555993164028,
+      "learning_rate": 6.295906587095023e-06,
+      "loss": 0.0293,
+      "step": 1250
+    },
+    {
+      "epoch": 0.4345258770406391,
+      "grad_norm": 0.3602112221728658,
+      "learning_rate": 6.2904719564501545e-06,
+      "loss": 0.0209,
+      "step": 1251
+    },
+    {
+      "epoch": 0.43487321986800975,
+      "grad_norm": 0.65050746016075,
+      "learning_rate": 6.285035691935495e-06,
+      "loss": 0.0201,
+      "step": 1252
+    },
+    {
+      "epoch": 0.43522056269538034,
+      "grad_norm": 0.43904022220763456,
+      "learning_rate": 6.279597800433915e-06,
+      "loss": 0.0244,
+      "step": 1253
+    },
+    {
+      "epoch": 0.43556790552275093,
+      "grad_norm": 1.0501132280980383,
+      "learning_rate": 6.274158288830339e-06,
+      "loss": 0.0152,
+      "step": 1254
+    },
+    {
+      "epoch": 0.4359152483501216,
+      "grad_norm": 1.368313597379767,
+      "learning_rate": 6.268717164011751e-06,
+      "loss": 0.0299,
+      "step": 1255
+    },
+    {
+      "epoch": 0.43626259117749216,
+      "grad_norm": 0.29948345258209563,
+      "learning_rate": 6.263274432867168e-06,
+      "loss": 0.0175,
+      "step": 1256
+    },
+    {
+      "epoch": 0.4366099340048628,
+      "grad_norm": 0.5520058587984689,
+      "learning_rate": 6.257830102287649e-06,
+      "loss": 0.0131,
+      "step": 1257
+    },
+    {
+      "epoch": 0.4369572768322334,
+      "grad_norm": 0.703456209843452,
+      "learning_rate": 6.252384179166272e-06,
+      "loss": 0.026,
+      "step": 1258
+    },
+    {
+      "epoch": 0.43730461965960404,
+      "grad_norm": 0.31650775648430113,
+      "learning_rate": 6.246936670398136e-06,
+      "loss": 0.0119,
+      "step": 1259
+    },
+    {
+      "epoch": 0.43765196248697463,
+      "grad_norm": 0.6575613262249127,
+      "learning_rate": 6.2414875828803446e-06,
+      "loss": 0.018,
+      "step": 1260
+    },
+    {
+      "epoch": 0.4379993053143453,
+      "grad_norm": 0.7650908913989808,
+      "learning_rate": 6.236036923512002e-06,
+      "loss": 0.0193,
+      "step": 1261
+    },
+    {
+      "epoch": 0.43834664814171587,
+      "grad_norm": 0.6645282671067363,
+      "learning_rate": 6.230584699194201e-06,
+      "loss": 0.0191,
+      "step": 1262
+    },
+    {
+      "epoch": 0.4386939909690865,
+      "grad_norm": 0.89119449808857,
+      "learning_rate": 6.225130916830017e-06,
+      "loss": 0.0165,
+      "step": 1263
+    },
+    {
+      "epoch": 0.4390413337964571,
+      "grad_norm": 1.4336588300538304,
+      "learning_rate": 6.2196755833244975e-06,
+      "loss": 0.0172,
+      "step": 1264
+    },
+    {
+      "epoch": 0.43938867662382775,
+      "grad_norm": 0.5141694363351154,
+      "learning_rate": 6.214218705584653e-06,
+      "loss": 0.017,
+      "step": 1265
+    },
+    {
+      "epoch": 0.43973601945119833,
+      "grad_norm": 0.7024489932275159,
+      "learning_rate": 6.208760290519451e-06,
+      "loss": 0.0266,
+      "step": 1266
+    },
+    {
+      "epoch": 0.4400833622785689,
+      "grad_norm": 0.5137267630010462,
+      "learning_rate": 6.203300345039804e-06,
+      "loss": 0.0208,
+      "step": 1267
+    },
+    {
+      "epoch": 0.44043070510593957,
+      "grad_norm": 0.9545655799932716,
+      "learning_rate": 6.197838876058564e-06,
+      "loss": 0.0212,
+      "step": 1268
+    },
+    {
+      "epoch": 0.44077804793331016,
+      "grad_norm": 1.274602132775298,
+      "learning_rate": 6.19237589049051e-06,
+      "loss": 0.02,
+      "step": 1269
+    },
+    {
+      "epoch": 0.4411253907606808,
+      "grad_norm": 1.1485868760171818,
+      "learning_rate": 6.186911395252342e-06,
+      "loss": 0.0325,
+      "step": 1270
+    },
+    {
+      "epoch": 0.4414727335880514,
+      "grad_norm": 1.0719692070927125,
+      "learning_rate": 6.181445397262671e-06,
+      "loss": 0.0246,
+      "step": 1271
+    },
+    {
+      "epoch": 0.44182007641542204,
+      "grad_norm": 0.5336479484346233,
+      "learning_rate": 6.175977903442008e-06,
+      "loss": 0.0154,
+      "step": 1272
+    },
+    {
+      "epoch": 0.4421674192427926,
+      "grad_norm": 0.6172067020914168,
+      "learning_rate": 6.170508920712765e-06,
+      "loss": 0.0188,
+      "step": 1273
+    },
+    {
+      "epoch": 0.44251476207016327,
+      "grad_norm": 0.9412324301612399,
+      "learning_rate": 6.165038455999233e-06,
+      "loss": 0.0216,
+      "step": 1274
+    },
+    {
+      "epoch": 0.44286210489753386,
+      "grad_norm": 0.5832155616020471,
+      "learning_rate": 6.159566516227582e-06,
+      "loss": 0.0214,
+      "step": 1275
+    },
+    {
+      "epoch": 0.4432094477249045,
+      "grad_norm": 0.5486296573893363,
+      "learning_rate": 6.154093108325846e-06,
+      "loss": 0.0163,
+      "step": 1276
+    },
+    {
+      "epoch": 0.4435567905522751,
+      "grad_norm": 0.6013895518841912,
+      "learning_rate": 6.148618239223924e-06,
+      "loss": 0.0146,
+      "step": 1277
+    },
+    {
+      "epoch": 0.44390413337964574,
+      "grad_norm": 0.886747002104457,
+      "learning_rate": 6.143141915853558e-06,
+      "loss": 0.0141,
+      "step": 1278
+    },
+    {
+      "epoch": 0.4442514762070163,
+      "grad_norm": 0.7983414083814292,
+      "learning_rate": 6.137664145148339e-06,
+      "loss": 0.0229,
+      "step": 1279
+    },
+    {
+      "epoch": 0.4445988190343869,
+      "grad_norm": 0.4982201099217589,
+      "learning_rate": 6.1321849340436824e-06,
+      "loss": 0.0226,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44494616186175756,
+      "grad_norm": 0.6104658236755754,
+      "learning_rate": 6.126704289476834e-06,
+      "loss": 0.0176,
+      "step": 1281
+    },
+    {
+      "epoch": 0.44529350468912815,
+      "grad_norm": 0.4850325700357511,
+      "learning_rate": 6.121222218386848e-06,
+      "loss": 0.0275,
+      "step": 1282
+    },
+    {
+      "epoch": 0.4456408475164988,
+      "grad_norm": 1.1453236673507199,
+      "learning_rate": 6.115738727714593e-06,
+      "loss": 0.0157,
+      "step": 1283
+    },
+    {
+      "epoch": 0.4459881903438694,
+      "grad_norm": 1.1501958397477723,
+      "learning_rate": 6.110253824402728e-06,
+      "loss": 0.0233,
+      "step": 1284
+    },
+    {
+      "epoch": 0.44633553317124003,
+      "grad_norm": 1.2142627028854323,
+      "learning_rate": 6.104767515395702e-06,
+      "loss": 0.0217,
+      "step": 1285
+    },
+    {
+      "epoch": 0.4466828759986106,
+      "grad_norm": 0.8286577032771891,
+      "learning_rate": 6.0992798076397465e-06,
+      "loss": 0.0281,
+      "step": 1286
+    },
+    {
+      "epoch": 0.44703021882598126,
+      "grad_norm": 1.1668550772372472,
+      "learning_rate": 6.093790708082861e-06,
+      "loss": 0.0122,
+      "step": 1287
+    },
+    {
+      "epoch": 0.44737756165335185,
+      "grad_norm": 0.8595111659192216,
+      "learning_rate": 6.088300223674808e-06,
+      "loss": 0.013,
+      "step": 1288
+    },
+    {
+      "epoch": 0.4477249044807225,
+      "grad_norm": 1.202956918343463,
+      "learning_rate": 6.0828083613671055e-06,
+      "loss": 0.0155,
+      "step": 1289
+    },
+    {
+      "epoch": 0.4480722473080931,
+      "grad_norm": 1.144767269040697,
+      "learning_rate": 6.077315128113011e-06,
+      "loss": 0.0234,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44841959013546373,
+      "grad_norm": 0.3829734675474034,
+      "learning_rate": 6.071820530867524e-06,
+      "loss": 0.017,
+      "step": 1291
+    },
+    {
+      "epoch": 0.4487669329628343,
+      "grad_norm": 0.705925698459192,
+      "learning_rate": 6.066324576587367e-06,
+      "loss": 0.0207,
+      "step": 1292
+    },
+    {
+      "epoch": 0.4491142757902049,
+      "grad_norm": 0.9927898285052602,
+      "learning_rate": 6.06082727223098e-06,
+      "loss": 0.0227,
+      "step": 1293
+    },
+    {
+      "epoch": 0.44946161861757555,
+      "grad_norm": 0.5058096386894589,
+      "learning_rate": 6.055328624758515e-06,
+      "loss": 0.025,
+      "step": 1294
+    },
+    {
+      "epoch": 0.44980896144494614,
+      "grad_norm": 0.9125209454948322,
+      "learning_rate": 6.0498286411318255e-06,
+      "loss": 0.0224,
+      "step": 1295
+    },
+    {
+      "epoch": 0.4501563042723168,
+      "grad_norm": 1.1242298901659808,
+      "learning_rate": 6.04432732831445e-06,
+      "loss": 0.0191,
+      "step": 1296
+    },
+    {
+      "epoch": 0.4505036470996874,
+      "grad_norm": 0.8939518763202791,
+      "learning_rate": 6.038824693271619e-06,
+      "loss": 0.0249,
+      "step": 1297
+    },
+    {
+      "epoch": 0.450850989927058,
+      "grad_norm": 1.0950080836356835,
+      "learning_rate": 6.033320742970229e-06,
+      "loss": 0.0248,
+      "step": 1298
+    },
+    {
+      "epoch": 0.4511983327544286,
+      "grad_norm": 0.8937794294216942,
+      "learning_rate": 6.027815484378848e-06,
+      "loss": 0.0247,
+      "step": 1299
+    },
+    {
+      "epoch": 0.45154567558179926,
+      "grad_norm": 1.211557408088087,
+      "learning_rate": 6.0223089244676965e-06,
+      "loss": 0.0257,
+      "step": 1300
+    },
+    {
+      "epoch": 0.45189301840916984,
+      "grad_norm": 0.7913795101404235,
+      "learning_rate": 6.016801070208644e-06,
+      "loss": 0.0201,
+      "step": 1301
+    },
+    {
+      "epoch": 0.4522403612365405,
+      "grad_norm": 1.022217982353667,
+      "learning_rate": 6.011291928575199e-06,
+      "loss": 0.0177,
+      "step": 1302
+    },
+    {
+      "epoch": 0.4525877040639111,
+      "grad_norm": 0.5643877352773056,
+      "learning_rate": 6.005781506542498e-06,
+      "loss": 0.018,
+      "step": 1303
+    },
+    {
+      "epoch": 0.45293504689128167,
+      "grad_norm": 0.46969855875618594,
+      "learning_rate": 6.000269811087304e-06,
+      "loss": 0.0164,
+      "step": 1304
+    },
+    {
+      "epoch": 0.4532823897186523,
+      "grad_norm": 0.64817857112924,
+      "learning_rate": 5.994756849187984e-06,
+      "loss": 0.0224,
+      "step": 1305
+    },
+    {
+      "epoch": 0.4536297325460229,
+      "grad_norm": 1.0142085570957111,
+      "learning_rate": 5.989242627824516e-06,
+      "loss": 0.0293,
+      "step": 1306
+    },
+    {
+      "epoch": 0.45397707537339355,
+      "grad_norm": 0.44811883050529416,
+      "learning_rate": 5.983727153978467e-06,
+      "loss": 0.0188,
+      "step": 1307
+    },
+    {
+      "epoch": 0.45432441820076414,
+      "grad_norm": 0.8638051888219721,
+      "learning_rate": 5.978210434632996e-06,
+      "loss": 0.0233,
+      "step": 1308
+    },
+    {
+      "epoch": 0.4546717610281348,
+      "grad_norm": 0.7100938005096301,
+      "learning_rate": 5.97269247677283e-06,
+      "loss": 0.0197,
+      "step": 1309
+    },
+    {
+      "epoch": 0.45501910385550537,
+      "grad_norm": 0.5543422204245909,
+      "learning_rate": 5.967173287384275e-06,
+      "loss": 0.0163,
+      "step": 1310
+    },
+    {
+      "epoch": 0.455366446682876,
+      "grad_norm": 0.8347623418009196,
+      "learning_rate": 5.961652873455186e-06,
+      "loss": 0.0142,
+      "step": 1311
+    },
+    {
+      "epoch": 0.4557137895102466,
+      "grad_norm": 0.2539498893920443,
+      "learning_rate": 5.956131241974976e-06,
+      "loss": 0.0132,
+      "step": 1312
+    },
+    {
+      "epoch": 0.45606113233761725,
+      "grad_norm": 1.1630972005065157,
+      "learning_rate": 5.950608399934594e-06,
+      "loss": 0.0242,
+      "step": 1313
+    },
+    {
+      "epoch": 0.45640847516498784,
+      "grad_norm": 1.0539497447728725,
+      "learning_rate": 5.945084354326527e-06,
+      "loss": 0.0179,
+      "step": 1314
+    },
+    {
+      "epoch": 0.4567558179923585,
+      "grad_norm": 0.838017365343182,
+      "learning_rate": 5.939559112144781e-06,
+      "loss": 0.0265,
+      "step": 1315
+    },
+    {
+      "epoch": 0.45710316081972907,
+      "grad_norm": 0.5217217424418361,
+      "learning_rate": 5.93403268038488e-06,
+      "loss": 0.0173,
+      "step": 1316
+    },
+    {
+      "epoch": 0.45745050364709966,
+      "grad_norm": 1.049739897466213,
+      "learning_rate": 5.928505066043852e-06,
+      "loss": 0.0154,
+      "step": 1317
+    },
+    {
+      "epoch": 0.4577978464744703,
+      "grad_norm": 0.514558882908952,
+      "learning_rate": 5.922976276120225e-06,
+      "loss": 0.0236,
+      "step": 1318
+    },
+    {
+      "epoch": 0.4581451893018409,
+      "grad_norm": 0.8103740247236859,
+      "learning_rate": 5.917446317614012e-06,
+      "loss": 0.0115,
+      "step": 1319
+    },
+    {
+      "epoch": 0.45849253212921154,
+      "grad_norm": 1.0239011468521235,
+      "learning_rate": 5.911915197526709e-06,
+      "loss": 0.0194,
+      "step": 1320
+    },
+    {
+      "epoch": 0.4588398749565821,
+      "grad_norm": 0.45055190346246493,
+      "learning_rate": 5.9063829228612805e-06,
+      "loss": 0.0244,
+      "step": 1321
+    },
+    {
+      "epoch": 0.4591872177839528,
+      "grad_norm": 0.4096700783073768,
+      "learning_rate": 5.900849500622153e-06,
+      "loss": 0.0192,
+      "step": 1322
+    },
+    {
+      "epoch": 0.45953456061132336,
+      "grad_norm": 5.0746307124667585,
+      "learning_rate": 5.895314937815206e-06,
+      "loss": 0.0249,
+      "step": 1323
+    },
+    {
+      "epoch": 0.459881903438694,
+      "grad_norm": 0.6287936978756935,
+      "learning_rate": 5.889779241447765e-06,
+      "loss": 0.0192,
+      "step": 1324
+    },
+    {
+      "epoch": 0.4602292462660646,
+      "grad_norm": 0.5423899884873425,
+      "learning_rate": 5.884242418528588e-06,
+      "loss": 0.0219,
+      "step": 1325
+    },
+    {
+      "epoch": 0.46057658909343524,
+      "grad_norm": 0.5268649486665123,
+      "learning_rate": 5.878704476067862e-06,
+      "loss": 0.0166,
+      "step": 1326
+    },
+    {
+      "epoch": 0.46092393192080583,
+      "grad_norm": 0.922315522446586,
+      "learning_rate": 5.873165421077186e-06,
+      "loss": 0.0244,
+      "step": 1327
+    },
+    {
+      "epoch": 0.4612712747481765,
+      "grad_norm": 0.6105099198285099,
+      "learning_rate": 5.867625260569575e-06,
+      "loss": 0.0278,
+      "step": 1328
+    },
+    {
+      "epoch": 0.46161861757554706,
+      "grad_norm": 1.472196223378207,
+      "learning_rate": 5.862084001559438e-06,
+      "loss": 0.0233,
+      "step": 1329
+    },
+    {
+      "epoch": 0.46196596040291765,
+      "grad_norm": 0.6221418880436498,
+      "learning_rate": 5.85654165106258e-06,
+      "loss": 0.0113,
+      "step": 1330
+    },
+    {
+      "epoch": 0.4623133032302883,
+      "grad_norm": 0.6121127340609553,
+      "learning_rate": 5.850998216096181e-06,
+      "loss": 0.0249,
+      "step": 1331
+    },
+    {
+      "epoch": 0.4626606460576589,
+      "grad_norm": 0.7691635062367261,
+      "learning_rate": 5.845453703678801e-06,
+      "loss": 0.0263,
+      "step": 1332
+    },
+    {
+      "epoch": 0.46300798888502953,
+      "grad_norm": 0.4516319265859705,
+      "learning_rate": 5.8399081208303595e-06,
+      "loss": 0.0217,
+      "step": 1333
+    },
+    {
+      "epoch": 0.4633553317124001,
+      "grad_norm": 1.0317569643624573,
+      "learning_rate": 5.834361474572134e-06,
+      "loss": 0.0174,
+      "step": 1334
+    },
+    {
+      "epoch": 0.46370267453977076,
+      "grad_norm": 0.6672608515849126,
+      "learning_rate": 5.828813771926746e-06,
+      "loss": 0.0202,
+      "step": 1335
+    },
+    {
+      "epoch": 0.46405001736714135,
+      "grad_norm": 2.1225607990268083,
+      "learning_rate": 5.823265019918156e-06,
+      "loss": 0.0202,
+      "step": 1336
+    },
+    {
+      "epoch": 0.464397360194512,
+      "grad_norm": 0.5824621371505889,
+      "learning_rate": 5.817715225571654e-06,
+      "loss": 0.0189,
+      "step": 1337
+    },
+    {
+      "epoch": 0.4647447030218826,
+      "grad_norm": 0.5556260631755074,
+      "learning_rate": 5.812164395913848e-06,
+      "loss": 0.02,
+      "step": 1338
+    },
+    {
+      "epoch": 0.46509204584925323,
+      "grad_norm": 0.5236250786169708,
+      "learning_rate": 5.806612537972658e-06,
+      "loss": 0.0295,
+      "step": 1339
+    },
+    {
+      "epoch": 0.4654393886766238,
+      "grad_norm": 0.915218352922461,
+      "learning_rate": 5.801059658777303e-06,
+      "loss": 0.0247,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46578673150399447,
+      "grad_norm": 0.8125906729186393,
+      "learning_rate": 5.7955057653583e-06,
+      "loss": 0.0168,
+      "step": 1341
+    },
+    {
+      "epoch": 0.46613407433136506,
+      "grad_norm": 0.500901462927243,
+      "learning_rate": 5.789950864747446e-06,
+      "loss": 0.0163,
+      "step": 1342
+    },
+    {
+      "epoch": 0.46648141715873565,
+      "grad_norm": 0.8445734370127098,
+      "learning_rate": 5.784394963977815e-06,
+      "loss": 0.0156,
+      "step": 1343
+    },
+    {
+      "epoch": 0.4668287599861063,
+      "grad_norm": 0.8202499613757581,
+      "learning_rate": 5.778838070083747e-06,
+      "loss": 0.0165,
+      "step": 1344
+    },
+    {
+      "epoch": 0.4671761028134769,
+      "grad_norm": 0.9521555125223035,
+      "learning_rate": 5.77328019010084e-06,
+      "loss": 0.0236,
+      "step": 1345
+    },
+    {
+      "epoch": 0.4675234456408475,
+      "grad_norm": 0.3706684952737426,
+      "learning_rate": 5.7677213310659375e-06,
+      "loss": 0.0134,
+      "step": 1346
+    },
+    {
+      "epoch": 0.4678707884682181,
+      "grad_norm": 0.4599731154160651,
+      "learning_rate": 5.762161500017128e-06,
+      "loss": 0.0155,
+      "step": 1347
+    },
+    {
+      "epoch": 0.46821813129558876,
+      "grad_norm": 0.2427917500418682,
+      "learning_rate": 5.756600703993725e-06,
+      "loss": 0.0089,
+      "step": 1348
+    },
+    {
+      "epoch": 0.46856547412295935,
+      "grad_norm": 0.7157300676183229,
+      "learning_rate": 5.751038950036267e-06,
+      "loss": 0.026,
+      "step": 1349
+    },
+    {
+      "epoch": 0.46891281695033,
+      "grad_norm": 0.5005873987026797,
+      "learning_rate": 5.745476245186506e-06,
+      "loss": 0.0213,
+      "step": 1350
+    },
+    {
+      "epoch": 0.4692601597777006,
+      "grad_norm": 0.9673326052882976,
+      "learning_rate": 5.739912596487396e-06,
+      "loss": 0.0269,
+      "step": 1351
+    },
+    {
+      "epoch": 0.4696075026050712,
+      "grad_norm": 0.457111905319674,
+      "learning_rate": 5.7343480109830865e-06,
+      "loss": 0.0177,
+      "step": 1352
+    },
+    {
+      "epoch": 0.4699548454324418,
+      "grad_norm": 0.8619795743501616,
+      "learning_rate": 5.728782495718912e-06,
+      "loss": 0.0156,
+      "step": 1353
+    },
+    {
+      "epoch": 0.47030218825981246,
+      "grad_norm": 0.40333546480267507,
+      "learning_rate": 5.7232160577413866e-06,
+      "loss": 0.0186,
+      "step": 1354
+    },
+    {
+      "epoch": 0.47064953108718305,
+      "grad_norm": 0.5088273149893132,
+      "learning_rate": 5.717648704098191e-06,
+      "loss": 0.0121,
+      "step": 1355
+    },
+    {
+      "epoch": 0.47099687391455364,
+      "grad_norm": 0.5242704677164807,
+      "learning_rate": 5.712080441838167e-06,
+      "loss": 0.0162,
+      "step": 1356
+    },
+    {
+      "epoch": 0.4713442167419243,
+      "grad_norm": 1.2256987250615423,
+      "learning_rate": 5.706511278011303e-06,
+      "loss": 0.0207,
+      "step": 1357
+    },
+    {
+      "epoch": 0.47169155956929487,
+      "grad_norm": 0.9464524842305495,
+      "learning_rate": 5.700941219668733e-06,
+      "loss": 0.0191,
+      "step": 1358
+    },
+    {
+      "epoch": 0.4720389023966655,
+      "grad_norm": 0.8126675490446509,
+      "learning_rate": 5.6953702738627215e-06,
+      "loss": 0.0195,
+      "step": 1359
+    },
+    {
+      "epoch": 0.4723862452240361,
+      "grad_norm": 1.0495606290514148,
+      "learning_rate": 5.689798447646657e-06,
+      "loss": 0.0216,
+      "step": 1360
+    },
+    {
+      "epoch": 0.47273358805140675,
+      "grad_norm": 0.5981699030064619,
+      "learning_rate": 5.684225748075044e-06,
+      "loss": 0.0143,
+      "step": 1361
+    },
+    {
+      "epoch": 0.47308093087877734,
+      "grad_norm": 0.478440133230351,
+      "learning_rate": 5.678652182203489e-06,
+      "loss": 0.0215,
+      "step": 1362
+    },
+    {
+      "epoch": 0.473428273706148,
+      "grad_norm": 0.9677500646745865,
+      "learning_rate": 5.6730777570887e-06,
+      "loss": 0.0195,
+      "step": 1363
+    },
+    {
+      "epoch": 0.4737756165335186,
+      "grad_norm": 0.7312470814458618,
+      "learning_rate": 5.667502479788467e-06,
+      "loss": 0.0109,
+      "step": 1364
+    },
+    {
+      "epoch": 0.4741229593608892,
+      "grad_norm": 0.36719507217506836,
+      "learning_rate": 5.6619263573616676e-06,
+      "loss": 0.0152,
+      "step": 1365
+    },
+    {
+      "epoch": 0.4744703021882598,
+      "grad_norm": 0.503824691086867,
+      "learning_rate": 5.6563493968682405e-06,
+      "loss": 0.021,
+      "step": 1366
+    },
+    {
+      "epoch": 0.47481764501563045,
+      "grad_norm": 1.1250583221317176,
+      "learning_rate": 5.6507716053691916e-06,
+      "loss": 0.0259,
+      "step": 1367
+    },
+    {
+      "epoch": 0.47516498784300104,
+      "grad_norm": 0.574390097728659,
+      "learning_rate": 5.645192989926577e-06,
+      "loss": 0.0237,
+      "step": 1368
+    },
+    {
+      "epoch": 0.47551233067037163,
+      "grad_norm": 1.5251526377041664,
+      "learning_rate": 5.639613557603494e-06,
+      "loss": 0.0281,
+      "step": 1369
+    },
+    {
+      "epoch": 0.4758596734977423,
+      "grad_norm": 0.6785324566307315,
+      "learning_rate": 5.634033315464076e-06,
+      "loss": 0.0124,
+      "step": 1370
+    },
+    {
+      "epoch": 0.47620701632511286,
+      "grad_norm": 0.5227461607632101,
+      "learning_rate": 5.628452270573483e-06,
+      "loss": 0.0177,
+      "step": 1371
+    },
+    {
+      "epoch": 0.4765543591524835,
+      "grad_norm": 1.1646192631978813,
+      "learning_rate": 5.6228704299978905e-06,
+      "loss": 0.0241,
+      "step": 1372
+    },
+    {
+      "epoch": 0.4769017019798541,
+      "grad_norm": 0.5872534337661728,
+      "learning_rate": 5.617287800804478e-06,
+      "loss": 0.0218,
+      "step": 1373
+    },
+    {
+      "epoch": 0.47724904480722474,
+      "grad_norm": 0.5225710058655455,
+      "learning_rate": 5.61170439006143e-06,
+      "loss": 0.0223,
+      "step": 1374
+    },
+    {
+      "epoch": 0.47759638763459533,
+      "grad_norm": 1.1334161288347557,
+      "learning_rate": 5.6061202048379125e-06,
+      "loss": 0.0254,
+      "step": 1375
+    },
+    {
+      "epoch": 0.477943730461966,
+      "grad_norm": 1.0865327629480914,
+      "learning_rate": 5.600535252204081e-06,
+      "loss": 0.0237,
+      "step": 1376
+    },
+    {
+      "epoch": 0.47829107328933657,
+      "grad_norm": 1.279397014560853,
+      "learning_rate": 5.5949495392310535e-06,
+      "loss": 0.0286,
+      "step": 1377
+    },
+    {
+      "epoch": 0.4786384161167072,
+      "grad_norm": 0.32759908340107924,
+      "learning_rate": 5.589363072990921e-06,
+      "loss": 0.0149,
+      "step": 1378
+    },
+    {
+      "epoch": 0.4789857589440778,
+      "grad_norm": 0.5810913478185202,
+      "learning_rate": 5.583775860556717e-06,
+      "loss": 0.0225,
+      "step": 1379
+    },
+    {
+      "epoch": 0.47933310177144844,
+      "grad_norm": 0.3907616220341205,
+      "learning_rate": 5.578187909002428e-06,
+      "loss": 0.0246,
+      "step": 1380
+    },
+    {
+      "epoch": 0.47968044459881903,
+      "grad_norm": 0.5016926806740227,
+      "learning_rate": 5.572599225402974e-06,
+      "loss": 0.0196,
+      "step": 1381
+    },
+    {
+      "epoch": 0.4800277874261896,
+      "grad_norm": 0.5103118178433353,
+      "learning_rate": 5.567009816834199e-06,
+      "loss": 0.0155,
+      "step": 1382
+    },
+    {
+      "epoch": 0.48037513025356027,
+      "grad_norm": 0.9141679829913468,
+      "learning_rate": 5.561419690372869e-06,
+      "loss": 0.025,
+      "step": 1383
+    },
+    {
+      "epoch": 0.48072247308093086,
+      "grad_norm": 0.5420971166376601,
+      "learning_rate": 5.555828853096656e-06,
+      "loss": 0.0164,
+      "step": 1384
+    },
+    {
+      "epoch": 0.4810698159083015,
+      "grad_norm": 0.6355535194466504,
+      "learning_rate": 5.5502373120841346e-06,
+      "loss": 0.0214,
+      "step": 1385
+    },
+    {
+      "epoch": 0.4814171587356721,
+      "grad_norm": 0.3671380997420511,
+      "learning_rate": 5.544645074414768e-06,
+      "loss": 0.0149,
+      "step": 1386
+    },
+    {
+      "epoch": 0.48176450156304274,
+      "grad_norm": 0.8978543017268891,
+      "learning_rate": 5.539052147168903e-06,
+      "loss": 0.0133,
+      "step": 1387
+    },
+    {
+      "epoch": 0.4821118443904133,
+      "grad_norm": 0.9548917687698516,
+      "learning_rate": 5.533458537427758e-06,
+      "loss": 0.0214,
+      "step": 1388
+    },
+    {
+      "epoch": 0.48245918721778397,
+      "grad_norm": 0.8756025165481638,
+      "learning_rate": 5.5278642522734175e-06,
+      "loss": 0.0225,
+      "step": 1389
+    },
+    {
+      "epoch": 0.48280653004515456,
+      "grad_norm": 0.9092163623639533,
+      "learning_rate": 5.52226929878882e-06,
+      "loss": 0.0146,
+      "step": 1390
+    },
+    {
+      "epoch": 0.4831538728725252,
+      "grad_norm": 0.39815894527941675,
+      "learning_rate": 5.516673684057747e-06,
+      "loss": 0.0166,
+      "step": 1391
+    },
+    {
+      "epoch": 0.4835012156998958,
+      "grad_norm": 0.2966587066038902,
+      "learning_rate": 5.511077415164825e-06,
+      "loss": 0.0159,
+      "step": 1392
+    },
+    {
+      "epoch": 0.48384855852726644,
+      "grad_norm": 0.4884869491519517,
+      "learning_rate": 5.505480499195502e-06,
+      "loss": 0.0166,
+      "step": 1393
+    },
+    {
+      "epoch": 0.484195901354637,
+      "grad_norm": 0.35698741794966127,
+      "learning_rate": 5.499882943236045e-06,
+      "loss": 0.0114,
+      "step": 1394
+    },
+    {
+      "epoch": 0.4845432441820076,
+      "grad_norm": 0.7090902718663725,
+      "learning_rate": 5.494284754373538e-06,
+      "loss": 0.0146,
+      "step": 1395
+    },
+    {
+      "epoch": 0.48489058700937826,
+      "grad_norm": 0.5026358347696848,
+      "learning_rate": 5.488685939695862e-06,
+      "loss": 0.0148,
+      "step": 1396
+    },
+    {
+      "epoch": 0.48523792983674885,
+      "grad_norm": 0.7545042851491367,
+      "learning_rate": 5.4830865062916835e-06,
+      "loss": 0.022,
+      "step": 1397
+    },
+    {
+      "epoch": 0.4855852726641195,
+      "grad_norm": 0.5369844064837273,
+      "learning_rate": 5.477486461250469e-06,
+      "loss": 0.0209,
+      "step": 1398
+    },
+    {
+      "epoch": 0.4859326154914901,
+      "grad_norm": 0.6861158062273787,
+      "learning_rate": 5.471885811662442e-06,
+      "loss": 0.0271,
+      "step": 1399
+    },
+    {
+      "epoch": 0.4862799583188607,
+      "grad_norm": 0.7673744461691656,
+      "learning_rate": 5.466284564618603e-06,
+      "loss": 0.0296,
+      "step": 1400
+    },
+    {
+      "epoch": 0.4866273011462313,
+      "grad_norm": 0.5475029370005332,
+      "learning_rate": 5.460682727210702e-06,
+      "loss": 0.0248,
+      "step": 1401
+    },
+    {
+      "epoch": 0.48697464397360196,
+      "grad_norm": 0.9512795806158605,
+      "learning_rate": 5.455080306531244e-06,
+      "loss": 0.0218,
+      "step": 1402
+    },
+    {
+      "epoch": 0.48732198680097255,
+      "grad_norm": 0.5444903179080952,
+      "learning_rate": 5.449477309673462e-06,
+      "loss": 0.0191,
+      "step": 1403
+    },
+    {
+      "epoch": 0.4876693296283432,
+      "grad_norm": 0.8861826575673567,
+      "learning_rate": 5.443873743731331e-06,
+      "loss": 0.0109,
+      "step": 1404
+    },
+    {
+      "epoch": 0.4880166724557138,
+      "grad_norm": 0.9771566586732063,
+      "learning_rate": 5.438269615799534e-06,
+      "loss": 0.0197,
+      "step": 1405
+    },
+    {
+      "epoch": 0.48836401528308443,
+      "grad_norm": 1.9427129264091787,
+      "learning_rate": 5.432664932973474e-06,
+      "loss": 0.0282,
+      "step": 1406
+    },
+    {
+      "epoch": 0.488711358110455,
+      "grad_norm": 1.7104554294560776,
+      "learning_rate": 5.427059702349255e-06,
+      "loss": 0.0204,
+      "step": 1407
+    },
+    {
+      "epoch": 0.4890587009378256,
+      "grad_norm": 1.526360473399325,
+      "learning_rate": 5.4214539310236716e-06,
+      "loss": 0.0235,
+      "step": 1408
+    },
+    {
+      "epoch": 0.48940604376519625,
+      "grad_norm": 0.4997923973605721,
+      "learning_rate": 5.4158476260942075e-06,
+      "loss": 0.0174,
+      "step": 1409
+    },
+    {
+      "epoch": 0.48975338659256684,
+      "grad_norm": 0.4525414144834577,
+      "learning_rate": 5.410240794659016e-06,
+      "loss": 0.0195,
+      "step": 1410
+    },
+    {
+      "epoch": 0.4901007294199375,
+      "grad_norm": 0.410000828468513,
+      "learning_rate": 5.4046334438169245e-06,
+      "loss": 0.018,
+      "step": 1411
+    },
+    {
+      "epoch": 0.4904480722473081,
+      "grad_norm": 0.8060585957623981,
+      "learning_rate": 5.39902558066741e-06,
+      "loss": 0.0186,
+      "step": 1412
+    },
+    {
+      "epoch": 0.4907954150746787,
+      "grad_norm": 2.1170543916864037,
+      "learning_rate": 5.393417212310605e-06,
+      "loss": 0.0307,
+      "step": 1413
+    },
+    {
+      "epoch": 0.4911427579020493,
+      "grad_norm": 0.8364632533994034,
+      "learning_rate": 5.387808345847277e-06,
+      "loss": 0.0214,
+      "step": 1414
+    },
+    {
+      "epoch": 0.49149010072941995,
+      "grad_norm": 1.3428878485248303,
+      "learning_rate": 5.382198988378829e-06,
+      "loss": 0.0218,
+      "step": 1415
+    },
+    {
+      "epoch": 0.49183744355679054,
+      "grad_norm": 0.6747820600344949,
+      "learning_rate": 5.376589147007279e-06,
+      "loss": 0.0184,
+      "step": 1416
+    },
+    {
+      "epoch": 0.4921847863841612,
+      "grad_norm": 0.8594425526853627,
+      "learning_rate": 5.3709788288352615e-06,
+      "loss": 0.0182,
+      "step": 1417
+    },
+    {
+      "epoch": 0.4925321292115318,
+      "grad_norm": 0.5487772743825399,
+      "learning_rate": 5.365368040966016e-06,
+      "loss": 0.0221,
+      "step": 1418
+    },
+    {
+      "epoch": 0.4928794720389024,
+      "grad_norm": 0.8978739980804368,
+      "learning_rate": 5.359756790503376e-06,
+      "loss": 0.0161,
+      "step": 1419
+    },
+    {
+      "epoch": 0.493226814866273,
+      "grad_norm": 1.165143564643093,
+      "learning_rate": 5.354145084551757e-06,
+      "loss": 0.0215,
+      "step": 1420
+    },
+    {
+      "epoch": 0.4935741576936436,
+      "grad_norm": 0.8557699204572717,
+      "learning_rate": 5.348532930216157e-06,
+      "loss": 0.0239,
+      "step": 1421
+    },
+    {
+      "epoch": 0.49392150052101425,
+      "grad_norm": 1.203420881332666,
+      "learning_rate": 5.342920334602137e-06,
+      "loss": 0.0199,
+      "step": 1422
+    },
+    {
+      "epoch": 0.49426884334838483,
+      "grad_norm": 0.49131730078379343,
+      "learning_rate": 5.337307304815817e-06,
+      "loss": 0.0211,
+      "step": 1423
+    },
+    {
+      "epoch": 0.4946161861757555,
+      "grad_norm": 0.7833574610940411,
+      "learning_rate": 5.331693847963871e-06,
+      "loss": 0.0167,
+      "step": 1424
+    },
+    {
+      "epoch": 0.49496352900312607,
+      "grad_norm": 0.4454425944376909,
+      "learning_rate": 5.32607997115351e-06,
+      "loss": 0.0218,
+      "step": 1425
+    },
+    {
+      "epoch": 0.4953108718304967,
+      "grad_norm": 0.4471730702798169,
+      "learning_rate": 5.320465681492478e-06,
+      "loss": 0.016,
+      "step": 1426
+    },
+    {
+      "epoch": 0.4956582146578673,
+      "grad_norm": 0.9677891658268274,
+      "learning_rate": 5.31485098608904e-06,
+      "loss": 0.0142,
+      "step": 1427
+    },
+    {
+      "epoch": 0.49600555748523795,
+      "grad_norm": 0.5375414719758962,
+      "learning_rate": 5.309235892051976e-06,
+      "loss": 0.0147,
+      "step": 1428
+    },
+    {
+      "epoch": 0.49635290031260854,
+      "grad_norm": 0.5293248840266622,
+      "learning_rate": 5.303620406490573e-06,
+      "loss": 0.0155,
+      "step": 1429
+    },
+    {
+      "epoch": 0.4967002431399792,
+      "grad_norm": 0.7923702737792816,
+      "learning_rate": 5.298004536514606e-06,
+      "loss": 0.0137,
+      "step": 1430
+    },
+    {
+      "epoch": 0.49704758596734977,
+      "grad_norm": 1.2096835572281968,
+      "learning_rate": 5.292388289234349e-06,
+      "loss": 0.0227,
+      "step": 1431
+    },
+    {
+      "epoch": 0.4973949287947204,
+      "grad_norm": 0.9552234758771334,
+      "learning_rate": 5.286771671760541e-06,
+      "loss": 0.0318,
+      "step": 1432
+    },
+    {
+      "epoch": 0.497742271622091,
+      "grad_norm": 0.9769572098077361,
+      "learning_rate": 5.2811546912044e-06,
+      "loss": 0.0211,
+      "step": 1433
+    },
+    {
+      "epoch": 0.4980896144494616,
+      "grad_norm": 1.3870302191088562,
+      "learning_rate": 5.275537354677595e-06,
+      "loss": 0.0262,
+      "step": 1434
+    },
+    {
+      "epoch": 0.49843695727683224,
+      "grad_norm": 0.455287552427971,
+      "learning_rate": 5.2699196692922546e-06,
+      "loss": 0.0153,
+      "step": 1435
+    },
+    {
+      "epoch": 0.4987843001042028,
+      "grad_norm": 0.45542023402718185,
+      "learning_rate": 5.264301642160939e-06,
+      "loss": 0.0133,
+      "step": 1436
+    },
+    {
+      "epoch": 0.49913164293157347,
+      "grad_norm": 0.5647376621549762,
+      "learning_rate": 5.2586832803966525e-06,
+      "loss": 0.0183,
+      "step": 1437
+    },
+    {
+      "epoch": 0.49947898575894406,
+      "grad_norm": 0.4851885791856978,
+      "learning_rate": 5.2530645911128135e-06,
+      "loss": 0.0159,
+      "step": 1438
+    },
+    {
+      "epoch": 0.4998263285863147,
+      "grad_norm": 0.8448863799291079,
+      "learning_rate": 5.247445581423257e-06,
+      "loss": 0.015,
+      "step": 1439
+    },
+    {
+      "epoch": 0.5001736714136853,
+      "grad_norm": 0.9930674011154853,
+      "learning_rate": 5.24182625844223e-06,
+      "loss": 0.0191,
+      "step": 1440
+    },
+    {
+      "epoch": 0.5005210142410559,
+      "grad_norm": 0.9939816699848936,
+      "learning_rate": 5.236206629284367e-06,
+      "loss": 0.0147,
+      "step": 1441
+    },
+    {
+      "epoch": 0.5008683570684266,
+      "grad_norm": 0.8530137576376285,
+      "learning_rate": 5.2305867010646975e-06,
+      "loss": 0.0131,
+      "step": 1442
+    },
+    {
+      "epoch": 0.5012156998957972,
+      "grad_norm": 0.6190001699293848,
+      "learning_rate": 5.224966480898624e-06,
+      "loss": 0.0205,
+      "step": 1443
+    },
+    {
+      "epoch": 0.5015630427231678,
+      "grad_norm": 0.5694946459177951,
+      "learning_rate": 5.219345975901925e-06,
+      "loss": 0.0125,
+      "step": 1444
+    },
+    {
+      "epoch": 0.5019103855505384,
+      "grad_norm": 0.6351471531109476,
+      "learning_rate": 5.2137251931907315e-06,
+      "loss": 0.0133,
+      "step": 1445
+    },
+    {
+      "epoch": 0.5022577283779089,
+      "grad_norm": 0.9258615858237176,
+      "learning_rate": 5.208104139881537e-06,
+      "loss": 0.0228,
+      "step": 1446
+    },
+    {
+      "epoch": 0.5026050712052796,
+      "grad_norm": 1.1728546136592717,
+      "learning_rate": 5.202482823091165e-06,
+      "loss": 0.027,
+      "step": 1447
+    },
+    {
+      "epoch": 0.5029524140326502,
+      "grad_norm": 0.4906390164717808,
+      "learning_rate": 5.196861249936782e-06,
+      "loss": 0.0164,
+      "step": 1448
+    },
+    {
+      "epoch": 0.5032997568600208,
+      "grad_norm": 0.8785844081111323,
+      "learning_rate": 5.191239427535876e-06,
+      "loss": 0.03,
+      "step": 1449
+    },
+    {
+      "epoch": 0.5036470996873914,
+      "grad_norm": 0.47561204331956825,
+      "learning_rate": 5.185617363006249e-06,
+      "loss": 0.0164,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5039944425147621,
+      "grad_norm": 1.0364600327319036,
+      "learning_rate": 5.179995063466011e-06,
+      "loss": 0.0224,
+      "step": 1451
+    },
+    {
+      "epoch": 0.5043417853421327,
+      "grad_norm": 0.4607390170994428,
+      "learning_rate": 5.174372536033572e-06,
+      "loss": 0.0186,
+      "step": 1452
+    },
+    {
+      "epoch": 0.5046891281695033,
+      "grad_norm": 1.258104528985896,
+      "learning_rate": 5.168749787827625e-06,
+      "loss": 0.0207,
+      "step": 1453
+    },
+    {
+      "epoch": 0.5050364709968739,
+      "grad_norm": 0.4431006510218478,
+      "learning_rate": 5.163126825967147e-06,
+      "loss": 0.0248,
+      "step": 1454
+    },
+    {
+      "epoch": 0.5053838138242446,
+      "grad_norm": 1.0796514546664886,
+      "learning_rate": 5.157503657571386e-06,
+      "loss": 0.0226,
+      "step": 1455
+    },
+    {
+      "epoch": 0.5057311566516152,
+      "grad_norm": 0.7305079505359174,
+      "learning_rate": 5.151880289759847e-06,
+      "loss": 0.0162,
+      "step": 1456
+    },
+    {
+      "epoch": 0.5060784994789858,
+      "grad_norm": 0.5142991616801853,
+      "learning_rate": 5.14625672965229e-06,
+      "loss": 0.0231,
+      "step": 1457
+    },
+    {
+      "epoch": 0.5064258423063563,
+      "grad_norm": 0.3053355660863188,
+      "learning_rate": 5.140632984368721e-06,
+      "loss": 0.0169,
+      "step": 1458
+    },
+    {
+      "epoch": 0.5067731851337269,
+      "grad_norm": 0.6123341558370508,
+      "learning_rate": 5.1350090610293765e-06,
+      "loss": 0.0229,
+      "step": 1459
+    },
+    {
+      "epoch": 0.5071205279610976,
+      "grad_norm": 0.4030206492046645,
+      "learning_rate": 5.12938496675472e-06,
+      "loss": 0.0175,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5074678707884682,
+      "grad_norm": 1.0886198281053343,
+      "learning_rate": 5.123760708665432e-06,
+      "loss": 0.0186,
+      "step": 1461
+    },
+    {
+      "epoch": 0.5078152136158388,
+      "grad_norm": 0.48403013770913567,
+      "learning_rate": 5.1181362938823995e-06,
+      "loss": 0.0186,
+      "step": 1462
+    },
+    {
+      "epoch": 0.5081625564432094,
+      "grad_norm": 0.740876359392756,
+      "learning_rate": 5.112511729526708e-06,
+      "loss": 0.0202,
+      "step": 1463
+    },
+    {
+      "epoch": 0.5085098992705801,
+      "grad_norm": 1.3075365516854722,
+      "learning_rate": 5.106887022719633e-06,
+      "loss": 0.0182,
+      "step": 1464
+    },
+    {
+      "epoch": 0.5088572420979507,
+      "grad_norm": 0.5392655690852264,
+      "learning_rate": 5.101262180582628e-06,
+      "loss": 0.0274,
+      "step": 1465
+    },
+    {
+      "epoch": 0.5092045849253213,
+      "grad_norm": 0.7466155039978699,
+      "learning_rate": 5.095637210237324e-06,
+      "loss": 0.0198,
+      "step": 1466
+    },
+    {
+      "epoch": 0.5095519277526919,
+      "grad_norm": 0.4642989380683977,
+      "learning_rate": 5.090012118805505e-06,
+      "loss": 0.0235,
+      "step": 1467
+    },
+    {
+      "epoch": 0.5098992705800626,
+      "grad_norm": 0.5150423584413911,
+      "learning_rate": 5.084386913409118e-06,
+      "loss": 0.0256,
+      "step": 1468
+    },
+    {
+      "epoch": 0.5102466134074332,
+      "grad_norm": 0.5093614850962735,
+      "learning_rate": 5.0787616011702455e-06,
+      "loss": 0.0195,
+      "step": 1469
+    },
+    {
+      "epoch": 0.5105939562348037,
+      "grad_norm": 0.7660637501450626,
+      "learning_rate": 5.073136189211114e-06,
+      "loss": 0.0176,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5109412990621743,
+      "grad_norm": 1.0473364217411674,
+      "learning_rate": 5.067510684654069e-06,
+      "loss": 0.0261,
+      "step": 1471
+    },
+    {
+      "epoch": 0.5112886418895449,
+      "grad_norm": 0.8782397765062985,
+      "learning_rate": 5.061885094621575e-06,
+      "loss": 0.0199,
+      "step": 1472
+    },
+    {
+      "epoch": 0.5116359847169156,
+      "grad_norm": 0.4607201639452861,
+      "learning_rate": 5.056259426236207e-06,
+      "loss": 0.0146,
+      "step": 1473
+    },
+    {
+      "epoch": 0.5119833275442862,
+      "grad_norm": 0.6661850564864881,
+      "learning_rate": 5.05063368662064e-06,
+      "loss": 0.0179,
+      "step": 1474
+    },
+    {
+      "epoch": 0.5123306703716568,
+      "grad_norm": 0.4772014886025317,
+      "learning_rate": 5.0450078828976326e-06,
+      "loss": 0.0198,
+      "step": 1475
+    },
+    {
+      "epoch": 0.5126780131990274,
+      "grad_norm": 1.305023780400015,
+      "learning_rate": 5.0393820221900325e-06,
+      "loss": 0.0219,
+      "step": 1476
+    },
+    {
+      "epoch": 0.5130253560263981,
+      "grad_norm": 0.8537297309135922,
+      "learning_rate": 5.0337561116207546e-06,
+      "loss": 0.0191,
+      "step": 1477
+    },
+    {
+      "epoch": 0.5133726988537687,
+      "grad_norm": 0.5326568118271483,
+      "learning_rate": 5.028130158312779e-06,
+      "loss": 0.0246,
+      "step": 1478
+    },
+    {
+      "epoch": 0.5137200416811393,
+      "grad_norm": 0.3866710810781936,
+      "learning_rate": 5.02250416938914e-06,
+      "loss": 0.0178,
+      "step": 1479
+    },
+    {
+      "epoch": 0.5140673845085099,
+      "grad_norm": 0.22396563755414556,
+      "learning_rate": 5.016878151972915e-06,
+      "loss": 0.0123,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5144147273358806,
+      "grad_norm": 0.4451777326858171,
+      "learning_rate": 5.01125211318722e-06,
+      "loss": 0.0214,
+      "step": 1481
+    },
+    {
+      "epoch": 0.5147620701632512,
+      "grad_norm": 0.6070766959527533,
+      "learning_rate": 5.005626060155194e-06,
+      "loss": 0.0252,
+      "step": 1482
+    },
+    {
+      "epoch": 0.5151094129906217,
+      "grad_norm": 1.0407371235155332,
+      "learning_rate": 5e-06,
+      "loss": 0.0221,
+      "step": 1483
+    },
+    {
+      "epoch": 0.5154567558179923,
+      "grad_norm": 0.9630652449889104,
+      "learning_rate": 4.994373939844807e-06,
+      "loss": 0.0292,
+      "step": 1484
+    },
+    {
+      "epoch": 0.5158040986453629,
+      "grad_norm": 0.5372033115683907,
+      "learning_rate": 4.988747886812781e-06,
+      "loss": 0.0332,
+      "step": 1485
+    },
+    {
+      "epoch": 0.5161514414727336,
+      "grad_norm": 0.42093625107990457,
+      "learning_rate": 4.983121848027088e-06,
+      "loss": 0.0154,
+      "step": 1486
+    },
+    {
+      "epoch": 0.5164987843001042,
+      "grad_norm": 0.5437961266541049,
+      "learning_rate": 4.977495830610862e-06,
+      "loss": 0.0162,
+      "step": 1487
+    },
+    {
+      "epoch": 0.5168461271274748,
+      "grad_norm": 0.6036534712036593,
+      "learning_rate": 4.9718698416872215e-06,
+      "loss": 0.0267,
+      "step": 1488
+    },
+    {
+      "epoch": 0.5171934699548454,
+      "grad_norm": 0.531411114394444,
+      "learning_rate": 4.966243888379245e-06,
+      "loss": 0.0273,
+      "step": 1489
+    },
+    {
+      "epoch": 0.5175408127822161,
+      "grad_norm": 0.907790815597591,
+      "learning_rate": 4.96061797780997e-06,
+      "loss": 0.0143,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5178881556095867,
+      "grad_norm": 1.0791005798691276,
+      "learning_rate": 4.954992117102369e-06,
+      "loss": 0.0203,
+      "step": 1491
+    },
+    {
+      "epoch": 0.5182354984369573,
+      "grad_norm": 1.0050228768407403,
+      "learning_rate": 4.949366313379362e-06,
+      "loss": 0.0159,
+      "step": 1492
+    },
+    {
+      "epoch": 0.5185828412643279,
+      "grad_norm": 0.31620179318811703,
+      "learning_rate": 4.943740573763794e-06,
+      "loss": 0.0181,
+      "step": 1493
+    },
+    {
+      "epoch": 0.5189301840916986,
+      "grad_norm": 0.693985447212686,
+      "learning_rate": 4.938114905378428e-06,
+      "loss": 0.0143,
+      "step": 1494
+    },
+    {
+      "epoch": 0.5192775269190691,
+      "grad_norm": 0.9981712152035553,
+      "learning_rate": 4.932489315345933e-06,
+      "loss": 0.022,
+      "step": 1495
+    },
+    {
+      "epoch": 0.5196248697464397,
+      "grad_norm": 0.47907771487206496,
+      "learning_rate": 4.9268638107888875e-06,
+      "loss": 0.0158,
+      "step": 1496
+    },
+    {
+      "epoch": 0.5199722125738103,
+      "grad_norm": 0.6623027870764544,
+      "learning_rate": 4.9212383988297545e-06,
+      "loss": 0.0134,
+      "step": 1497
+    },
+    {
+      "epoch": 0.5203195554011809,
+      "grad_norm": 1.0794034174253464,
+      "learning_rate": 4.9156130865908845e-06,
+      "loss": 0.0292,
+      "step": 1498
+    },
+    {
+      "epoch": 0.5206668982285516,
+      "grad_norm": 0.3138682448100754,
+      "learning_rate": 4.9099878811944965e-06,
+      "loss": 0.0155,
+      "step": 1499
+    },
+    {
+      "epoch": 0.5210142410559222,
+      "grad_norm": 0.7583850057995437,
+      "learning_rate": 4.904362789762677e-06,
+      "loss": 0.0147,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5213615838832928,
+      "grad_norm": 0.7155832485484563,
+      "learning_rate": 4.898737819417372e-06,
+      "loss": 0.0203,
+      "step": 1501
+    },
+    {
+      "epoch": 0.5217089267106634,
+      "grad_norm": 0.35672739687797744,
+      "learning_rate": 4.893112977280369e-06,
+      "loss": 0.0069,
+      "step": 1502
+    },
+    {
+      "epoch": 0.5220562695380341,
+      "grad_norm": 0.987017131142079,
+      "learning_rate": 4.887488270473294e-06,
+      "loss": 0.0232,
+      "step": 1503
+    },
+    {
+      "epoch": 0.5224036123654047,
+      "grad_norm": 0.34804891464014953,
+      "learning_rate": 4.881863706117601e-06,
+      "loss": 0.0134,
+      "step": 1504
+    },
+    {
+      "epoch": 0.5227509551927753,
+      "grad_norm": 0.6251911038205994,
+      "learning_rate": 4.876239291334568e-06,
+      "loss": 0.0202,
+      "step": 1505
+    },
+    {
+      "epoch": 0.5230982980201458,
+      "grad_norm": 1.3491600406620532,
+      "learning_rate": 4.8706150332452815e-06,
+      "loss": 0.0237,
+      "step": 1506
+    },
+    {
+      "epoch": 0.5234456408475165,
+      "grad_norm": 0.6296332865053156,
+      "learning_rate": 4.864990938970624e-06,
+      "loss": 0.0162,
+      "step": 1507
+    },
+    {
+      "epoch": 0.5237929836748871,
+      "grad_norm": 0.6175960968496856,
+      "learning_rate": 4.85936701563128e-06,
+      "loss": 0.0223,
+      "step": 1508
+    },
+    {
+      "epoch": 0.5241403265022577,
+      "grad_norm": 0.6135026142466486,
+      "learning_rate": 4.85374327034771e-06,
+      "loss": 0.0192,
+      "step": 1509
+    },
+    {
+      "epoch": 0.5244876693296283,
+      "grad_norm": 0.8443887642336403,
+      "learning_rate": 4.848119710240156e-06,
+      "loss": 0.0244,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5248350121569989,
+      "grad_norm": 0.7215133934035978,
+      "learning_rate": 4.842496342428616e-06,
+      "loss": 0.0306,
+      "step": 1511
+    },
+    {
+      "epoch": 0.5251823549843696,
+      "grad_norm": 0.3768241222787331,
+      "learning_rate": 4.8368731740328536e-06,
+      "loss": 0.0139,
+      "step": 1512
+    },
+    {
+      "epoch": 0.5255296978117402,
+      "grad_norm": 0.3510966580810152,
+      "learning_rate": 4.8312502121723755e-06,
+      "loss": 0.0137,
+      "step": 1513
+    },
+    {
+      "epoch": 0.5258770406391108,
+      "grad_norm": 0.9763840468277634,
+      "learning_rate": 4.825627463966431e-06,
+      "loss": 0.0218,
+      "step": 1514
+    },
+    {
+      "epoch": 0.5262243834664814,
+      "grad_norm": 0.9974349698601298,
+      "learning_rate": 4.8200049365339905e-06,
+      "loss": 0.0229,
+      "step": 1515
+    },
+    {
+      "epoch": 0.5265717262938521,
+      "grad_norm": 0.395840326358885,
+      "learning_rate": 4.814382636993753e-06,
+      "loss": 0.0224,
+      "step": 1516
+    },
+    {
+      "epoch": 0.5269190691212227,
+      "grad_norm": 0.7070158097071757,
+      "learning_rate": 4.808760572464126e-06,
+      "loss": 0.0237,
+      "step": 1517
+    },
+    {
+      "epoch": 0.5272664119485932,
+      "grad_norm": 0.34439167819910455,
+      "learning_rate": 4.80313875006322e-06,
+      "loss": 0.0209,
+      "step": 1518
+    },
+    {
+      "epoch": 0.5276137547759638,
+      "grad_norm": 0.333265253316496,
+      "learning_rate": 4.7975171769088366e-06,
+      "loss": 0.0203,
+      "step": 1519
+    },
+    {
+      "epoch": 0.5279610976033345,
+      "grad_norm": 0.7184684037655478,
+      "learning_rate": 4.791895860118465e-06,
+      "loss": 0.0182,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5283084404307051,
+      "grad_norm": 0.8473043460095674,
+      "learning_rate": 4.7862748068092685e-06,
+      "loss": 0.0174,
+      "step": 1521
+    },
+    {
+      "epoch": 0.5286557832580757,
+      "grad_norm": 0.376539100887896,
+      "learning_rate": 4.780654024098076e-06,
+      "loss": 0.0167,
+      "step": 1522
+    },
+    {
+      "epoch": 0.5290031260854463,
+      "grad_norm": 1.1649611690839885,
+      "learning_rate": 4.775033519101378e-06,
+      "loss": 0.0217,
+      "step": 1523
+    },
+    {
+      "epoch": 0.5293504689128169,
+      "grad_norm": 0.62058919496297,
+      "learning_rate": 4.769413298935305e-06,
+      "loss": 0.0136,
+      "step": 1524
+    },
+    {
+      "epoch": 0.5296978117401876,
+      "grad_norm": 0.5251282587265036,
+      "learning_rate": 4.763793370715635e-06,
+      "loss": 0.0219,
+      "step": 1525
+    },
+    {
+      "epoch": 0.5300451545675582,
+      "grad_norm": 0.6236545128248215,
+      "learning_rate": 4.758173741557772e-06,
+      "loss": 0.0181,
+      "step": 1526
+    },
+    {
+      "epoch": 0.5303924973949288,
+      "grad_norm": 0.9945963688156065,
+      "learning_rate": 4.752554418576744e-06,
+      "loss": 0.0203,
+      "step": 1527
+    },
+    {
+      "epoch": 0.5307398402222994,
+      "grad_norm": 0.45683681677249527,
+      "learning_rate": 4.746935408887188e-06,
+      "loss": 0.0094,
+      "step": 1528
+    },
+    {
+      "epoch": 0.5310871830496701,
+      "grad_norm": 0.2846095731280691,
+      "learning_rate": 4.741316719603348e-06,
+      "loss": 0.0102,
+      "step": 1529
+    },
+    {
+      "epoch": 0.5314345258770407,
+      "grad_norm": 0.3347130074226901,
+      "learning_rate": 4.735698357839061e-06,
+      "loss": 0.016,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5317818687044112,
+      "grad_norm": 0.28383945406870187,
+      "learning_rate": 4.730080330707748e-06,
+      "loss": 0.018,
+      "step": 1531
+    },
+    {
+      "epoch": 0.5321292115317818,
+      "grad_norm": 0.39618404004198604,
+      "learning_rate": 4.724462645322406e-06,
+      "loss": 0.0146,
+      "step": 1532
+    },
+    {
+      "epoch": 0.5324765543591525,
+      "grad_norm": 0.48107525602946016,
+      "learning_rate": 4.718845308795601e-06,
+      "loss": 0.0084,
+      "step": 1533
+    },
+    {
+      "epoch": 0.5328238971865231,
+      "grad_norm": 0.7775999794726074,
+      "learning_rate": 4.71322832823946e-06,
+      "loss": 0.017,
+      "step": 1534
+    },
+    {
+      "epoch": 0.5331712400138937,
+      "grad_norm": 1.069007509619891,
+      "learning_rate": 4.707611710765654e-06,
+      "loss": 0.0288,
+      "step": 1535
+    },
+    {
+      "epoch": 0.5335185828412643,
+      "grad_norm": 0.39849903984305235,
+      "learning_rate": 4.701995463485395e-06,
+      "loss": 0.0117,
+      "step": 1536
+    },
+    {
+      "epoch": 0.5338659256686349,
+      "grad_norm": 0.677236043933078,
+      "learning_rate": 4.696379593509429e-06,
+      "loss": 0.015,
+      "step": 1537
+    },
+    {
+      "epoch": 0.5342132684960056,
+      "grad_norm": 0.4265486094005142,
+      "learning_rate": 4.690764107948025e-06,
+      "loss": 0.0122,
+      "step": 1538
+    },
+    {
+      "epoch": 0.5345606113233762,
+      "grad_norm": 0.4715494638944048,
+      "learning_rate": 4.685149013910962e-06,
+      "loss": 0.0156,
+      "step": 1539
+    },
+    {
+      "epoch": 0.5349079541507468,
+      "grad_norm": 0.7150624971216185,
+      "learning_rate": 4.6795343185075235e-06,
+      "loss": 0.0169,
+      "step": 1540
+    },
+    {
+      "epoch": 0.5352552969781174,
+      "grad_norm": 0.6525023335363956,
+      "learning_rate": 4.6739200288464905e-06,
+      "loss": 0.0291,
+      "step": 1541
+    },
+    {
+      "epoch": 0.5356026398054881,
+      "grad_norm": 0.47839611155737516,
+      "learning_rate": 4.668306152036129e-06,
+      "loss": 0.0162,
+      "step": 1542
+    },
+    {
+      "epoch": 0.5359499826328586,
+      "grad_norm": 0.7470286100606716,
+      "learning_rate": 4.662692695184184e-06,
+      "loss": 0.026,
+      "step": 1543
+    },
+    {
+      "epoch": 0.5362973254602292,
+      "grad_norm": 0.4794922661466092,
+      "learning_rate": 4.657079665397865e-06,
+      "loss": 0.0233,
+      "step": 1544
+    },
+    {
+      "epoch": 0.5366446682875998,
+      "grad_norm": 0.7903918813270774,
+      "learning_rate": 4.651467069783845e-06,
+      "loss": 0.0232,
+      "step": 1545
+    },
+    {
+      "epoch": 0.5369920111149705,
+      "grad_norm": 1.610986156977707,
+      "learning_rate": 4.645854915448243e-06,
+      "loss": 0.0298,
+      "step": 1546
+    },
+    {
+      "epoch": 0.5373393539423411,
+      "grad_norm": 0.6560078832036462,
+      "learning_rate": 4.640243209496627e-06,
+      "loss": 0.0186,
+      "step": 1547
+    },
+    {
+      "epoch": 0.5376866967697117,
+      "grad_norm": 1.6026250043381345,
+      "learning_rate": 4.634631959033985e-06,
+      "loss": 0.0218,
+      "step": 1548
+    },
+    {
+      "epoch": 0.5380340395970823,
+      "grad_norm": 0.6898766822043673,
+      "learning_rate": 4.62902117116474e-06,
+      "loss": 0.017,
+      "step": 1549
+    },
+    {
+      "epoch": 0.5383813824244529,
+      "grad_norm": 0.5467408900610035,
+      "learning_rate": 4.623410852992724e-06,
+      "loss": 0.0233,
+      "step": 1550
+    },
+    {
+      "epoch": 0.5387287252518236,
+      "grad_norm": 1.0551467144246778,
+      "learning_rate": 4.617801011621175e-06,
+      "loss": 0.0208,
+      "step": 1551
+    },
+    {
+      "epoch": 0.5390760680791942,
+      "grad_norm": 1.162941958095386,
+      "learning_rate": 4.6121916541527235e-06,
+      "loss": 0.0259,
+      "step": 1552
+    },
+    {
+      "epoch": 0.5394234109065648,
+      "grad_norm": 0.3761300022318631,
+      "learning_rate": 4.606582787689396e-06,
+      "loss": 0.0185,
+      "step": 1553
+    },
+    {
+      "epoch": 0.5397707537339353,
+      "grad_norm": 0.6376765879619961,
+      "learning_rate": 4.600974419332591e-06,
+      "loss": 0.0172,
+      "step": 1554
+    },
+    {
+      "epoch": 0.540118096561306,
+      "grad_norm": 0.8179982640637906,
+      "learning_rate": 4.595366556183079e-06,
+      "loss": 0.0192,
+      "step": 1555
+    },
+    {
+      "epoch": 0.5404654393886766,
+      "grad_norm": 0.6660409394522981,
+      "learning_rate": 4.589759205340986e-06,
+      "loss": 0.0201,
+      "step": 1556
+    },
+    {
+      "epoch": 0.5408127822160472,
+      "grad_norm": 0.37364095824936294,
+      "learning_rate": 4.584152373905794e-06,
+      "loss": 0.0242,
+      "step": 1557
+    },
+    {
+      "epoch": 0.5411601250434178,
+      "grad_norm": 0.6214912023671911,
+      "learning_rate": 4.578546068976329e-06,
+      "loss": 0.0186,
+      "step": 1558
+    },
+    {
+      "epoch": 0.5415074678707885,
+      "grad_norm": 0.3570650197481583,
+      "learning_rate": 4.572940297650747e-06,
+      "loss": 0.0137,
+      "step": 1559
+    },
+    {
+      "epoch": 0.5418548106981591,
+      "grad_norm": 0.5029530205971685,
+      "learning_rate": 4.567335067026528e-06,
+      "loss": 0.0144,
+      "step": 1560
+    },
+    {
+      "epoch": 0.5422021535255297,
+      "grad_norm": 0.8095089739524693,
+      "learning_rate": 4.561730384200467e-06,
+      "loss": 0.021,
+      "step": 1561
+    },
+    {
+      "epoch": 0.5425494963529003,
+      "grad_norm": 0.45079050575045515,
+      "learning_rate": 4.556126256268671e-06,
+      "loss": 0.0199,
+      "step": 1562
+    },
+    {
+      "epoch": 0.5428968391802709,
+      "grad_norm": 0.544012994922186,
+      "learning_rate": 4.550522690326538e-06,
+      "loss": 0.0178,
+      "step": 1563
+    },
+    {
+      "epoch": 0.5432441820076416,
+      "grad_norm": 0.4351027380527296,
+      "learning_rate": 4.544919693468759e-06,
+      "loss": 0.0187,
+      "step": 1564
+    },
+    {
+      "epoch": 0.5435915248350122,
+      "grad_norm": 0.735577031685772,
+      "learning_rate": 4.539317272789299e-06,
+      "loss": 0.0176,
+      "step": 1565
+    },
+    {
+      "epoch": 0.5439388676623828,
+      "grad_norm": 0.7448332665427014,
+      "learning_rate": 4.533715435381398e-06,
+      "loss": 0.0203,
+      "step": 1566
+    },
+    {
+      "epoch": 0.5442862104897533,
+      "grad_norm": 0.8816422119244176,
+      "learning_rate": 4.528114188337559e-06,
+      "loss": 0.0205,
+      "step": 1567
+    },
+    {
+      "epoch": 0.544633553317124,
+      "grad_norm": 0.7125240794309594,
+      "learning_rate": 4.522513538749534e-06,
+      "loss": 0.019,
+      "step": 1568
+    },
+    {
+      "epoch": 0.5449808961444946,
+      "grad_norm": 0.6691735284658429,
+      "learning_rate": 4.516913493708317e-06,
+      "loss": 0.021,
+      "step": 1569
+    },
+    {
+      "epoch": 0.5453282389718652,
+      "grad_norm": 0.8271292240524128,
+      "learning_rate": 4.511314060304141e-06,
+      "loss": 0.0189,
+      "step": 1570
+    },
+    {
+      "epoch": 0.5456755817992358,
+      "grad_norm": 0.9574204304906477,
+      "learning_rate": 4.505715245626462e-06,
+      "loss": 0.023,
+      "step": 1571
+    },
+    {
+      "epoch": 0.5460229246266065,
+      "grad_norm": 0.6547097514027208,
+      "learning_rate": 4.500117056763956e-06,
+      "loss": 0.0225,
+      "step": 1572
+    },
+    {
+      "epoch": 0.5463702674539771,
+      "grad_norm": 1.0813224646564668,
+      "learning_rate": 4.494519500804501e-06,
+      "loss": 0.0173,
+      "step": 1573
+    },
+    {
+      "epoch": 0.5467176102813477,
+      "grad_norm": 0.9496654452606548,
+      "learning_rate": 4.488922584835177e-06,
+      "loss": 0.0146,
+      "step": 1574
+    },
+    {
+      "epoch": 0.5470649531087183,
+      "grad_norm": 0.49774440254371105,
+      "learning_rate": 4.483326315942253e-06,
+      "loss": 0.0157,
+      "step": 1575
+    },
+    {
+      "epoch": 0.5474122959360889,
+      "grad_norm": 0.4088463704083283,
+      "learning_rate": 4.477730701211183e-06,
+      "loss": 0.017,
+      "step": 1576
+    },
+    {
+      "epoch": 0.5477596387634596,
+      "grad_norm": 0.6482438150567166,
+      "learning_rate": 4.472135747726583e-06,
+      "loss": 0.0265,
+      "step": 1577
+    },
+    {
+      "epoch": 0.5481069815908302,
+      "grad_norm": 1.1293197974606506,
+      "learning_rate": 4.466541462572243e-06,
+      "loss": 0.0189,
+      "step": 1578
+    },
+    {
+      "epoch": 0.5484543244182007,
+      "grad_norm": 0.5822884700457213,
+      "learning_rate": 4.460947852831097e-06,
+      "loss": 0.0183,
+      "step": 1579
+    },
+    {
+      "epoch": 0.5488016672455713,
+      "grad_norm": 0.5604041072365497,
+      "learning_rate": 4.455354925585234e-06,
+      "loss": 0.0283,
+      "step": 1580
+    },
+    {
+      "epoch": 0.549149010072942,
+      "grad_norm": 0.44375445497432964,
+      "learning_rate": 4.449762687915866e-06,
+      "loss": 0.0131,
+      "step": 1581
+    },
+    {
+      "epoch": 0.5494963529003126,
+      "grad_norm": 1.0388835523370303,
+      "learning_rate": 4.444171146903345e-06,
+      "loss": 0.016,
+      "step": 1582
+    },
+    {
+      "epoch": 0.5498436957276832,
+      "grad_norm": 0.6674378343975669,
+      "learning_rate": 4.438580309627132e-06,
+      "loss": 0.0198,
+      "step": 1583
+    },
+    {
+      "epoch": 0.5501910385550538,
+      "grad_norm": 0.39270508786634684,
+      "learning_rate": 4.4329901831658035e-06,
+      "loss": 0.0128,
+      "step": 1584
+    },
+    {
+      "epoch": 0.5505383813824245,
+      "grad_norm": 0.7015995821207273,
+      "learning_rate": 4.427400774597028e-06,
+      "loss": 0.0246,
+      "step": 1585
+    },
+    {
+      "epoch": 0.5508857242097951,
+      "grad_norm": 0.4993737094804665,
+      "learning_rate": 4.421812090997573e-06,
+      "loss": 0.0231,
+      "step": 1586
+    },
+    {
+      "epoch": 0.5512330670371657,
+      "grad_norm": 0.549124903537338,
+      "learning_rate": 4.4162241394432834e-06,
+      "loss": 0.0219,
+      "step": 1587
+    },
+    {
+      "epoch": 0.5515804098645363,
+      "grad_norm": 0.9191413987980975,
+      "learning_rate": 4.4106369270090814e-06,
+      "loss": 0.0128,
+      "step": 1588
+    },
+    {
+      "epoch": 0.5519277526919069,
+      "grad_norm": 0.47115453283840025,
+      "learning_rate": 4.405050460768947e-06,
+      "loss": 0.0151,
+      "step": 1589
+    },
+    {
+      "epoch": 0.5522750955192776,
+      "grad_norm": 0.5360090594122309,
+      "learning_rate": 4.3994647477959205e-06,
+      "loss": 0.0194,
+      "step": 1590
+    },
+    {
+      "epoch": 0.5526224383466481,
+      "grad_norm": 0.6178093745994127,
+      "learning_rate": 4.393879795162088e-06,
+      "loss": 0.0185,
+      "step": 1591
+    },
+    {
+      "epoch": 0.5529697811740187,
+      "grad_norm": 0.5705601551799697,
+      "learning_rate": 4.388295609938572e-06,
+      "loss": 0.0194,
+      "step": 1592
+    },
+    {
+      "epoch": 0.5533171240013893,
+      "grad_norm": 0.6254120310087067,
+      "learning_rate": 4.3827121991955235e-06,
+      "loss": 0.0181,
+      "step": 1593
+    },
+    {
+      "epoch": 0.55366446682876,
+      "grad_norm": 0.3870391915004128,
+      "learning_rate": 4.37712957000211e-06,
+      "loss": 0.0148,
+      "step": 1594
+    },
+    {
+      "epoch": 0.5540118096561306,
+      "grad_norm": 0.9142400891756279,
+      "learning_rate": 4.371547729426517e-06,
+      "loss": 0.0171,
+      "step": 1595
+    },
+    {
+      "epoch": 0.5543591524835012,
+      "grad_norm": 0.49985847505258046,
+      "learning_rate": 4.365966684535925e-06,
+      "loss": 0.0201,
+      "step": 1596
+    },
+    {
+      "epoch": 0.5547064953108718,
+      "grad_norm": 1.3223267741569067,
+      "learning_rate": 4.360386442396508e-06,
+      "loss": 0.0191,
+      "step": 1597
+    },
+    {
+      "epoch": 0.5550538381382425,
+      "grad_norm": 0.3660339615210978,
+      "learning_rate": 4.354807010073425e-06,
+      "loss": 0.0116,
+      "step": 1598
+    },
+    {
+      "epoch": 0.5554011809656131,
+      "grad_norm": 1.0210647191568252,
+      "learning_rate": 4.349228394630808e-06,
+      "loss": 0.0294,
+      "step": 1599
+    },
+    {
+      "epoch": 0.5557485237929837,
+      "grad_norm": 0.5885533492032752,
+      "learning_rate": 4.3436506031317594e-06,
+      "loss": 0.0175,
+      "step": 1600
+    },
+    {
+      "epoch": 0.5560958666203543,
+      "grad_norm": 0.6342172768582458,
+      "learning_rate": 4.338073642638334e-06,
+      "loss": 0.014,
+      "step": 1601
+    },
+    {
+      "epoch": 0.5564432094477249,
+      "grad_norm": 0.5569542724926432,
+      "learning_rate": 4.3324975202115345e-06,
+      "loss": 0.0195,
+      "step": 1602
+    },
+    {
+      "epoch": 0.5567905522750956,
+      "grad_norm": 0.7223546169720154,
+      "learning_rate": 4.326922242911302e-06,
+      "loss": 0.0142,
+      "step": 1603
+    },
+    {
+      "epoch": 0.5571378951024661,
+      "grad_norm": 0.3683037424124626,
+      "learning_rate": 4.321347817796511e-06,
+      "loss": 0.0196,
+      "step": 1604
+    },
+    {
+      "epoch": 0.5574852379298367,
+      "grad_norm": 1.2092694294770592,
+      "learning_rate": 4.3157742519249576e-06,
+      "loss": 0.0251,
+      "step": 1605
+    },
+    {
+      "epoch": 0.5578325807572073,
+      "grad_norm": 0.4910970055629151,
+      "learning_rate": 4.3102015523533436e-06,
+      "loss": 0.0164,
+      "step": 1606
+    },
+    {
+      "epoch": 0.558179923584578,
+      "grad_norm": 0.36288578186019704,
+      "learning_rate": 4.304629726137279e-06,
+      "loss": 0.0142,
+      "step": 1607
+    },
+    {
+      "epoch": 0.5585272664119486,
+      "grad_norm": 0.5776094098850557,
+      "learning_rate": 4.299058780331267e-06,
+      "loss": 0.0147,
+      "step": 1608
+    },
+    {
+      "epoch": 0.5588746092393192,
+      "grad_norm": 0.4606276309503416,
+      "learning_rate": 4.293488721988698e-06,
+      "loss": 0.0171,
+      "step": 1609
+    },
+    {
+      "epoch": 0.5592219520666898,
+      "grad_norm": 1.9663566917991069,
+      "learning_rate": 4.287919558161835e-06,
+      "loss": 0.0173,
+      "step": 1610
+    },
+    {
+      "epoch": 0.5595692948940605,
+      "grad_norm": 0.9161917744112766,
+      "learning_rate": 4.28235129590181e-06,
+      "loss": 0.0164,
+      "step": 1611
+    },
+    {
+      "epoch": 0.5599166377214311,
+      "grad_norm": 0.3841077769256294,
+      "learning_rate": 4.276783942258613e-06,
+      "loss": 0.0136,
+      "step": 1612
+    },
+    {
+      "epoch": 0.5602639805488017,
+      "grad_norm": 0.40125517813129213,
+      "learning_rate": 4.27121750428109e-06,
+      "loss": 0.0181,
+      "step": 1613
+    },
+    {
+      "epoch": 0.5606113233761723,
+      "grad_norm": 0.8054623880288195,
+      "learning_rate": 4.265651989016915e-06,
+      "loss": 0.019,
+      "step": 1614
+    },
+    {
+      "epoch": 0.5609586662035428,
+      "grad_norm": 0.9023479920127903,
+      "learning_rate": 4.260087403512605e-06,
+      "loss": 0.0185,
+      "step": 1615
+    },
+    {
+      "epoch": 0.5613060090309135,
+      "grad_norm": 0.4648304244403317,
+      "learning_rate": 4.254523754813495e-06,
+      "loss": 0.0193,
+      "step": 1616
+    },
+    {
+      "epoch": 0.5616533518582841,
+      "grad_norm": 0.9102616101626732,
+      "learning_rate": 4.2489610499637346e-06,
+      "loss": 0.0173,
+      "step": 1617
+    },
+    {
+      "epoch": 0.5620006946856547,
+      "grad_norm": 0.5742925429597601,
+      "learning_rate": 4.243399296006276e-06,
+      "loss": 0.0224,
+      "step": 1618
+    },
+    {
+      "epoch": 0.5623480375130253,
+      "grad_norm": 1.1494398692641448,
+      "learning_rate": 4.237838499982874e-06,
+      "loss": 0.0195,
+      "step": 1619
+    },
+    {
+      "epoch": 0.562695380340396,
+      "grad_norm": 0.8792697527207582,
+      "learning_rate": 4.232278668934063e-06,
+      "loss": 0.0135,
+      "step": 1620
+    },
+    {
+      "epoch": 0.5630427231677666,
+      "grad_norm": 0.46740643343738697,
+      "learning_rate": 4.226719809899163e-06,
+      "loss": 0.0199,
+      "step": 1621
+    },
+    {
+      "epoch": 0.5633900659951372,
+      "grad_norm": 1.1632556584909721,
+      "learning_rate": 4.221161929916255e-06,
+      "loss": 0.0304,
+      "step": 1622
+    },
+    {
+      "epoch": 0.5637374088225078,
+      "grad_norm": 0.5407257650582535,
+      "learning_rate": 4.2156050360221855e-06,
+      "loss": 0.0109,
+      "step": 1623
+    },
+    {
+      "epoch": 0.5640847516498785,
+      "grad_norm": 0.6586876902269023,
+      "learning_rate": 4.210049135252554e-06,
+      "loss": 0.0149,
+      "step": 1624
+    },
+    {
+      "epoch": 0.5644320944772491,
+      "grad_norm": 0.7069930675788888,
+      "learning_rate": 4.204494234641701e-06,
+      "loss": 0.0099,
+      "step": 1625
+    },
+    {
+      "epoch": 0.5647794373046197,
+      "grad_norm": 0.8706649712202481,
+      "learning_rate": 4.198940341222699e-06,
+      "loss": 0.0259,
+      "step": 1626
+    },
+    {
+      "epoch": 0.5651267801319902,
+      "grad_norm": 0.4998724109402863,
+      "learning_rate": 4.193387462027343e-06,
+      "loss": 0.019,
+      "step": 1627
+    },
+    {
+      "epoch": 0.5654741229593608,
+      "grad_norm": 0.6031533765692226,
+      "learning_rate": 4.1878356040861525e-06,
+      "loss": 0.0192,
+      "step": 1628
+    },
+    {
+      "epoch": 0.5658214657867315,
+      "grad_norm": 0.7768746063951399,
+      "learning_rate": 4.182284774428348e-06,
+      "loss": 0.0232,
+      "step": 1629
+    },
+    {
+      "epoch": 0.5661688086141021,
+      "grad_norm": 0.46684861318924753,
+      "learning_rate": 4.176734980081845e-06,
+      "loss": 0.0174,
+      "step": 1630
+    },
+    {
+      "epoch": 0.5665161514414727,
+      "grad_norm": 1.2406437022629773,
+      "learning_rate": 4.171186228073256e-06,
+      "loss": 0.026,
+      "step": 1631
+    },
+    {
+      "epoch": 0.5668634942688433,
+      "grad_norm": 0.42437690132062417,
+      "learning_rate": 4.165638525427867e-06,
+      "loss": 0.0181,
+      "step": 1632
+    },
+    {
+      "epoch": 0.567210837096214,
+      "grad_norm": 0.4618761042287134,
+      "learning_rate": 4.160091879169642e-06,
+      "loss": 0.0179,
+      "step": 1633
+    },
+    {
+      "epoch": 0.5675581799235846,
+      "grad_norm": 1.1659549434015535,
+      "learning_rate": 4.154546296321201e-06,
+      "loss": 0.016,
+      "step": 1634
+    },
+    {
+      "epoch": 0.5679055227509552,
+      "grad_norm": 0.8666502913074675,
+      "learning_rate": 4.14900178390382e-06,
+      "loss": 0.0216,
+      "step": 1635
+    },
+    {
+      "epoch": 0.5682528655783258,
+      "grad_norm": 1.5564805764903915,
+      "learning_rate": 4.143458348937421e-06,
+      "loss": 0.021,
+      "step": 1636
+    },
+    {
+      "epoch": 0.5686002084056965,
+      "grad_norm": 1.1023216366479238,
+      "learning_rate": 4.137915998440564e-06,
+      "loss": 0.0182,
+      "step": 1637
+    },
+    {
+      "epoch": 0.5689475512330671,
+      "grad_norm": 0.6297911166518874,
+      "learning_rate": 4.132374739430427e-06,
+      "loss": 0.0238,
+      "step": 1638
+    },
+    {
+      "epoch": 0.5692948940604377,
+      "grad_norm": 1.2566214150953217,
+      "learning_rate": 4.126834578922816e-06,
+      "loss": 0.0165,
+      "step": 1639
+    },
+    {
+      "epoch": 0.5696422368878082,
+      "grad_norm": 0.49215545540415767,
+      "learning_rate": 4.121295523932141e-06,
+      "loss": 0.013,
+      "step": 1640
+    },
+    {
+      "epoch": 0.5699895797151788,
+      "grad_norm": 0.4038473277561727,
+      "learning_rate": 4.115757581471412e-06,
+      "loss": 0.0156,
+      "step": 1641
+    },
+    {
+      "epoch": 0.5703369225425495,
+      "grad_norm": 0.5382862896323736,
+      "learning_rate": 4.110220758552236e-06,
+      "loss": 0.0329,
+      "step": 1642
+    },
+    {
+      "epoch": 0.5706842653699201,
+      "grad_norm": 1.1409766712611276,
+      "learning_rate": 4.104685062184795e-06,
+      "loss": 0.0222,
+      "step": 1643
+    },
+    {
+      "epoch": 0.5710316081972907,
+      "grad_norm": 1.0556759974525016,
+      "learning_rate": 4.0991504993778485e-06,
+      "loss": 0.0267,
+      "step": 1644
+    },
+    {
+      "epoch": 0.5713789510246613,
+      "grad_norm": 0.3881030370366468,
+      "learning_rate": 4.09361707713872e-06,
+      "loss": 0.0194,
+      "step": 1645
+    },
+    {
+      "epoch": 0.571726293852032,
+      "grad_norm": 0.4925901427231959,
+      "learning_rate": 4.088084802473294e-06,
+      "loss": 0.0195,
+      "step": 1646
+    },
+    {
+      "epoch": 0.5720736366794026,
+      "grad_norm": 0.5519961509625639,
+      "learning_rate": 4.0825536823859895e-06,
+      "loss": 0.0244,
+      "step": 1647
+    },
+    {
+      "epoch": 0.5724209795067732,
+      "grad_norm": 0.30715357160720397,
+      "learning_rate": 4.077023723879777e-06,
+      "loss": 0.0107,
+      "step": 1648
+    },
+    {
+      "epoch": 0.5727683223341438,
+      "grad_norm": 0.5494519120842487,
+      "learning_rate": 4.0714949339561495e-06,
+      "loss": 0.0167,
+      "step": 1649
+    },
+    {
+      "epoch": 0.5731156651615145,
+      "grad_norm": 0.3374000317637518,
+      "learning_rate": 4.065967319615123e-06,
+      "loss": 0.0126,
+      "step": 1650
+    },
+    {
+      "epoch": 0.573463007988885,
+      "grad_norm": 0.6301590836558902,
+      "learning_rate": 4.06044088785522e-06,
+      "loss": 0.0122,
+      "step": 1651
+    },
+    {
+      "epoch": 0.5738103508162556,
+      "grad_norm": 1.309711699788591,
+      "learning_rate": 4.054915645673475e-06,
+      "loss": 0.0225,
+      "step": 1652
+    },
+    {
+      "epoch": 0.5741576936436262,
+      "grad_norm": 0.8057778580750404,
+      "learning_rate": 4.049391600065407e-06,
+      "loss": 0.0324,
+      "step": 1653
+    },
+    {
+      "epoch": 0.5745050364709968,
+      "grad_norm": 0.6846173645835267,
+      "learning_rate": 4.043868758025027e-06,
+      "loss": 0.0149,
+      "step": 1654
+    },
+    {
+      "epoch": 0.5748523792983675,
+      "grad_norm": 0.983006587398994,
+      "learning_rate": 4.038347126544816e-06,
+      "loss": 0.0202,
+      "step": 1655
+    },
+    {
+      "epoch": 0.5751997221257381,
+      "grad_norm": 0.4000451176994636,
+      "learning_rate": 4.032826712615727e-06,
+      "loss": 0.0146,
+      "step": 1656
+    },
+    {
+      "epoch": 0.5755470649531087,
+      "grad_norm": 0.6557928230605302,
+      "learning_rate": 4.02730752322717e-06,
+      "loss": 0.0185,
+      "step": 1657
+    },
+    {
+      "epoch": 0.5758944077804793,
+      "grad_norm": 0.39754184151942606,
+      "learning_rate": 4.021789565367007e-06,
+      "loss": 0.0137,
+      "step": 1658
+    },
+    {
+      "epoch": 0.57624175060785,
+      "grad_norm": 0.4884113890090465,
+      "learning_rate": 4.016272846021534e-06,
+      "loss": 0.0227,
+      "step": 1659
+    },
+    {
+      "epoch": 0.5765890934352206,
+      "grad_norm": 0.7745976315929322,
+      "learning_rate": 4.010757372175485e-06,
+      "loss": 0.0208,
+      "step": 1660
+    },
+    {
+      "epoch": 0.5769364362625912,
+      "grad_norm": 0.4156533149774267,
+      "learning_rate": 4.005243150812017e-06,
+      "loss": 0.0202,
+      "step": 1661
+    },
+    {
+      "epoch": 0.5772837790899618,
+      "grad_norm": 0.5057128862814502,
+      "learning_rate": 3.999730188912698e-06,
+      "loss": 0.0164,
+      "step": 1662
+    },
+    {
+      "epoch": 0.5776311219173325,
+      "grad_norm": 0.3684895934447478,
+      "learning_rate": 3.994218493457503e-06,
+      "loss": 0.0181,
+      "step": 1663
+    },
+    {
+      "epoch": 0.577978464744703,
+      "grad_norm": 0.3642952572602709,
+      "learning_rate": 3.988708071424803e-06,
+      "loss": 0.013,
+      "step": 1664
+    },
+    {
+      "epoch": 0.5783258075720736,
+      "grad_norm": 0.9994996067257631,
+      "learning_rate": 3.983198929791357e-06,
+      "loss": 0.0249,
+      "step": 1665
+    },
+    {
+      "epoch": 0.5786731503994442,
+      "grad_norm": 0.3528391599891256,
+      "learning_rate": 3.977691075532305e-06,
+      "loss": 0.0134,
+      "step": 1666
+    },
+    {
+      "epoch": 0.5790204932268148,
+      "grad_norm": 0.4584118521555668,
+      "learning_rate": 3.9721845156211535e-06,
+      "loss": 0.0232,
+      "step": 1667
+    },
+    {
+      "epoch": 0.5793678360541855,
+      "grad_norm": 0.6681777982150439,
+      "learning_rate": 3.966679257029772e-06,
+      "loss": 0.0171,
+      "step": 1668
+    },
+    {
+      "epoch": 0.5797151788815561,
+      "grad_norm": 0.9132998487091103,
+      "learning_rate": 3.961175306728382e-06,
+      "loss": 0.0152,
+      "step": 1669
+    },
+    {
+      "epoch": 0.5800625217089267,
+      "grad_norm": 0.6259597245750973,
+      "learning_rate": 3.955672671685552e-06,
+      "loss": 0.0131,
+      "step": 1670
+    },
+    {
+      "epoch": 0.5804098645362973,
+      "grad_norm": 0.4912882189301569,
+      "learning_rate": 3.950171358868177e-06,
+      "loss": 0.0177,
+      "step": 1671
+    },
+    {
+      "epoch": 0.580757207363668,
+      "grad_norm": 0.9807953706195164,
+      "learning_rate": 3.944671375241485e-06,
+      "loss": 0.0187,
+      "step": 1672
+    },
+    {
+      "epoch": 0.5811045501910386,
+      "grad_norm": 0.9981909862170388,
+      "learning_rate": 3.939172727769021e-06,
+      "loss": 0.0179,
+      "step": 1673
+    },
+    {
+      "epoch": 0.5814518930184092,
+      "grad_norm": 0.6504993573564067,
+      "learning_rate": 3.933675423412636e-06,
+      "loss": 0.0179,
+      "step": 1674
+    },
+    {
+      "epoch": 0.5817992358457797,
+      "grad_norm": 0.513719596446845,
+      "learning_rate": 3.928179469132477e-06,
+      "loss": 0.0189,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5821465786731504,
+      "grad_norm": 0.5682561853770528,
+      "learning_rate": 3.9226848718869905e-06,
+      "loss": 0.0162,
+      "step": 1676
+    },
+    {
+      "epoch": 0.582493921500521,
+      "grad_norm": 0.5510658649823071,
+      "learning_rate": 3.917191638632897e-06,
+      "loss": 0.0182,
+      "step": 1677
+    },
+    {
+      "epoch": 0.5828412643278916,
+      "grad_norm": 0.7127772237528726,
+      "learning_rate": 3.911699776325191e-06,
+      "loss": 0.0204,
+      "step": 1678
+    },
+    {
+      "epoch": 0.5831886071552622,
+      "grad_norm": 0.4103249102772578,
+      "learning_rate": 3.906209291917141e-06,
+      "loss": 0.0121,
+      "step": 1679
+    },
+    {
+      "epoch": 0.5835359499826328,
+      "grad_norm": 0.5473544971789079,
+      "learning_rate": 3.900720192360255e-06,
+      "loss": 0.027,
+      "step": 1680
+    },
+    {
+      "epoch": 0.5838832928100035,
+      "grad_norm": 0.39319317320120634,
+      "learning_rate": 3.895232484604299e-06,
+      "loss": 0.0142,
+      "step": 1681
+    },
+    {
+      "epoch": 0.5842306356373741,
+      "grad_norm": 0.5760042859983425,
+      "learning_rate": 3.889746175597274e-06,
+      "loss": 0.0075,
+      "step": 1682
+    },
+    {
+      "epoch": 0.5845779784647447,
+      "grad_norm": 0.5384592196281841,
+      "learning_rate": 3.884261272285409e-06,
+      "loss": 0.0162,
+      "step": 1683
+    },
+    {
+      "epoch": 0.5849253212921153,
+      "grad_norm": 0.5430426343042051,
+      "learning_rate": 3.8787777816131525e-06,
+      "loss": 0.0132,
+      "step": 1684
+    },
+    {
+      "epoch": 0.585272664119486,
+      "grad_norm": 0.49939162433255296,
+      "learning_rate": 3.873295710523168e-06,
+      "loss": 0.017,
+      "step": 1685
+    },
+    {
+      "epoch": 0.5856200069468566,
+      "grad_norm": 0.6649735504087517,
+      "learning_rate": 3.867815065956319e-06,
+      "loss": 0.0173,
+      "step": 1686
+    },
+    {
+      "epoch": 0.5859673497742272,
+      "grad_norm": 0.878731758013457,
+      "learning_rate": 3.862335854851664e-06,
+      "loss": 0.0225,
+      "step": 1687
+    },
+    {
+      "epoch": 0.5863146926015977,
+      "grad_norm": 0.6320050052238549,
+      "learning_rate": 3.856858084146444e-06,
+      "loss": 0.0216,
+      "step": 1688
+    },
+    {
+      "epoch": 0.5866620354289684,
+      "grad_norm": 0.9414138865126228,
+      "learning_rate": 3.851381760776077e-06,
+      "loss": 0.0211,
+      "step": 1689
+    },
+    {
+      "epoch": 0.587009378256339,
+      "grad_norm": 0.7355116706005304,
+      "learning_rate": 3.845906891674155e-06,
+      "loss": 0.0193,
+      "step": 1690
+    },
+    {
+      "epoch": 0.5873567210837096,
+      "grad_norm": 0.5623242566677557,
+      "learning_rate": 3.8404334837724205e-06,
+      "loss": 0.0201,
+      "step": 1691
+    },
+    {
+      "epoch": 0.5877040639110802,
+      "grad_norm": 0.5982393659307643,
+      "learning_rate": 3.834961544000769e-06,
+      "loss": 0.024,
+      "step": 1692
+    },
+    {
+      "epoch": 0.5880514067384508,
+      "grad_norm": 0.5212721299546837,
+      "learning_rate": 3.8294910792872355e-06,
+      "loss": 0.0194,
+      "step": 1693
+    },
+    {
+      "epoch": 0.5883987495658215,
+      "grad_norm": 0.9699697508233305,
+      "learning_rate": 3.824022096557992e-06,
+      "loss": 0.0246,
+      "step": 1694
+    },
+    {
+      "epoch": 0.5887460923931921,
+      "grad_norm": 0.6506075968369963,
+      "learning_rate": 3.8185546027373325e-06,
+      "loss": 0.0181,
+      "step": 1695
+    },
+    {
+      "epoch": 0.5890934352205627,
+      "grad_norm": 1.1944480293880813,
+      "learning_rate": 3.81308860474766e-06,
+      "loss": 0.0181,
+      "step": 1696
+    },
+    {
+      "epoch": 0.5894407780479333,
+      "grad_norm": 0.8174985452252524,
+      "learning_rate": 3.807624109509491e-06,
+      "loss": 0.0205,
+      "step": 1697
+    },
+    {
+      "epoch": 0.589788120875304,
+      "grad_norm": 0.9297467401132783,
+      "learning_rate": 3.802161123941436e-06,
+      "loss": 0.0181,
+      "step": 1698
+    },
+    {
+      "epoch": 0.5901354637026746,
+      "grad_norm": 1.1350479422662159,
+      "learning_rate": 3.7966996549601968e-06,
+      "loss": 0.022,
+      "step": 1699
+    },
+    {
+      "epoch": 0.5904828065300451,
+      "grad_norm": 1.0512262693137089,
+      "learning_rate": 3.7912397094805508e-06,
+      "loss": 0.0132,
+      "step": 1700
+    },
+    {
+      "epoch": 0.5908301493574157,
+      "grad_norm": 1.2991204215566619,
+      "learning_rate": 3.785781294415349e-06,
+      "loss": 0.0254,
+      "step": 1701
+    },
+    {
+      "epoch": 0.5911774921847864,
+      "grad_norm": 0.5226731278223045,
+      "learning_rate": 3.780324416675504e-06,
+      "loss": 0.0236,
+      "step": 1702
+    },
+    {
+      "epoch": 0.591524835012157,
+      "grad_norm": 0.7465469795767791,
+      "learning_rate": 3.7748690831699858e-06,
+      "loss": 0.0199,
+      "step": 1703
+    },
+    {
+      "epoch": 0.5918721778395276,
+      "grad_norm": 0.39840312715848447,
+      "learning_rate": 3.7694153008058005e-06,
+      "loss": 0.0144,
+      "step": 1704
+    },
+    {
+      "epoch": 0.5922195206668982,
+      "grad_norm": 0.62786992782903,
+      "learning_rate": 3.7639630764879996e-06,
+      "loss": 0.0152,
+      "step": 1705
+    },
+    {
+      "epoch": 0.5925668634942688,
+      "grad_norm": 0.6340042279876545,
+      "learning_rate": 3.7585124171196563e-06,
+      "loss": 0.0208,
+      "step": 1706
+    },
+    {
+      "epoch": 0.5929142063216395,
+      "grad_norm": 0.3158368812851925,
+      "learning_rate": 3.7530633296018664e-06,
+      "loss": 0.0158,
+      "step": 1707
+    },
+    {
+      "epoch": 0.5932615491490101,
+      "grad_norm": 0.6469044082386353,
+      "learning_rate": 3.747615820833729e-06,
+      "loss": 0.0155,
+      "step": 1708
+    },
+    {
+      "epoch": 0.5936088919763807,
+      "grad_norm": 0.5138966576020948,
+      "learning_rate": 3.7421698977123533e-06,
+      "loss": 0.0141,
+      "step": 1709
+    },
+    {
+      "epoch": 0.5939562348037513,
+      "grad_norm": 1.0369760306163973,
+      "learning_rate": 3.736725567132833e-06,
+      "loss": 0.0182,
+      "step": 1710
+    },
+    {
+      "epoch": 0.594303577631122,
+      "grad_norm": 0.8561393743072047,
+      "learning_rate": 3.731282835988252e-06,
+      "loss": 0.0195,
+      "step": 1711
+    },
+    {
+      "epoch": 0.5946509204584925,
+      "grad_norm": 0.3744227619737715,
+      "learning_rate": 3.725841711169662e-06,
+      "loss": 0.0125,
+      "step": 1712
+    },
+    {
+      "epoch": 0.5949982632858631,
+      "grad_norm": 0.44797174308593246,
+      "learning_rate": 3.7204021995660865e-06,
+      "loss": 0.023,
+      "step": 1713
+    },
+    {
+      "epoch": 0.5953456061132337,
+      "grad_norm": 0.5892043090184133,
+      "learning_rate": 3.7149643080645055e-06,
+      "loss": 0.0237,
+      "step": 1714
+    },
+    {
+      "epoch": 0.5956929489406044,
+      "grad_norm": 1.2898844568707646,
+      "learning_rate": 3.7095280435498476e-06,
+      "loss": 0.0208,
+      "step": 1715
+    },
+    {
+      "epoch": 0.596040291767975,
+      "grad_norm": 0.32721676156286766,
+      "learning_rate": 3.7040934129049794e-06,
+      "loss": 0.0121,
+      "step": 1716
+    },
+    {
+      "epoch": 0.5963876345953456,
+      "grad_norm": 0.8377490237817452,
+      "learning_rate": 3.6986604230106993e-06,
+      "loss": 0.0202,
+      "step": 1717
+    },
+    {
+      "epoch": 0.5967349774227162,
+      "grad_norm": 1.3845327772962783,
+      "learning_rate": 3.6932290807457326e-06,
+      "loss": 0.0232,
+      "step": 1718
+    },
+    {
+      "epoch": 0.5970823202500868,
+      "grad_norm": 0.5700167886253666,
+      "learning_rate": 3.6877993929867146e-06,
+      "loss": 0.0121,
+      "step": 1719
+    },
+    {
+      "epoch": 0.5974296630774575,
+      "grad_norm": 0.6239261997221,
+      "learning_rate": 3.6823713666081864e-06,
+      "loss": 0.0122,
+      "step": 1720
+    },
+    {
+      "epoch": 0.5977770059048281,
+      "grad_norm": 0.486573773418933,
+      "learning_rate": 3.676945008482585e-06,
+      "loss": 0.0163,
+      "step": 1721
+    },
+    {
+      "epoch": 0.5981243487321987,
+      "grad_norm": 0.8899160417072958,
+      "learning_rate": 3.671520325480235e-06,
+      "loss": 0.0152,
+      "step": 1722
+    },
+    {
+      "epoch": 0.5984716915595693,
+      "grad_norm": 0.6204522930553988,
+      "learning_rate": 3.6660973244693443e-06,
+      "loss": 0.0136,
+      "step": 1723
+    },
+    {
+      "epoch": 0.59881903438694,
+      "grad_norm": 0.6351473439134602,
+      "learning_rate": 3.6606760123159867e-06,
+      "loss": 0.0124,
+      "step": 1724
+    },
+    {
+      "epoch": 0.5991663772143105,
+      "grad_norm": 0.43393477090853044,
+      "learning_rate": 3.6552563958840994e-06,
+      "loss": 0.011,
+      "step": 1725
+    },
+    {
+      "epoch": 0.5995137200416811,
+      "grad_norm": 0.5715600265285414,
+      "learning_rate": 3.6498384820354693e-06,
+      "loss": 0.0182,
+      "step": 1726
+    },
+    {
+      "epoch": 0.5998610628690517,
+      "grad_norm": 0.8520750202499213,
+      "learning_rate": 3.6444222776297356e-06,
+      "loss": 0.0156,
+      "step": 1727
+    },
+    {
+      "epoch": 0.6002084056964224,
+      "grad_norm": 0.919378604709449,
+      "learning_rate": 3.6390077895243676e-06,
+      "loss": 0.0124,
+      "step": 1728
+    },
+    {
+      "epoch": 0.600555748523793,
+      "grad_norm": 0.5905220548628203,
+      "learning_rate": 3.6335950245746593e-06,
+      "loss": 0.0148,
+      "step": 1729
+    },
+    {
+      "epoch": 0.6009030913511636,
+      "grad_norm": 0.45040394474125944,
+      "learning_rate": 3.6281839896337277e-06,
+      "loss": 0.0191,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6012504341785342,
+      "grad_norm": 0.5316669983658013,
+      "learning_rate": 3.6227746915524964e-06,
+      "loss": 0.016,
+      "step": 1731
+    },
+    {
+      "epoch": 0.6015977770059048,
+      "grad_norm": 0.4609606149512745,
+      "learning_rate": 3.6173671371796946e-06,
+      "loss": 0.0148,
+      "step": 1732
+    },
+    {
+      "epoch": 0.6019451198332755,
+      "grad_norm": 0.41149254163013027,
+      "learning_rate": 3.6119613333618386e-06,
+      "loss": 0.0133,
+      "step": 1733
+    },
+    {
+      "epoch": 0.6022924626606461,
+      "grad_norm": 1.325138871791973,
+      "learning_rate": 3.606557286943229e-06,
+      "loss": 0.0172,
+      "step": 1734
+    },
+    {
+      "epoch": 0.6026398054880167,
+      "grad_norm": 0.44045335177247696,
+      "learning_rate": 3.601155004765943e-06,
+      "loss": 0.009,
+      "step": 1735
+    },
+    {
+      "epoch": 0.6029871483153872,
+      "grad_norm": 0.6696722263208196,
+      "learning_rate": 3.5957544936698272e-06,
+      "loss": 0.018,
+      "step": 1736
+    },
+    {
+      "epoch": 0.603334491142758,
+      "grad_norm": 1.0184366709017578,
+      "learning_rate": 3.5903557604924764e-06,
+      "loss": 0.0137,
+      "step": 1737
+    },
+    {
+      "epoch": 0.6036818339701285,
+      "grad_norm": 0.43044179281096573,
+      "learning_rate": 3.5849588120692446e-06,
+      "loss": 0.0104,
+      "step": 1738
+    },
+    {
+      "epoch": 0.6040291767974991,
+      "grad_norm": 0.3641616392590476,
+      "learning_rate": 3.5795636552332203e-06,
+      "loss": 0.0106,
+      "step": 1739
+    },
+    {
+      "epoch": 0.6043765196248697,
+      "grad_norm": 0.7270321800788714,
+      "learning_rate": 3.5741702968152263e-06,
+      "loss": 0.0262,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6047238624522404,
+      "grad_norm": 0.9048183481948355,
+      "learning_rate": 3.5687787436438044e-06,
+      "loss": 0.0179,
+      "step": 1741
+    },
+    {
+      "epoch": 0.605071205279611,
+      "grad_norm": 1.0702279052136698,
+      "learning_rate": 3.5633890025452162e-06,
+      "loss": 0.022,
+      "step": 1742
+    },
+    {
+      "epoch": 0.6054185481069816,
+      "grad_norm": 2.7184991390092743,
+      "learning_rate": 3.5580010803434254e-06,
+      "loss": 0.0263,
+      "step": 1743
+    },
+    {
+      "epoch": 0.6057658909343522,
+      "grad_norm": 0.5480717716362404,
+      "learning_rate": 3.552614983860096e-06,
+      "loss": 0.025,
+      "step": 1744
+    },
+    {
+      "epoch": 0.6061132337617228,
+      "grad_norm": 0.5845582912450411,
+      "learning_rate": 3.547230719914575e-06,
+      "loss": 0.0121,
+      "step": 1745
+    },
+    {
+      "epoch": 0.6064605765890935,
+      "grad_norm": 0.706154766729172,
+      "learning_rate": 3.541848295323893e-06,
+      "loss": 0.0117,
+      "step": 1746
+    },
+    {
+      "epoch": 0.6068079194164641,
+      "grad_norm": 0.6491617666226633,
+      "learning_rate": 3.536467716902754e-06,
+      "loss": 0.0227,
+      "step": 1747
+    },
+    {
+      "epoch": 0.6071552622438346,
+      "grad_norm": 0.5492141027171614,
+      "learning_rate": 3.5310889914635205e-06,
+      "loss": 0.0269,
+      "step": 1748
+    },
+    {
+      "epoch": 0.6075026050712052,
+      "grad_norm": 0.5439963844863106,
+      "learning_rate": 3.5257121258162092e-06,
+      "loss": 0.0148,
+      "step": 1749
+    },
+    {
+      "epoch": 0.6078499478985759,
+      "grad_norm": 1.3037710079779645,
+      "learning_rate": 3.5203371267684827e-06,
+      "loss": 0.0235,
+      "step": 1750
+    },
+    {
+      "epoch": 0.6081972907259465,
+      "grad_norm": 0.5104769061119815,
+      "learning_rate": 3.5149640011256438e-06,
+      "loss": 0.0198,
+      "step": 1751
+    },
+    {
+      "epoch": 0.6085446335533171,
+      "grad_norm": 0.3616939548509015,
+      "learning_rate": 3.5095927556906193e-06,
+      "loss": 0.0132,
+      "step": 1752
+    },
+    {
+      "epoch": 0.6088919763806877,
+      "grad_norm": 1.4959841416039057,
+      "learning_rate": 3.504223397263955e-06,
+      "loss": 0.0291,
+      "step": 1753
+    },
+    {
+      "epoch": 0.6092393192080584,
+      "grad_norm": 1.3319542700316978,
+      "learning_rate": 3.498855932643811e-06,
+      "loss": 0.0282,
+      "step": 1754
+    },
+    {
+      "epoch": 0.609586662035429,
+      "grad_norm": 0.3634235019022115,
+      "learning_rate": 3.4934903686259445e-06,
+      "loss": 0.0092,
+      "step": 1755
+    },
+    {
+      "epoch": 0.6099340048627996,
+      "grad_norm": 0.26043584442925766,
+      "learning_rate": 3.4881267120037143e-06,
+      "loss": 0.0109,
+      "step": 1756
+    },
+    {
+      "epoch": 0.6102813476901702,
+      "grad_norm": 0.6366073358882772,
+      "learning_rate": 3.4827649695680578e-06,
+      "loss": 0.0285,
+      "step": 1757
+    },
+    {
+      "epoch": 0.6106286905175408,
+      "grad_norm": 1.3257848109030876,
+      "learning_rate": 3.4774051481074885e-06,
+      "loss": 0.0223,
+      "step": 1758
+    },
+    {
+      "epoch": 0.6109760333449115,
+      "grad_norm": 0.4347554670924022,
+      "learning_rate": 3.472047254408091e-06,
+      "loss": 0.0156,
+      "step": 1759
+    },
+    {
+      "epoch": 0.611323376172282,
+      "grad_norm": 0.5238569207098802,
+      "learning_rate": 3.466691295253508e-06,
+      "loss": 0.0184,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6116707189996526,
+      "grad_norm": 0.5864988138134234,
+      "learning_rate": 3.4613372774249355e-06,
+      "loss": 0.0192,
+      "step": 1761
+    },
+    {
+      "epoch": 0.6120180618270232,
+      "grad_norm": 0.4574609279537837,
+      "learning_rate": 3.455985207701105e-06,
+      "loss": 0.0208,
+      "step": 1762
+    },
+    {
+      "epoch": 0.6123654046543939,
+      "grad_norm": 0.6793197760170264,
+      "learning_rate": 3.4506350928582878e-06,
+      "loss": 0.0168,
+      "step": 1763
+    },
+    {
+      "epoch": 0.6127127474817645,
+      "grad_norm": 0.8979077135480431,
+      "learning_rate": 3.4452869396702754e-06,
+      "loss": 0.0193,
+      "step": 1764
+    },
+    {
+      "epoch": 0.6130600903091351,
+      "grad_norm": 1.5966359913446138,
+      "learning_rate": 3.439940754908382e-06,
+      "loss": 0.018,
+      "step": 1765
+    },
+    {
+      "epoch": 0.6134074331365057,
+      "grad_norm": 0.3362644354278385,
+      "learning_rate": 3.4345965453414222e-06,
+      "loss": 0.011,
+      "step": 1766
+    },
+    {
+      "epoch": 0.6137547759638764,
+      "grad_norm": 0.9478737143318928,
+      "learning_rate": 3.429254317735714e-06,
+      "loss": 0.0229,
+      "step": 1767
+    },
+    {
+      "epoch": 0.614102118791247,
+      "grad_norm": 0.5981398739431912,
+      "learning_rate": 3.423914078855064e-06,
+      "loss": 0.014,
+      "step": 1768
+    },
+    {
+      "epoch": 0.6144494616186176,
+      "grad_norm": 1.2492792940989146,
+      "learning_rate": 3.418575835460767e-06,
+      "loss": 0.0159,
+      "step": 1769
+    },
+    {
+      "epoch": 0.6147968044459882,
+      "grad_norm": 0.5085854726417735,
+      "learning_rate": 3.4132395943115803e-06,
+      "loss": 0.0206,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6151441472733588,
+      "grad_norm": 0.3446709459428576,
+      "learning_rate": 3.4079053621637346e-06,
+      "loss": 0.0164,
+      "step": 1771
+    },
+    {
+      "epoch": 0.6154914901007295,
+      "grad_norm": 0.5545946109354595,
+      "learning_rate": 3.402573145770916e-06,
+      "loss": 0.0157,
+      "step": 1772
+    },
+    {
+      "epoch": 0.6158388329281,
+      "grad_norm": 0.9460892604144455,
+      "learning_rate": 3.3972429518842566e-06,
+      "loss": 0.0383,
+      "step": 1773
+    },
+    {
+      "epoch": 0.6161861757554706,
+      "grad_norm": 0.4367290077518178,
+      "learning_rate": 3.3919147872523257e-06,
+      "loss": 0.0159,
+      "step": 1774
+    },
+    {
+      "epoch": 0.6165335185828412,
+      "grad_norm": 0.45854417375935685,
+      "learning_rate": 3.3865886586211285e-06,
+      "loss": 0.0209,
+      "step": 1775
+    },
+    {
+      "epoch": 0.6168808614102119,
+      "grad_norm": 0.6207845739989777,
+      "learning_rate": 3.38126457273409e-06,
+      "loss": 0.0273,
+      "step": 1776
+    },
+    {
+      "epoch": 0.6172282042375825,
+      "grad_norm": 0.8158895384794037,
+      "learning_rate": 3.3759425363320482e-06,
+      "loss": 0.0205,
+      "step": 1777
+    },
+    {
+      "epoch": 0.6175755470649531,
+      "grad_norm": 0.4183124541564,
+      "learning_rate": 3.3706225561532457e-06,
+      "loss": 0.017,
+      "step": 1778
+    },
+    {
+      "epoch": 0.6179228898923237,
+      "grad_norm": 0.5875812018496149,
+      "learning_rate": 3.365304638933322e-06,
+      "loss": 0.0226,
+      "step": 1779
+    },
+    {
+      "epoch": 0.6182702327196944,
+      "grad_norm": 0.4777993091444752,
+      "learning_rate": 3.359988791405309e-06,
+      "loss": 0.0187,
+      "step": 1780
+    },
+    {
+      "epoch": 0.618617575547065,
+      "grad_norm": 0.48444108861589447,
+      "learning_rate": 3.3546750202996136e-06,
+      "loss": 0.0185,
+      "step": 1781
+    },
+    {
+      "epoch": 0.6189649183744356,
+      "grad_norm": 0.4370386372661205,
+      "learning_rate": 3.349363332344013e-06,
+      "loss": 0.0114,
+      "step": 1782
+    },
+    {
+      "epoch": 0.6193122612018062,
+      "grad_norm": 0.3131795277127707,
+      "learning_rate": 3.3440537342636483e-06,
+      "loss": 0.0162,
+      "step": 1783
+    },
+    {
+      "epoch": 0.6196596040291767,
+      "grad_norm": 0.5666886716999282,
+      "learning_rate": 3.338746232781017e-06,
+      "loss": 0.013,
+      "step": 1784
+    },
+    {
+      "epoch": 0.6200069468565474,
+      "grad_norm": 0.30906398669450097,
+      "learning_rate": 3.333440834615961e-06,
+      "loss": 0.019,
+      "step": 1785
+    },
+    {
+      "epoch": 0.620354289683918,
+      "grad_norm": 0.26856637246235177,
+      "learning_rate": 3.3281375464856556e-06,
+      "loss": 0.0118,
+      "step": 1786
+    },
+    {
+      "epoch": 0.6207016325112886,
+      "grad_norm": 0.449006224074791,
+      "learning_rate": 3.322836375104608e-06,
+      "loss": 0.0157,
+      "step": 1787
+    },
+    {
+      "epoch": 0.6210489753386592,
+      "grad_norm": 0.9291779318103394,
+      "learning_rate": 3.3175373271846434e-06,
+      "loss": 0.0165,
+      "step": 1788
+    },
+    {
+      "epoch": 0.6213963181660299,
+      "grad_norm": 0.8633956849713047,
+      "learning_rate": 3.3122404094349037e-06,
+      "loss": 0.0159,
+      "step": 1789
+    },
+    {
+      "epoch": 0.6217436609934005,
+      "grad_norm": 0.43556926498499643,
+      "learning_rate": 3.3069456285618263e-06,
+      "loss": 0.0205,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6220910038207711,
+      "grad_norm": 0.4644429061948247,
+      "learning_rate": 3.3016529912691476e-06,
+      "loss": 0.0174,
+      "step": 1791
+    },
+    {
+      "epoch": 0.6224383466481417,
+      "grad_norm": 0.4928017975644182,
+      "learning_rate": 3.2963625042578875e-06,
+      "loss": 0.0195,
+      "step": 1792
+    },
+    {
+      "epoch": 0.6227856894755124,
+      "grad_norm": 0.39183290751040956,
+      "learning_rate": 3.2910741742263495e-06,
+      "loss": 0.0119,
+      "step": 1793
+    },
+    {
+      "epoch": 0.623133032302883,
+      "grad_norm": 0.5461515569612917,
+      "learning_rate": 3.2857880078700953e-06,
+      "loss": 0.018,
+      "step": 1794
+    },
+    {
+      "epoch": 0.6234803751302536,
+      "grad_norm": 0.7497392589922698,
+      "learning_rate": 3.2805040118819574e-06,
+      "loss": 0.0261,
+      "step": 1795
+    },
+    {
+      "epoch": 0.6238277179576242,
+      "grad_norm": 0.802555297478242,
+      "learning_rate": 3.2752221929520164e-06,
+      "loss": 0.0158,
+      "step": 1796
+    },
+    {
+      "epoch": 0.6241750607849947,
+      "grad_norm": 0.576305432582867,
+      "learning_rate": 3.2699425577675935e-06,
+      "loss": 0.0158,
+      "step": 1797
+    },
+    {
+      "epoch": 0.6245224036123654,
+      "grad_norm": 0.5846916212785859,
+      "learning_rate": 3.2646651130132533e-06,
+      "loss": 0.0201,
+      "step": 1798
+    },
+    {
+      "epoch": 0.624869746439736,
+      "grad_norm": 0.4703989112906414,
+      "learning_rate": 3.2593898653707773e-06,
+      "loss": 0.0153,
+      "step": 1799
+    },
+    {
+      "epoch": 0.6252170892671066,
+      "grad_norm": 0.5825293697611765,
+      "learning_rate": 3.254116821519171e-06,
+      "loss": 0.0127,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6255644320944772,
+      "grad_norm": 0.4387644946692422,
+      "learning_rate": 3.2488459881346483e-06,
+      "loss": 0.02,
+      "step": 1801
+    },
+    {
+      "epoch": 0.6259117749218479,
+      "grad_norm": 0.4193154456369623,
+      "learning_rate": 3.2435773718906284e-06,
+      "loss": 0.0244,
+      "step": 1802
+    },
+    {
+      "epoch": 0.6262591177492185,
+      "grad_norm": 0.39690908782734347,
+      "learning_rate": 3.238310979457713e-06,
+      "loss": 0.0077,
+      "step": 1803
+    },
+    {
+      "epoch": 0.6266064605765891,
+      "grad_norm": 1.233442099164475,
+      "learning_rate": 3.233046817503699e-06,
+      "loss": 0.0186,
+      "step": 1804
+    },
+    {
+      "epoch": 0.6269538034039597,
+      "grad_norm": 1.6141038525452736,
+      "learning_rate": 3.2277848926935528e-06,
+      "loss": 0.0281,
+      "step": 1805
+    },
+    {
+      "epoch": 0.6273011462313304,
+      "grad_norm": 0.7139609495476705,
+      "learning_rate": 3.2225252116894155e-06,
+      "loss": 0.0258,
+      "step": 1806
+    },
+    {
+      "epoch": 0.627648489058701,
+      "grad_norm": 0.7513223031684714,
+      "learning_rate": 3.2172677811505766e-06,
+      "loss": 0.0225,
+      "step": 1807
+    },
+    {
+      "epoch": 0.6279958318860716,
+      "grad_norm": 0.8056062866255712,
+      "learning_rate": 3.2120126077334844e-06,
+      "loss": 0.0158,
+      "step": 1808
+    },
+    {
+      "epoch": 0.6283431747134421,
+      "grad_norm": 0.523846225092398,
+      "learning_rate": 3.2067596980917282e-06,
+      "loss": 0.0189,
+      "step": 1809
+    },
+    {
+      "epoch": 0.6286905175408127,
+      "grad_norm": 0.6441750993488121,
+      "learning_rate": 3.20150905887603e-06,
+      "loss": 0.0282,
+      "step": 1810
+    },
+    {
+      "epoch": 0.6290378603681834,
+      "grad_norm": 1.0001326879988315,
+      "learning_rate": 3.1962606967342356e-06,
+      "loss": 0.0201,
+      "step": 1811
+    },
+    {
+      "epoch": 0.629385203195554,
+      "grad_norm": 0.5018682485066364,
+      "learning_rate": 3.191014618311309e-06,
+      "loss": 0.0192,
+      "step": 1812
+    },
+    {
+      "epoch": 0.6297325460229246,
+      "grad_norm": 0.8969102194509483,
+      "learning_rate": 3.185770830249326e-06,
+      "loss": 0.0176,
+      "step": 1813
+    },
+    {
+      "epoch": 0.6300798888502952,
+      "grad_norm": 0.9231788627357475,
+      "learning_rate": 3.1805293391874604e-06,
+      "loss": 0.0239,
+      "step": 1814
+    },
+    {
+      "epoch": 0.6304272316776659,
+      "grad_norm": 0.4778698280139559,
+      "learning_rate": 3.1752901517619733e-06,
+      "loss": 0.0174,
+      "step": 1815
+    },
+    {
+      "epoch": 0.6307745745050365,
+      "grad_norm": 0.7761414647990466,
+      "learning_rate": 3.1700532746062148e-06,
+      "loss": 0.0237,
+      "step": 1816
+    },
+    {
+      "epoch": 0.6311219173324071,
+      "grad_norm": 0.49952446026830283,
+      "learning_rate": 3.1648187143506095e-06,
+      "loss": 0.0145,
+      "step": 1817
+    },
+    {
+      "epoch": 0.6314692601597777,
+      "grad_norm": 0.6271738528394313,
+      "learning_rate": 3.159586477622647e-06,
+      "loss": 0.0176,
+      "step": 1818
+    },
+    {
+      "epoch": 0.6318166029871484,
+      "grad_norm": 0.2541241193835112,
+      "learning_rate": 3.1543565710468743e-06,
+      "loss": 0.0113,
+      "step": 1819
+    },
+    {
+      "epoch": 0.632163945814519,
+      "grad_norm": 0.6685782936301835,
+      "learning_rate": 3.14912900124489e-06,
+      "loss": 0.0191,
+      "step": 1820
+    },
+    {
+      "epoch": 0.6325112886418895,
+      "grad_norm": 0.376455933499608,
+      "learning_rate": 3.1439037748353316e-06,
+      "loss": 0.0233,
+      "step": 1821
+    },
+    {
+      "epoch": 0.6328586314692601,
+      "grad_norm": 0.42194631315752845,
+      "learning_rate": 3.1386808984338758e-06,
+      "loss": 0.0192,
+      "step": 1822
+    },
+    {
+      "epoch": 0.6332059742966307,
+      "grad_norm": 0.5449588242931194,
+      "learning_rate": 3.1334603786532147e-06,
+      "loss": 0.0254,
+      "step": 1823
+    },
+    {
+      "epoch": 0.6335533171240014,
+      "grad_norm": 0.22995057823733359,
+      "learning_rate": 3.128242222103064e-06,
+      "loss": 0.0109,
+      "step": 1824
+    },
+    {
+      "epoch": 0.633900659951372,
+      "grad_norm": 0.48777742225584697,
+      "learning_rate": 3.123026435390144e-06,
+      "loss": 0.0231,
+      "step": 1825
+    },
+    {
+      "epoch": 0.6342480027787426,
+      "grad_norm": 0.3576773382846993,
+      "learning_rate": 3.117813025118178e-06,
+      "loss": 0.0152,
+      "step": 1826
+    },
+    {
+      "epoch": 0.6345953456061132,
+      "grad_norm": 0.2928349054625997,
+      "learning_rate": 3.112601997887873e-06,
+      "loss": 0.017,
+      "step": 1827
+    },
+    {
+      "epoch": 0.6349426884334839,
+      "grad_norm": 0.5006421572877223,
+      "learning_rate": 3.107393360296927e-06,
+      "loss": 0.0167,
+      "step": 1828
+    },
+    {
+      "epoch": 0.6352900312608545,
+      "grad_norm": 0.4143562291378625,
+      "learning_rate": 3.1021871189400077e-06,
+      "loss": 0.013,
+      "step": 1829
+    },
+    {
+      "epoch": 0.6356373740882251,
+      "grad_norm": 1.0102185381541648,
+      "learning_rate": 3.096983280408754e-06,
+      "loss": 0.032,
+      "step": 1830
+    },
+    {
+      "epoch": 0.6359847169155957,
+      "grad_norm": 0.5038454928470375,
+      "learning_rate": 3.091781851291753e-06,
+      "loss": 0.0195,
+      "step": 1831
+    },
+    {
+      "epoch": 0.6363320597429664,
+      "grad_norm": 0.9531592705539882,
+      "learning_rate": 3.0865828381745515e-06,
+      "loss": 0.0257,
+      "step": 1832
+    },
+    {
+      "epoch": 0.636679402570337,
+      "grad_norm": 0.7522264848981152,
+      "learning_rate": 3.0813862476396323e-06,
+      "loss": 0.0193,
+      "step": 1833
+    },
+    {
+      "epoch": 0.6370267453977075,
+      "grad_norm": 0.4979623790577095,
+      "learning_rate": 3.07619208626641e-06,
+      "loss": 0.018,
+      "step": 1834
+    },
+    {
+      "epoch": 0.6373740882250781,
+      "grad_norm": 1.156688985936038,
+      "learning_rate": 3.0710003606312292e-06,
+      "loss": 0.023,
+      "step": 1835
+    },
+    {
+      "epoch": 0.6377214310524487,
+      "grad_norm": 0.7241396081280072,
+      "learning_rate": 3.065811077307342e-06,
+      "loss": 0.0221,
+      "step": 1836
+    },
+    {
+      "epoch": 0.6380687738798194,
+      "grad_norm": 0.4866234304079524,
+      "learning_rate": 3.060624242864916e-06,
+      "loss": 0.0123,
+      "step": 1837
+    },
+    {
+      "epoch": 0.63841611670719,
+      "grad_norm": 0.4496824373815353,
+      "learning_rate": 3.0554398638710136e-06,
+      "loss": 0.011,
+      "step": 1838
+    },
+    {
+      "epoch": 0.6387634595345606,
+      "grad_norm": 0.7591414273681181,
+      "learning_rate": 3.050257946889594e-06,
+      "loss": 0.0151,
+      "step": 1839
+    },
+    {
+      "epoch": 0.6391108023619312,
+      "grad_norm": 0.37588557206384826,
+      "learning_rate": 3.045078498481491e-06,
+      "loss": 0.0172,
+      "step": 1840
+    },
+    {
+      "epoch": 0.6394581451893019,
+      "grad_norm": 0.5904069516982642,
+      "learning_rate": 3.0399015252044185e-06,
+      "loss": 0.0209,
+      "step": 1841
+    },
+    {
+      "epoch": 0.6398054880166725,
+      "grad_norm": 0.5503945994322635,
+      "learning_rate": 3.0347270336129554e-06,
+      "loss": 0.0161,
+      "step": 1842
+    },
+    {
+      "epoch": 0.6401528308440431,
+      "grad_norm": 0.8287427058376795,
+      "learning_rate": 3.02955503025854e-06,
+      "loss": 0.0231,
+      "step": 1843
+    },
+    {
+      "epoch": 0.6405001736714137,
+      "grad_norm": 0.8132926128652753,
+      "learning_rate": 3.0243855216894557e-06,
+      "loss": 0.0222,
+      "step": 1844
+    },
+    {
+      "epoch": 0.6408475164987844,
+      "grad_norm": 0.34673836499554306,
+      "learning_rate": 3.0192185144508336e-06,
+      "loss": 0.0096,
+      "step": 1845
+    },
+    {
+      "epoch": 0.6411948593261549,
+      "grad_norm": 0.6070961612743545,
+      "learning_rate": 3.0140540150846324e-06,
+      "loss": 0.0255,
+      "step": 1846
+    },
+    {
+      "epoch": 0.6415422021535255,
+      "grad_norm": 0.37708682146770417,
+      "learning_rate": 3.00889203012964e-06,
+      "loss": 0.0142,
+      "step": 1847
+    },
+    {
+      "epoch": 0.6418895449808961,
+      "grad_norm": 0.4761721285878218,
+      "learning_rate": 3.0037325661214555e-06,
+      "loss": 0.0172,
+      "step": 1848
+    },
+    {
+      "epoch": 0.6422368878082667,
+      "grad_norm": 0.43162467961906015,
+      "learning_rate": 2.99857562959249e-06,
+      "loss": 0.014,
+      "step": 1849
+    },
+    {
+      "epoch": 0.6425842306356374,
+      "grad_norm": 0.4418457043261697,
+      "learning_rate": 2.9934212270719555e-06,
+      "loss": 0.0178,
+      "step": 1850
+    },
+    {
+      "epoch": 0.642931573463008,
+      "grad_norm": 0.736952954263188,
+      "learning_rate": 2.988269365085854e-06,
+      "loss": 0.022,
+      "step": 1851
+    },
+    {
+      "epoch": 0.6432789162903786,
+      "grad_norm": 0.5793129124364786,
+      "learning_rate": 2.983120050156969e-06,
+      "loss": 0.0166,
+      "step": 1852
+    },
+    {
+      "epoch": 0.6436262591177492,
+      "grad_norm": 0.4059013512127999,
+      "learning_rate": 2.9779732888048607e-06,
+      "loss": 0.0134,
+      "step": 1853
+    },
+    {
+      "epoch": 0.6439736019451199,
+      "grad_norm": 0.5434418757990271,
+      "learning_rate": 2.9728290875458597e-06,
+      "loss": 0.0184,
+      "step": 1854
+    },
+    {
+      "epoch": 0.6443209447724905,
+      "grad_norm": 0.6080536174329941,
+      "learning_rate": 2.967687452893051e-06,
+      "loss": 0.0151,
+      "step": 1855
+    },
+    {
+      "epoch": 0.6446682875998611,
+      "grad_norm": 0.3794734225409095,
+      "learning_rate": 2.9625483913562696e-06,
+      "loss": 0.0051,
+      "step": 1856
+    },
+    {
+      "epoch": 0.6450156304272316,
+      "grad_norm": 0.4025726954879858,
+      "learning_rate": 2.957411909442095e-06,
+      "loss": 0.0138,
+      "step": 1857
+    },
+    {
+      "epoch": 0.6453629732546023,
+      "grad_norm": 0.9717609808679778,
+      "learning_rate": 2.95227801365384e-06,
+      "loss": 0.0234,
+      "step": 1858
+    },
+    {
+      "epoch": 0.6457103160819729,
+      "grad_norm": 0.9793666448755983,
+      "learning_rate": 2.947146710491545e-06,
+      "loss": 0.0243,
+      "step": 1859
+    },
+    {
+      "epoch": 0.6460576589093435,
+      "grad_norm": 0.6545294963642488,
+      "learning_rate": 2.942018006451961e-06,
+      "loss": 0.0198,
+      "step": 1860
+    },
+    {
+      "epoch": 0.6464050017367141,
+      "grad_norm": 0.543720085501545,
+      "learning_rate": 2.9368919080285574e-06,
+      "loss": 0.0148,
+      "step": 1861
+    },
+    {
+      "epoch": 0.6467523445640847,
+      "grad_norm": 0.5001478214174131,
+      "learning_rate": 2.9317684217114977e-06,
+      "loss": 0.0153,
+      "step": 1862
+    },
+    {
+      "epoch": 0.6470996873914554,
+      "grad_norm": 0.8634338411303856,
+      "learning_rate": 2.9266475539876447e-06,
+      "loss": 0.0153,
+      "step": 1863
+    },
+    {
+      "epoch": 0.647447030218826,
+      "grad_norm": 0.3794776934466785,
+      "learning_rate": 2.921529311340537e-06,
+      "loss": 0.0145,
+      "step": 1864
+    },
+    {
+      "epoch": 0.6477943730461966,
+      "grad_norm": 0.5456181574813935,
+      "learning_rate": 2.916413700250397e-06,
+      "loss": 0.0223,
+      "step": 1865
+    },
+    {
+      "epoch": 0.6481417158735672,
+      "grad_norm": 0.7344763458259643,
+      "learning_rate": 2.9113007271941118e-06,
+      "loss": 0.0263,
+      "step": 1866
+    },
+    {
+      "epoch": 0.6484890587009379,
+      "grad_norm": 0.42076158970673955,
+      "learning_rate": 2.9061903986452323e-06,
+      "loss": 0.0171,
+      "step": 1867
+    },
+    {
+      "epoch": 0.6488364015283085,
+      "grad_norm": 0.5699294311522565,
+      "learning_rate": 2.9010827210739557e-06,
+      "loss": 0.0223,
+      "step": 1868
+    },
+    {
+      "epoch": 0.649183744355679,
+      "grad_norm": 0.7798211226417748,
+      "learning_rate": 2.895977700947124e-06,
+      "loss": 0.0223,
+      "step": 1869
+    },
+    {
+      "epoch": 0.6495310871830496,
+      "grad_norm": 0.6520811194584589,
+      "learning_rate": 2.890875344728218e-06,
+      "loss": 0.0253,
+      "step": 1870
+    },
+    {
+      "epoch": 0.6498784300104203,
+      "grad_norm": 0.5155130215134897,
+      "learning_rate": 2.8857756588773457e-06,
+      "loss": 0.012,
+      "step": 1871
+    },
+    {
+      "epoch": 0.6502257728377909,
+      "grad_norm": 0.6695698436955599,
+      "learning_rate": 2.88067864985123e-06,
+      "loss": 0.0189,
+      "step": 1872
+    },
+    {
+      "epoch": 0.6505731156651615,
+      "grad_norm": 0.5691514228238037,
+      "learning_rate": 2.875584324103205e-06,
+      "loss": 0.022,
+      "step": 1873
+    },
+    {
+      "epoch": 0.6509204584925321,
+      "grad_norm": 0.7629069218607217,
+      "learning_rate": 2.8704926880832117e-06,
+      "loss": 0.0201,
+      "step": 1874
+    },
+    {
+      "epoch": 0.6512678013199027,
+      "grad_norm": 0.4637233958651768,
+      "learning_rate": 2.865403748237784e-06,
+      "loss": 0.0207,
+      "step": 1875
+    },
+    {
+      "epoch": 0.6516151441472734,
+      "grad_norm": 0.702869419072292,
+      "learning_rate": 2.860317511010041e-06,
+      "loss": 0.0173,
+      "step": 1876
+    },
+    {
+      "epoch": 0.651962486974644,
+      "grad_norm": 0.9400016037194756,
+      "learning_rate": 2.855233982839678e-06,
+      "loss": 0.0148,
+      "step": 1877
+    },
+    {
+      "epoch": 0.6523098298020146,
+      "grad_norm": 0.4312318941463103,
+      "learning_rate": 2.8501531701629658e-06,
+      "loss": 0.0156,
+      "step": 1878
+    },
+    {
+      "epoch": 0.6526571726293852,
+      "grad_norm": 0.8618641382406451,
+      "learning_rate": 2.845075079412731e-06,
+      "loss": 0.0163,
+      "step": 1879
+    },
+    {
+      "epoch": 0.6530045154567559,
+      "grad_norm": 0.7664712979039602,
+      "learning_rate": 2.8399997170183625e-06,
+      "loss": 0.0156,
+      "step": 1880
+    },
+    {
+      "epoch": 0.6533518582841265,
+      "grad_norm": 0.9200611024228682,
+      "learning_rate": 2.8349270894057822e-06,
+      "loss": 0.0202,
+      "step": 1881
+    },
+    {
+      "epoch": 0.653699201111497,
+      "grad_norm": 0.509279359509403,
+      "learning_rate": 2.8298572029974624e-06,
+      "loss": 0.0182,
+      "step": 1882
+    },
+    {
+      "epoch": 0.6540465439388676,
+      "grad_norm": 0.6113621614337158,
+      "learning_rate": 2.824790064212396e-06,
+      "loss": 0.019,
+      "step": 1883
+    },
+    {
+      "epoch": 0.6543938867662383,
+      "grad_norm": 0.6185846522906022,
+      "learning_rate": 2.8197256794661023e-06,
+      "loss": 0.0278,
+      "step": 1884
+    },
+    {
+      "epoch": 0.6547412295936089,
+      "grad_norm": 0.36054081111019864,
+      "learning_rate": 2.814664055170609e-06,
+      "loss": 0.0148,
+      "step": 1885
+    },
+    {
+      "epoch": 0.6550885724209795,
+      "grad_norm": 0.46357034653680274,
+      "learning_rate": 2.809605197734454e-06,
+      "loss": 0.0174,
+      "step": 1886
+    },
+    {
+      "epoch": 0.6554359152483501,
+      "grad_norm": 1.1073051215502399,
+      "learning_rate": 2.804549113562667e-06,
+      "loss": 0.0198,
+      "step": 1887
+    },
+    {
+      "epoch": 0.6557832580757207,
+      "grad_norm": 0.2971627242142241,
+      "learning_rate": 2.7994958090567715e-06,
+      "loss": 0.0175,
+      "step": 1888
+    },
+    {
+      "epoch": 0.6561306009030914,
+      "grad_norm": 0.36560530569067073,
+      "learning_rate": 2.7944452906147656e-06,
+      "loss": 0.0134,
+      "step": 1889
+    },
+    {
+      "epoch": 0.656477943730462,
+      "grad_norm": 0.3549139995169762,
+      "learning_rate": 2.7893975646311276e-06,
+      "loss": 0.0132,
+      "step": 1890
+    },
+    {
+      "epoch": 0.6568252865578326,
+      "grad_norm": 0.4732148819461115,
+      "learning_rate": 2.784352637496792e-06,
+      "loss": 0.0148,
+      "step": 1891
+    },
+    {
+      "epoch": 0.6571726293852032,
+      "grad_norm": 0.40360522752171646,
+      "learning_rate": 2.7793105155991584e-06,
+      "loss": 0.0158,
+      "step": 1892
+    },
+    {
+      "epoch": 0.6575199722125739,
+      "grad_norm": 0.7623569339524721,
+      "learning_rate": 2.774271205322066e-06,
+      "loss": 0.0177,
+      "step": 1893
+    },
+    {
+      "epoch": 0.6578673150399444,
+      "grad_norm": 1.367237702513199,
+      "learning_rate": 2.769234713045798e-06,
+      "loss": 0.0228,
+      "step": 1894
+    },
+    {
+      "epoch": 0.658214657867315,
+      "grad_norm": 0.675243175984947,
+      "learning_rate": 2.764201045147071e-06,
+      "loss": 0.0105,
+      "step": 1895
+    },
+    {
+      "epoch": 0.6585620006946856,
+      "grad_norm": 0.9439336995618546,
+      "learning_rate": 2.7591702079990277e-06,
+      "loss": 0.0181,
+      "step": 1896
+    },
+    {
+      "epoch": 0.6589093435220563,
+      "grad_norm": 0.5276873160106954,
+      "learning_rate": 2.754142207971221e-06,
+      "loss": 0.0099,
+      "step": 1897
+    },
+    {
+      "epoch": 0.6592566863494269,
+      "grad_norm": 0.569158156227268,
+      "learning_rate": 2.749117051429612e-06,
+      "loss": 0.0129,
+      "step": 1898
+    },
+    {
+      "epoch": 0.6596040291767975,
+      "grad_norm": 0.5137501939519034,
+      "learning_rate": 2.7440947447365664e-06,
+      "loss": 0.0141,
+      "step": 1899
+    },
+    {
+      "epoch": 0.6599513720041681,
+      "grad_norm": 0.7005662168512808,
+      "learning_rate": 2.739075294250841e-06,
+      "loss": 0.0201,
+      "step": 1900
+    },
+    {
+      "epoch": 0.6602987148315387,
+      "grad_norm": 0.7235707108221842,
+      "learning_rate": 2.7340587063275736e-06,
+      "loss": 0.0208,
+      "step": 1901
+    },
+    {
+      "epoch": 0.6606460576589094,
+      "grad_norm": 0.710616178049963,
+      "learning_rate": 2.7290449873182755e-06,
+      "loss": 0.0149,
+      "step": 1902
+    },
+    {
+      "epoch": 0.66099340048628,
+      "grad_norm": 0.48224009794578837,
+      "learning_rate": 2.7240341435708316e-06,
+      "loss": 0.0099,
+      "step": 1903
+    },
+    {
+      "epoch": 0.6613407433136506,
+      "grad_norm": 0.7252312473237166,
+      "learning_rate": 2.7190261814294873e-06,
+      "loss": 0.0197,
+      "step": 1904
+    },
+    {
+      "epoch": 0.6616880861410211,
+      "grad_norm": 0.4216043473064591,
+      "learning_rate": 2.714021107234831e-06,
+      "loss": 0.012,
+      "step": 1905
+    },
+    {
+      "epoch": 0.6620354289683918,
+      "grad_norm": 0.556375549796574,
+      "learning_rate": 2.7090189273238e-06,
+      "loss": 0.0203,
+      "step": 1906
+    },
+    {
+      "epoch": 0.6623827717957624,
+      "grad_norm": 1.4931977841749278,
+      "learning_rate": 2.7040196480296677e-06,
+      "loss": 0.031,
+      "step": 1907
+    },
+    {
+      "epoch": 0.662730114623133,
+      "grad_norm": 0.6694116899417739,
+      "learning_rate": 2.6990232756820396e-06,
+      "loss": 0.015,
+      "step": 1908
+    },
+    {
+      "epoch": 0.6630774574505036,
+      "grad_norm": 0.7127860652752314,
+      "learning_rate": 2.6940298166068255e-06,
+      "loss": 0.0151,
+      "step": 1909
+    },
+    {
+      "epoch": 0.6634248002778743,
+      "grad_norm": 0.5936061309196379,
+      "learning_rate": 2.6890392771262618e-06,
+      "loss": 0.0169,
+      "step": 1910
+    },
+    {
+      "epoch": 0.6637721431052449,
+      "grad_norm": 0.7367028197758674,
+      "learning_rate": 2.684051663558884e-06,
+      "loss": 0.0209,
+      "step": 1911
+    },
+    {
+      "epoch": 0.6641194859326155,
+      "grad_norm": 0.6177550979223532,
+      "learning_rate": 2.6790669822195202e-06,
+      "loss": 0.0177,
+      "step": 1912
+    },
+    {
+      "epoch": 0.6644668287599861,
+      "grad_norm": 1.1958625864238086,
+      "learning_rate": 2.6740852394192896e-06,
+      "loss": 0.0196,
+      "step": 1913
+    },
+    {
+      "epoch": 0.6648141715873567,
+      "grad_norm": 0.46507928549224564,
+      "learning_rate": 2.6691064414655864e-06,
+      "loss": 0.0123,
+      "step": 1914
+    },
+    {
+      "epoch": 0.6651615144147274,
+      "grad_norm": 0.5722070188703421,
+      "learning_rate": 2.664130594662083e-06,
+      "loss": 0.0224,
+      "step": 1915
+    },
+    {
+      "epoch": 0.665508857242098,
+      "grad_norm": 0.9009029726789275,
+      "learning_rate": 2.6591577053087084e-06,
+      "loss": 0.015,
+      "step": 1916
+    },
+    {
+      "epoch": 0.6658562000694686,
+      "grad_norm": 0.5063091771412535,
+      "learning_rate": 2.6541877797016535e-06,
+      "loss": 0.0138,
+      "step": 1917
+    },
+    {
+      "epoch": 0.6662035428968391,
+      "grad_norm": 0.45072043809201573,
+      "learning_rate": 2.6492208241333494e-06,
+      "loss": 0.016,
+      "step": 1918
+    },
+    {
+      "epoch": 0.6665508857242098,
+      "grad_norm": 0.39804190140708595,
+      "learning_rate": 2.6442568448924754e-06,
+      "loss": 0.0187,
+      "step": 1919
+    },
+    {
+      "epoch": 0.6668982285515804,
+      "grad_norm": 0.3519857413757391,
+      "learning_rate": 2.6392958482639343e-06,
+      "loss": 0.0153,
+      "step": 1920
+    },
+    {
+      "epoch": 0.667245571378951,
+      "grad_norm": 0.7945291965330532,
+      "learning_rate": 2.63433784052886e-06,
+      "loss": 0.0156,
+      "step": 1921
+    },
+    {
+      "epoch": 0.6675929142063216,
+      "grad_norm": 0.49910796071011826,
+      "learning_rate": 2.6293828279645938e-06,
+      "loss": 0.015,
+      "step": 1922
+    },
+    {
+      "epoch": 0.6679402570336923,
+      "grad_norm": 0.6606502012921845,
+      "learning_rate": 2.6244308168446958e-06,
+      "loss": 0.0329,
+      "step": 1923
+    },
+    {
+      "epoch": 0.6682875998610629,
+      "grad_norm": 0.5820565201288449,
+      "learning_rate": 2.6194818134389143e-06,
+      "loss": 0.0199,
+      "step": 1924
+    },
+    {
+      "epoch": 0.6686349426884335,
+      "grad_norm": 0.6077071562564756,
+      "learning_rate": 2.614535824013199e-06,
+      "loss": 0.0212,
+      "step": 1925
+    },
+    {
+      "epoch": 0.6689822855158041,
+      "grad_norm": 0.9648264400971077,
+      "learning_rate": 2.6095928548296773e-06,
+      "loss": 0.0308,
+      "step": 1926
+    },
+    {
+      "epoch": 0.6693296283431747,
+      "grad_norm": 0.6590307822381734,
+      "learning_rate": 2.6046529121466537e-06,
+      "loss": 0.016,
+      "step": 1927
+    },
+    {
+      "epoch": 0.6696769711705454,
+      "grad_norm": 0.647299441184019,
+      "learning_rate": 2.5997160022186028e-06,
+      "loss": 0.0259,
+      "step": 1928
+    },
+    {
+      "epoch": 0.670024313997916,
+      "grad_norm": 0.6135071672946737,
+      "learning_rate": 2.594782131296163e-06,
+      "loss": 0.0087,
+      "step": 1929
+    },
+    {
+      "epoch": 0.6703716568252865,
+      "grad_norm": 1.2610496636585344,
+      "learning_rate": 2.589851305626116e-06,
+      "loss": 0.0239,
+      "step": 1930
+    },
+    {
+      "epoch": 0.6707189996526571,
+      "grad_norm": 0.3954009222925035,
+      "learning_rate": 2.5849235314513923e-06,
+      "loss": 0.0122,
+      "step": 1931
+    },
+    {
+      "epoch": 0.6710663424800278,
+      "grad_norm": 0.5156724723554922,
+      "learning_rate": 2.57999881501106e-06,
+      "loss": 0.0241,
+      "step": 1932
+    },
+    {
+      "epoch": 0.6714136853073984,
+      "grad_norm": 0.8288162912643778,
+      "learning_rate": 2.575077162540318e-06,
+      "loss": 0.0258,
+      "step": 1933
+    },
+    {
+      "epoch": 0.671761028134769,
+      "grad_norm": 0.3963519210942034,
+      "learning_rate": 2.570158580270481e-06,
+      "loss": 0.0156,
+      "step": 1934
+    },
+    {
+      "epoch": 0.6721083709621396,
+      "grad_norm": 0.23269559096942974,
+      "learning_rate": 2.565243074428976e-06,
+      "loss": 0.0104,
+      "step": 1935
+    },
+    {
+      "epoch": 0.6724557137895103,
+      "grad_norm": 0.5076358202743023,
+      "learning_rate": 2.5603306512393387e-06,
+      "loss": 0.0165,
+      "step": 1936
+    },
+    {
+      "epoch": 0.6728030566168809,
+      "grad_norm": 0.4803777245917521,
+      "learning_rate": 2.555421316921203e-06,
+      "loss": 0.0161,
+      "step": 1937
+    },
+    {
+      "epoch": 0.6731503994442515,
+      "grad_norm": 1.7249480773077772,
+      "learning_rate": 2.5505150776902877e-06,
+      "loss": 0.0299,
+      "step": 1938
+    },
+    {
+      "epoch": 0.6734977422716221,
+      "grad_norm": 0.5695623395281586,
+      "learning_rate": 2.5456119397583923e-06,
+      "loss": 0.0226,
+      "step": 1939
+    },
+    {
+      "epoch": 0.6738450850989927,
+      "grad_norm": 0.6090762926972669,
+      "learning_rate": 2.540711909333394e-06,
+      "loss": 0.0138,
+      "step": 1940
+    },
+    {
+      "epoch": 0.6741924279263634,
+      "grad_norm": 0.8830972943432503,
+      "learning_rate": 2.535814992619237e-06,
+      "loss": 0.024,
+      "step": 1941
+    },
+    {
+      "epoch": 0.674539770753734,
+      "grad_norm": 0.672673665772259,
+      "learning_rate": 2.5309211958159135e-06,
+      "loss": 0.0143,
+      "step": 1942
+    },
+    {
+      "epoch": 0.6748871135811045,
+      "grad_norm": 0.44731362978442346,
+      "learning_rate": 2.526030525119475e-06,
+      "loss": 0.0204,
+      "step": 1943
+    },
+    {
+      "epoch": 0.6752344564084751,
+      "grad_norm": 0.3413850862351891,
+      "learning_rate": 2.521142986722014e-06,
+      "loss": 0.0168,
+      "step": 1944
+    },
+    {
+      "epoch": 0.6755817992358458,
+      "grad_norm": 1.356486822592823,
+      "learning_rate": 2.516258586811653e-06,
+      "loss": 0.0181,
+      "step": 1945
+    },
+    {
+      "epoch": 0.6759291420632164,
+      "grad_norm": 0.5590617678902554,
+      "learning_rate": 2.5113773315725407e-06,
+      "loss": 0.0167,
+      "step": 1946
+    },
+    {
+      "epoch": 0.676276484890587,
+      "grad_norm": 1.1480811352959883,
+      "learning_rate": 2.5064992271848504e-06,
+      "loss": 0.014,
+      "step": 1947
+    },
+    {
+      "epoch": 0.6766238277179576,
+      "grad_norm": 1.1960798962121544,
+      "learning_rate": 2.5016242798247623e-06,
+      "loss": 0.0217,
+      "step": 1948
+    },
+    {
+      "epoch": 0.6769711705453283,
+      "grad_norm": 1.4243039228724932,
+      "learning_rate": 2.496752495664457e-06,
+      "loss": 0.0233,
+      "step": 1949
+    },
+    {
+      "epoch": 0.6773185133726989,
+      "grad_norm": 1.0255082402145865,
+      "learning_rate": 2.491883880872115e-06,
+      "loss": 0.0272,
+      "step": 1950
+    },
+    {
+      "epoch": 0.6776658562000695,
+      "grad_norm": 0.7762367778334762,
+      "learning_rate": 2.487018441611899e-06,
+      "loss": 0.019,
+      "step": 1951
+    },
+    {
+      "epoch": 0.6780131990274401,
+      "grad_norm": 0.6073071568906027,
+      "learning_rate": 2.482156184043958e-06,
+      "loss": 0.0259,
+      "step": 1952
+    },
+    {
+      "epoch": 0.6783605418548107,
+      "grad_norm": 0.7255049359634366,
+      "learning_rate": 2.4772971143244033e-06,
+      "loss": 0.0257,
+      "step": 1953
+    },
+    {
+      "epoch": 0.6787078846821814,
+      "grad_norm": 0.8117217423490592,
+      "learning_rate": 2.4724412386053208e-06,
+      "loss": 0.0133,
+      "step": 1954
+    },
+    {
+      "epoch": 0.6790552275095519,
+      "grad_norm": 0.43153608387126113,
+      "learning_rate": 2.4675885630347423e-06,
+      "loss": 0.0144,
+      "step": 1955
+    },
+    {
+      "epoch": 0.6794025703369225,
+      "grad_norm": 0.31393892010490915,
+      "learning_rate": 2.4627390937566566e-06,
+      "loss": 0.0122,
+      "step": 1956
+    },
+    {
+      "epoch": 0.6797499131642931,
+      "grad_norm": 0.49140451557972653,
+      "learning_rate": 2.457892836910985e-06,
+      "loss": 0.0192,
+      "step": 1957
+    },
+    {
+      "epoch": 0.6800972559916638,
+      "grad_norm": 0.32574396229410224,
+      "learning_rate": 2.4530497986335888e-06,
+      "loss": 0.0111,
+      "step": 1958
+    },
+    {
+      "epoch": 0.6804445988190344,
+      "grad_norm": 0.5568119405077808,
+      "learning_rate": 2.4482099850562496e-06,
+      "loss": 0.0199,
+      "step": 1959
+    },
+    {
+      "epoch": 0.680791941646405,
+      "grad_norm": 1.3668016268563767,
+      "learning_rate": 2.4433734023066662e-06,
+      "loss": 0.018,
+      "step": 1960
+    },
+    {
+      "epoch": 0.6811392844737756,
+      "grad_norm": 0.375357049604211,
+      "learning_rate": 2.438540056508449e-06,
+      "loss": 0.009,
+      "step": 1961
+    },
+    {
+      "epoch": 0.6814866273011463,
+      "grad_norm": 0.7359559626951213,
+      "learning_rate": 2.4337099537811114e-06,
+      "loss": 0.0124,
+      "step": 1962
+    },
+    {
+      "epoch": 0.6818339701285169,
+      "grad_norm": 0.6182352327681956,
+      "learning_rate": 2.4288831002400574e-06,
+      "loss": 0.0159,
+      "step": 1963
+    },
+    {
+      "epoch": 0.6821813129558875,
+      "grad_norm": 0.9957988289105886,
+      "learning_rate": 2.4240595019965755e-06,
+      "loss": 0.0226,
+      "step": 1964
+    },
+    {
+      "epoch": 0.682528655783258,
+      "grad_norm": 0.5234908070793383,
+      "learning_rate": 2.4192391651578384e-06,
+      "loss": 0.0176,
+      "step": 1965
+    },
+    {
+      "epoch": 0.6828759986106286,
+      "grad_norm": 0.8440220877656187,
+      "learning_rate": 2.4144220958268883e-06,
+      "loss": 0.0202,
+      "step": 1966
+    },
+    {
+      "epoch": 0.6832233414379993,
+      "grad_norm": 0.5472941032734703,
+      "learning_rate": 2.409608300102627e-06,
+      "loss": 0.0197,
+      "step": 1967
+    },
+    {
+      "epoch": 0.6835706842653699,
+      "grad_norm": 0.37475787591312015,
+      "learning_rate": 2.404797784079811e-06,
+      "loss": 0.0159,
+      "step": 1968
+    },
+    {
+      "epoch": 0.6839180270927405,
+      "grad_norm": 0.41355656518558104,
+      "learning_rate": 2.3999905538490487e-06,
+      "loss": 0.0069,
+      "step": 1969
+    },
+    {
+      "epoch": 0.6842653699201111,
+      "grad_norm": 1.433554613823539,
+      "learning_rate": 2.395186615496789e-06,
+      "loss": 0.0268,
+      "step": 1970
+    },
+    {
+      "epoch": 0.6846127127474818,
+      "grad_norm": 1.1620950677445072,
+      "learning_rate": 2.390385975105308e-06,
+      "loss": 0.0218,
+      "step": 1971
+    },
+    {
+      "epoch": 0.6849600555748524,
+      "grad_norm": 0.37254931419419035,
+      "learning_rate": 2.3855886387527062e-06,
+      "loss": 0.0112,
+      "step": 1972
+    },
+    {
+      "epoch": 0.685307398402223,
+      "grad_norm": 0.4943387705379173,
+      "learning_rate": 2.3807946125129056e-06,
+      "loss": 0.0205,
+      "step": 1973
+    },
+    {
+      "epoch": 0.6856547412295936,
+      "grad_norm": 0.7882856389300709,
+      "learning_rate": 2.3760039024556387e-06,
+      "loss": 0.0198,
+      "step": 1974
+    },
+    {
+      "epoch": 0.6860020840569643,
+      "grad_norm": 1.2217971154250373,
+      "learning_rate": 2.371216514646428e-06,
+      "loss": 0.0135,
+      "step": 1975
+    },
+    {
+      "epoch": 0.6863494268843349,
+      "grad_norm": 0.6619688439348922,
+      "learning_rate": 2.3664324551466007e-06,
+      "loss": 0.0163,
+      "step": 1976
+    },
+    {
+      "epoch": 0.6866967697117055,
+      "grad_norm": 0.4763487225723533,
+      "learning_rate": 2.361651730013269e-06,
+      "loss": 0.0148,
+      "step": 1977
+    },
+    {
+      "epoch": 0.687044112539076,
+      "grad_norm": 0.38275391078177107,
+      "learning_rate": 2.356874345299319e-06,
+      "loss": 0.013,
+      "step": 1978
+    },
+    {
+      "epoch": 0.6873914553664466,
+      "grad_norm": 0.7080298751661178,
+      "learning_rate": 2.3521003070534065e-06,
+      "loss": 0.0205,
+      "step": 1979
+    },
+    {
+      "epoch": 0.6877387981938173,
+      "grad_norm": 0.9866082614707886,
+      "learning_rate": 2.347329621319957e-06,
+      "loss": 0.0203,
+      "step": 1980
+    },
+    {
+      "epoch": 0.6880861410211879,
+      "grad_norm": 0.5240609641902144,
+      "learning_rate": 2.3425622941391485e-06,
+      "loss": 0.0144,
+      "step": 1981
+    },
+    {
+      "epoch": 0.6884334838485585,
+      "grad_norm": 0.5375077106046096,
+      "learning_rate": 2.3377983315469045e-06,
+      "loss": 0.0151,
+      "step": 1982
+    },
+    {
+      "epoch": 0.6887808266759291,
+      "grad_norm": 0.9763971466094741,
+      "learning_rate": 2.3330377395748878e-06,
+      "loss": 0.0168,
+      "step": 1983
+    },
+    {
+      "epoch": 0.6891281695032998,
+      "grad_norm": 0.6777867916323312,
+      "learning_rate": 2.328280524250498e-06,
+      "loss": 0.0196,
+      "step": 1984
+    },
+    {
+      "epoch": 0.6894755123306704,
+      "grad_norm": 0.6717483439928033,
+      "learning_rate": 2.3235266915968586e-06,
+      "loss": 0.0297,
+      "step": 1985
+    },
+    {
+      "epoch": 0.689822855158041,
+      "grad_norm": 0.659428567653895,
+      "learning_rate": 2.3187762476328086e-06,
+      "loss": 0.0154,
+      "step": 1986
+    },
+    {
+      "epoch": 0.6901701979854116,
+      "grad_norm": 0.48768769720756106,
+      "learning_rate": 2.3140291983728936e-06,
+      "loss": 0.0175,
+      "step": 1987
+    },
+    {
+      "epoch": 0.6905175408127823,
+      "grad_norm": 0.6148456012864478,
+      "learning_rate": 2.3092855498273674e-06,
+      "loss": 0.0177,
+      "step": 1988
+    },
+    {
+      "epoch": 0.6908648836401529,
+      "grad_norm": 0.5297818902205734,
+      "learning_rate": 2.3045453080021775e-06,
+      "loss": 0.0134,
+      "step": 1989
+    },
+    {
+      "epoch": 0.6912122264675234,
+      "grad_norm": 0.6794761025303996,
+      "learning_rate": 2.2998084788989514e-06,
+      "loss": 0.0148,
+      "step": 1990
+    },
+    {
+      "epoch": 0.691559569294894,
+      "grad_norm": 0.4063679508309279,
+      "learning_rate": 2.2950750685150045e-06,
+      "loss": 0.0087,
+      "step": 1991
+    },
+    {
+      "epoch": 0.6919069121222646,
+      "grad_norm": 0.6095392803855911,
+      "learning_rate": 2.290345082843318e-06,
+      "loss": 0.0164,
+      "step": 1992
+    },
+    {
+      "epoch": 0.6922542549496353,
+      "grad_norm": 0.7441531444252605,
+      "learning_rate": 2.285618527872537e-06,
+      "loss": 0.0269,
+      "step": 1993
+    },
+    {
+      "epoch": 0.6926015977770059,
+      "grad_norm": 0.5717566310228904,
+      "learning_rate": 2.2808954095869653e-06,
+      "loss": 0.0131,
+      "step": 1994
+    },
+    {
+      "epoch": 0.6929489406043765,
+      "grad_norm": 0.6596567441963315,
+      "learning_rate": 2.2761757339665576e-06,
+      "loss": 0.0204,
+      "step": 1995
+    },
+    {
+      "epoch": 0.6932962834317471,
+      "grad_norm": 0.608634510397694,
+      "learning_rate": 2.2714595069869044e-06,
+      "loss": 0.0201,
+      "step": 1996
+    },
+    {
+      "epoch": 0.6936436262591178,
+      "grad_norm": 0.8865789858612154,
+      "learning_rate": 2.2667467346192325e-06,
+      "loss": 0.012,
+      "step": 1997
+    },
+    {
+      "epoch": 0.6939909690864884,
+      "grad_norm": 0.7362361397592723,
+      "learning_rate": 2.2620374228303944e-06,
+      "loss": 0.0193,
+      "step": 1998
+    },
+    {
+      "epoch": 0.694338311913859,
+      "grad_norm": 0.42850603654943825,
+      "learning_rate": 2.2573315775828655e-06,
+      "loss": 0.016,
+      "step": 1999
+    },
+    {
+      "epoch": 0.6946856547412296,
+      "grad_norm": 0.46854751088934754,
+      "learning_rate": 2.2526292048347246e-06,
+      "loss": 0.0184,
+      "step": 2000
+    },
+    {
+      "epoch": 0.6950329975686003,
+      "grad_norm": 0.6941328111513891,
+      "learning_rate": 2.2479303105396576e-06,
+      "loss": 0.0252,
+      "step": 2001
+    },
+    {
+      "epoch": 0.6953803403959709,
+      "grad_norm": 0.5761073352376276,
+      "learning_rate": 2.2432349006469468e-06,
+      "loss": 0.0282,
+      "step": 2002
+    },
+    {
+      "epoch": 0.6957276832233414,
+      "grad_norm": 0.5163738104718749,
+      "learning_rate": 2.2385429811014654e-06,
+      "loss": 0.0175,
+      "step": 2003
+    },
+    {
+      "epoch": 0.696075026050712,
+      "grad_norm": 0.5317310005945695,
+      "learning_rate": 2.2338545578436623e-06,
+      "loss": 0.0251,
+      "step": 2004
+    },
+    {
+      "epoch": 0.6964223688780826,
+      "grad_norm": 0.407307788912211,
+      "learning_rate": 2.2291696368095595e-06,
+      "loss": 0.0191,
+      "step": 2005
+    },
+    {
+      "epoch": 0.6967697117054533,
+      "grad_norm": 0.9234682659843976,
+      "learning_rate": 2.2244882239307497e-06,
+      "loss": 0.0277,
+      "step": 2006
+    },
+    {
+      "epoch": 0.6971170545328239,
+      "grad_norm": 0.4498944617872236,
+      "learning_rate": 2.2198103251343856e-06,
+      "loss": 0.0153,
+      "step": 2007
+    },
+    {
+      "epoch": 0.6974643973601945,
+      "grad_norm": 0.3445137017847305,
+      "learning_rate": 2.215135946343159e-06,
+      "loss": 0.0151,
+      "step": 2008
+    },
+    {
+      "epoch": 0.6978117401875651,
+      "grad_norm": 0.4550738434401644,
+      "learning_rate": 2.2104650934753157e-06,
+      "loss": 0.0198,
+      "step": 2009
+    },
+    {
+      "epoch": 0.6981590830149358,
+      "grad_norm": 0.4009917053292384,
+      "learning_rate": 2.2057977724446365e-06,
+      "loss": 0.0181,
+      "step": 2010
+    },
+    {
+      "epoch": 0.6985064258423064,
+      "grad_norm": 0.7002896810570791,
+      "learning_rate": 2.201133989160427e-06,
+      "loss": 0.0177,
+      "step": 2011
+    },
+    {
+      "epoch": 0.698853768669677,
+      "grad_norm": 0.6950029004017243,
+      "learning_rate": 2.1964737495275122e-06,
+      "loss": 0.0182,
+      "step": 2012
+    },
+    {
+      "epoch": 0.6992011114970476,
+      "grad_norm": 0.4199036421278682,
+      "learning_rate": 2.191817059446236e-06,
+      "loss": 0.0165,
+      "step": 2013
+    },
+    {
+      "epoch": 0.6995484543244183,
+      "grad_norm": 0.5045888082776578,
+      "learning_rate": 2.1871639248124465e-06,
+      "loss": 0.018,
+      "step": 2014
+    },
+    {
+      "epoch": 0.6998957971517888,
+      "grad_norm": 0.6958614459354281,
+      "learning_rate": 2.182514351517488e-06,
+      "loss": 0.0176,
+      "step": 2015
+    },
+    {
+      "epoch": 0.7002431399791594,
+      "grad_norm": 0.46312645884075154,
+      "learning_rate": 2.1778683454481946e-06,
+      "loss": 0.0202,
+      "step": 2016
+    },
+    {
+      "epoch": 0.70059048280653,
+      "grad_norm": 0.4799742697736775,
+      "learning_rate": 2.1732259124868883e-06,
+      "loss": 0.015,
+      "step": 2017
+    },
+    {
+      "epoch": 0.7009378256339006,
+      "grad_norm": 0.33079209530907566,
+      "learning_rate": 2.1685870585113666e-06,
+      "loss": 0.0107,
+      "step": 2018
+    },
+    {
+      "epoch": 0.7012851684612713,
+      "grad_norm": 0.5741815050233295,
+      "learning_rate": 2.1639517893948926e-06,
+      "loss": 0.0168,
+      "step": 2019
+    },
+    {
+      "epoch": 0.7016325112886419,
+      "grad_norm": 0.8688246670359403,
+      "learning_rate": 2.1593201110061906e-06,
+      "loss": 0.0179,
+      "step": 2020
+    },
+    {
+      "epoch": 0.7019798541160125,
+      "grad_norm": 0.47477951851972594,
+      "learning_rate": 2.154692029209442e-06,
+      "loss": 0.0131,
+      "step": 2021
+    },
+    {
+      "epoch": 0.7023271969433831,
+      "grad_norm": 1.249034719322415,
+      "learning_rate": 2.1500675498642746e-06,
+      "loss": 0.0237,
+      "step": 2022
+    },
+    {
+      "epoch": 0.7026745397707538,
+      "grad_norm": 1.4687157493254421,
+      "learning_rate": 2.145446678825751e-06,
+      "loss": 0.0217,
+      "step": 2023
+    },
+    {
+      "epoch": 0.7030218825981244,
+      "grad_norm": 0.38679253812681325,
+      "learning_rate": 2.140829421944367e-06,
+      "loss": 0.0181,
+      "step": 2024
+    },
+    {
+      "epoch": 0.703369225425495,
+      "grad_norm": 0.4319877902351082,
+      "learning_rate": 2.136215785066046e-06,
+      "loss": 0.0077,
+      "step": 2025
+    },
+    {
+      "epoch": 0.7037165682528655,
+      "grad_norm": 1.2606438198673813,
+      "learning_rate": 2.1316057740321212e-06,
+      "loss": 0.0298,
+      "step": 2026
+    },
+    {
+      "epoch": 0.7040639110802361,
+      "grad_norm": 0.7241622040775874,
+      "learning_rate": 2.1269993946793414e-06,
+      "loss": 0.0127,
+      "step": 2027
+    },
+    {
+      "epoch": 0.7044112539076068,
+      "grad_norm": 0.7645077476228309,
+      "learning_rate": 2.1223966528398577e-06,
+      "loss": 0.0249,
+      "step": 2028
+    },
+    {
+      "epoch": 0.7047585967349774,
+      "grad_norm": 0.6348186548588091,
+      "learning_rate": 2.11779755434121e-06,
+      "loss": 0.0142,
+      "step": 2029
+    },
+    {
+      "epoch": 0.705105939562348,
+      "grad_norm": 0.5434630709530021,
+      "learning_rate": 2.113202105006327e-06,
+      "loss": 0.0148,
+      "step": 2030
+    },
+    {
+      "epoch": 0.7054532823897186,
+      "grad_norm": 0.7784838597282087,
+      "learning_rate": 2.1086103106535214e-06,
+      "loss": 0.0197,
+      "step": 2031
+    },
+    {
+      "epoch": 0.7058006252170893,
+      "grad_norm": 1.0412089267934344,
+      "learning_rate": 2.104022177096477e-06,
+      "loss": 0.0207,
+      "step": 2032
+    },
+    {
+      "epoch": 0.7061479680444599,
+      "grad_norm": 0.5527143676070977,
+      "learning_rate": 2.0994377101442387e-06,
+      "loss": 0.0183,
+      "step": 2033
+    },
+    {
+      "epoch": 0.7064953108718305,
+      "grad_norm": 1.1085673643586902,
+      "learning_rate": 2.09485691560121e-06,
+      "loss": 0.0153,
+      "step": 2034
+    },
+    {
+      "epoch": 0.7068426536992011,
+      "grad_norm": 0.3690027590773899,
+      "learning_rate": 2.0902797992671485e-06,
+      "loss": 0.009,
+      "step": 2035
+    },
+    {
+      "epoch": 0.7071899965265718,
+      "grad_norm": 0.4620521220600708,
+      "learning_rate": 2.0857063669371545e-06,
+      "loss": 0.0117,
+      "step": 2036
+    },
+    {
+      "epoch": 0.7075373393539424,
+      "grad_norm": 0.7530837284547335,
+      "learning_rate": 2.081136624401661e-06,
+      "loss": 0.0109,
+      "step": 2037
+    },
+    {
+      "epoch": 0.707884682181313,
+      "grad_norm": 1.090635412666822,
+      "learning_rate": 2.076570577446428e-06,
+      "loss": 0.0197,
+      "step": 2038
+    },
+    {
+      "epoch": 0.7082320250086835,
+      "grad_norm": 0.5421154350298361,
+      "learning_rate": 2.0720082318525405e-06,
+      "loss": 0.021,
+      "step": 2039
+    },
+    {
+      "epoch": 0.7085793678360541,
+      "grad_norm": 0.675053408269602,
+      "learning_rate": 2.0674495933963997e-06,
+      "loss": 0.021,
+      "step": 2040
+    },
+    {
+      "epoch": 0.7089267106634248,
+      "grad_norm": 0.6491462598285759,
+      "learning_rate": 2.062894667849702e-06,
+      "loss": 0.0332,
+      "step": 2041
+    },
+    {
+      "epoch": 0.7092740534907954,
+      "grad_norm": 0.4073609530744494,
+      "learning_rate": 2.058343460979454e-06,
+      "loss": 0.0158,
+      "step": 2042
+    },
+    {
+      "epoch": 0.709621396318166,
+      "grad_norm": 0.6131847890148999,
+      "learning_rate": 2.0537959785479517e-06,
+      "loss": 0.0179,
+      "step": 2043
+    },
+    {
+      "epoch": 0.7099687391455366,
+      "grad_norm": 0.7676568413480472,
+      "learning_rate": 2.049252226312772e-06,
+      "loss": 0.0231,
+      "step": 2044
+    },
+    {
+      "epoch": 0.7103160819729073,
+      "grad_norm": 0.6639394977446624,
+      "learning_rate": 2.04471221002677e-06,
+      "loss": 0.0255,
+      "step": 2045
+    },
+    {
+      "epoch": 0.7106634248002779,
+      "grad_norm": 0.6596012575548285,
+      "learning_rate": 2.0401759354380728e-06,
+      "loss": 0.0197,
+      "step": 2046
+    },
+    {
+      "epoch": 0.7110107676276485,
+      "grad_norm": 0.7895955447252033,
+      "learning_rate": 2.035643408290071e-06,
+      "loss": 0.0258,
+      "step": 2047
+    },
+    {
+      "epoch": 0.7113581104550191,
+      "grad_norm": 0.6477495390019533,
+      "learning_rate": 2.0311146343214073e-06,
+      "loss": 0.0102,
+      "step": 2048
+    },
+    {
+      "epoch": 0.7117054532823898,
+      "grad_norm": 1.064483048484202,
+      "learning_rate": 2.0265896192659717e-06,
+      "loss": 0.0227,
+      "step": 2049
+    },
+    {
+      "epoch": 0.7120527961097604,
+      "grad_norm": 0.6008333641221011,
+      "learning_rate": 2.0220683688528988e-06,
+      "loss": 0.0128,
+      "step": 2050
+    },
+    {
+      "epoch": 0.712400138937131,
+      "grad_norm": 0.7586223473422381,
+      "learning_rate": 2.0175508888065563e-06,
+      "loss": 0.0207,
+      "step": 2051
+    },
+    {
+      "epoch": 0.7127474817645015,
+      "grad_norm": 0.6734658920557806,
+      "learning_rate": 2.013037184846537e-06,
+      "loss": 0.0146,
+      "step": 2052
+    },
+    {
+      "epoch": 0.7130948245918721,
+      "grad_norm": 0.43786239417939005,
+      "learning_rate": 2.0085272626876496e-06,
+      "loss": 0.025,
+      "step": 2053
+    },
+    {
+      "epoch": 0.7134421674192428,
+      "grad_norm": 0.43477244650995783,
+      "learning_rate": 2.00402112803992e-06,
+      "loss": 0.0224,
+      "step": 2054
+    },
+    {
+      "epoch": 0.7137895102466134,
+      "grad_norm": 0.738487309514483,
+      "learning_rate": 1.9995187866085786e-06,
+      "loss": 0.0203,
+      "step": 2055
+    },
+    {
+      "epoch": 0.714136853073984,
+      "grad_norm": 0.370434091066062,
+      "learning_rate": 1.9950202440940496e-06,
+      "loss": 0.0159,
+      "step": 2056
+    },
+    {
+      "epoch": 0.7144841959013546,
+      "grad_norm": 0.6924851015465208,
+      "learning_rate": 1.9905255061919464e-06,
+      "loss": 0.0131,
+      "step": 2057
+    },
+    {
+      "epoch": 0.7148315387287253,
+      "grad_norm": 0.5673188746010913,
+      "learning_rate": 1.9860345785930726e-06,
+      "loss": 0.0172,
+      "step": 2058
+    },
+    {
+      "epoch": 0.7151788815560959,
+      "grad_norm": 0.3438440626602846,
+      "learning_rate": 1.9815474669833985e-06,
+      "loss": 0.0145,
+      "step": 2059
+    },
+    {
+      "epoch": 0.7155262243834665,
+      "grad_norm": 0.45539417984892605,
+      "learning_rate": 1.977064177044071e-06,
+      "loss": 0.0183,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7158735672108371,
+      "grad_norm": 1.484472402316562,
+      "learning_rate": 1.972584714451392e-06,
+      "loss": 0.0221,
+      "step": 2061
+    },
+    {
+      "epoch": 0.7162209100382078,
+      "grad_norm": 0.5404845318844554,
+      "learning_rate": 1.9681090848768237e-06,
+      "loss": 0.0195,
+      "step": 2062
+    },
+    {
+      "epoch": 0.7165682528655783,
+      "grad_norm": 0.47909379000212127,
+      "learning_rate": 1.9636372939869677e-06,
+      "loss": 0.0169,
+      "step": 2063
+    },
+    {
+      "epoch": 0.7169155956929489,
+      "grad_norm": 0.5622427498011785,
+      "learning_rate": 1.9591693474435735e-06,
+      "loss": 0.0304,
+      "step": 2064
+    },
+    {
+      "epoch": 0.7172629385203195,
+      "grad_norm": 0.7467315115677308,
+      "learning_rate": 1.9547052509035164e-06,
+      "loss": 0.0165,
+      "step": 2065
+    },
+    {
+      "epoch": 0.7176102813476901,
+      "grad_norm": 0.5056808774501013,
+      "learning_rate": 1.9502450100188037e-06,
+      "loss": 0.0215,
+      "step": 2066
+    },
+    {
+      "epoch": 0.7179576241750608,
+      "grad_norm": 1.4391119548200895,
+      "learning_rate": 1.9457886304365533e-06,
+      "loss": 0.0231,
+      "step": 2067
+    },
+    {
+      "epoch": 0.7183049670024314,
+      "grad_norm": 0.6476533862493838,
+      "learning_rate": 1.9413361177990015e-06,
+      "loss": 0.0241,
+      "step": 2068
+    },
+    {
+      "epoch": 0.718652309829802,
+      "grad_norm": 0.530719837422982,
+      "learning_rate": 1.9368874777434864e-06,
+      "loss": 0.0199,
+      "step": 2069
+    },
+    {
+      "epoch": 0.7189996526571726,
+      "grad_norm": 0.519016068544451,
+      "learning_rate": 1.932442715902441e-06,
+      "loss": 0.0201,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7193469954845433,
+      "grad_norm": 0.704754836241455,
+      "learning_rate": 1.9280018379033884e-06,
+      "loss": 0.0292,
+      "step": 2071
+    },
+    {
+      "epoch": 0.7196943383119139,
+      "grad_norm": 0.5687572422985191,
+      "learning_rate": 1.923564849368936e-06,
+      "loss": 0.0214,
+      "step": 2072
+    },
+    {
+      "epoch": 0.7200416811392845,
+      "grad_norm": 0.3734820838812652,
+      "learning_rate": 1.919131755916771e-06,
+      "loss": 0.0187,
+      "step": 2073
+    },
+    {
+      "epoch": 0.720389023966655,
+      "grad_norm": 0.3791262946433928,
+      "learning_rate": 1.9147025631596362e-06,
+      "loss": 0.02,
+      "step": 2074
+    },
+    {
+      "epoch": 0.7207363667940258,
+      "grad_norm": 0.4220291272178966,
+      "learning_rate": 1.9102772767053467e-06,
+      "loss": 0.0249,
+      "step": 2075
+    },
+    {
+      "epoch": 0.7210837096213963,
+      "grad_norm": 0.5391949275021815,
+      "learning_rate": 1.9058559021567718e-06,
+      "loss": 0.0192,
+      "step": 2076
+    },
+    {
+      "epoch": 0.7214310524487669,
+      "grad_norm": 1.4908074892631773,
+      "learning_rate": 1.9014384451118229e-06,
+      "loss": 0.0206,
+      "step": 2077
+    },
+    {
+      "epoch": 0.7217783952761375,
+      "grad_norm": 0.711125711354139,
+      "learning_rate": 1.897024911163451e-06,
+      "loss": 0.0247,
+      "step": 2078
+    },
+    {
+      "epoch": 0.7221257381035081,
+      "grad_norm": 1.0280025482928667,
+      "learning_rate": 1.892615305899645e-06,
+      "loss": 0.027,
+      "step": 2079
+    },
+    {
+      "epoch": 0.7224730809308788,
+      "grad_norm": 0.6833450502355668,
+      "learning_rate": 1.8882096349034184e-06,
+      "loss": 0.0264,
+      "step": 2080
+    },
+    {
+      "epoch": 0.7228204237582494,
+      "grad_norm": 0.7449238137674074,
+      "learning_rate": 1.8838079037528012e-06,
+      "loss": 0.0226,
+      "step": 2081
+    },
+    {
+      "epoch": 0.72316776658562,
+      "grad_norm": 0.745751826998084,
+      "learning_rate": 1.879410118020834e-06,
+      "loss": 0.0306,
+      "step": 2082
+    },
+    {
+      "epoch": 0.7235151094129906,
+      "grad_norm": 0.3996052258819095,
+      "learning_rate": 1.8750162832755669e-06,
+      "loss": 0.0172,
+      "step": 2083
+    },
+    {
+      "epoch": 0.7238624522403613,
+      "grad_norm": 0.9447949016982862,
+      "learning_rate": 1.870626405080046e-06,
+      "loss": 0.02,
+      "step": 2084
+    },
+    {
+      "epoch": 0.7242097950677319,
+      "grad_norm": 0.7392285624719003,
+      "learning_rate": 1.8662404889923058e-06,
+      "loss": 0.0147,
+      "step": 2085
+    },
+    {
+      "epoch": 0.7245571378951025,
+      "grad_norm": 0.506159285212194,
+      "learning_rate": 1.8618585405653639e-06,
+      "loss": 0.0188,
+      "step": 2086
+    },
+    {
+      "epoch": 0.724904480722473,
+      "grad_norm": 0.5116325753968864,
+      "learning_rate": 1.8574805653472178e-06,
+      "loss": 0.0143,
+      "step": 2087
+    },
+    {
+      "epoch": 0.7252518235498437,
+      "grad_norm": 0.7618959377624293,
+      "learning_rate": 1.8531065688808346e-06,
+      "loss": 0.0205,
+      "step": 2088
+    },
+    {
+      "epoch": 0.7255991663772143,
+      "grad_norm": 0.6293199274360505,
+      "learning_rate": 1.848736556704141e-06,
+      "loss": 0.0122,
+      "step": 2089
+    },
+    {
+      "epoch": 0.7259465092045849,
+      "grad_norm": 0.9005126434840245,
+      "learning_rate": 1.8443705343500185e-06,
+      "loss": 0.0146,
+      "step": 2090
+    },
+    {
+      "epoch": 0.7262938520319555,
+      "grad_norm": 1.1648440673296412,
+      "learning_rate": 1.840008507346302e-06,
+      "loss": 0.0241,
+      "step": 2091
+    },
+    {
+      "epoch": 0.7266411948593261,
+      "grad_norm": 0.5855114659510036,
+      "learning_rate": 1.8356504812157623e-06,
+      "loss": 0.0207,
+      "step": 2092
+    },
+    {
+      "epoch": 0.7269885376866968,
+      "grad_norm": 0.5068661065981903,
+      "learning_rate": 1.831296461476109e-06,
+      "loss": 0.0146,
+      "step": 2093
+    },
+    {
+      "epoch": 0.7273358805140674,
+      "grad_norm": 0.5837482870866362,
+      "learning_rate": 1.826946453639976e-06,
+      "loss": 0.0163,
+      "step": 2094
+    },
+    {
+      "epoch": 0.727683223341438,
+      "grad_norm": 0.6908638053788853,
+      "learning_rate": 1.822600463214922e-06,
+      "loss": 0.0189,
+      "step": 2095
+    },
+    {
+      "epoch": 0.7280305661688086,
+      "grad_norm": 0.6943496634239659,
+      "learning_rate": 1.818258495703412e-06,
+      "loss": 0.0207,
+      "step": 2096
+    },
+    {
+      "epoch": 0.7283779089961793,
+      "grad_norm": 0.8592869838174363,
+      "learning_rate": 1.813920556602826e-06,
+      "loss": 0.0214,
+      "step": 2097
+    },
+    {
+      "epoch": 0.7287252518235499,
+      "grad_norm": 0.9767864211568215,
+      "learning_rate": 1.8095866514054372e-06,
+      "loss": 0.0151,
+      "step": 2098
+    },
+    {
+      "epoch": 0.7290725946509204,
+      "grad_norm": 0.24915866573453857,
+      "learning_rate": 1.805256785598416e-06,
+      "loss": 0.0104,
+      "step": 2099
+    },
+    {
+      "epoch": 0.729419937478291,
+      "grad_norm": 0.34781878859312965,
+      "learning_rate": 1.8009309646638128e-06,
+      "loss": 0.0116,
+      "step": 2100
+    },
+    {
+      "epoch": 0.7297672803056617,
+      "grad_norm": 0.6277095397756866,
+      "learning_rate": 1.7966091940785653e-06,
+      "loss": 0.016,
+      "step": 2101
+    },
+    {
+      "epoch": 0.7301146231330323,
+      "grad_norm": 0.7019890305202732,
+      "learning_rate": 1.792291479314473e-06,
+      "loss": 0.0188,
+      "step": 2102
+    },
+    {
+      "epoch": 0.7304619659604029,
+      "grad_norm": 0.4558680260256953,
+      "learning_rate": 1.7879778258382103e-06,
+      "loss": 0.0153,
+      "step": 2103
+    },
+    {
+      "epoch": 0.7308093087877735,
+      "grad_norm": 0.5209152944828331,
+      "learning_rate": 1.7836682391113002e-06,
+      "loss": 0.0205,
+      "step": 2104
+    },
+    {
+      "epoch": 0.7311566516151441,
+      "grad_norm": 0.6854168427153546,
+      "learning_rate": 1.7793627245901236e-06,
+      "loss": 0.0175,
+      "step": 2105
+    },
+    {
+      "epoch": 0.7315039944425148,
+      "grad_norm": 0.594503372287282,
+      "learning_rate": 1.775061287725906e-06,
+      "loss": 0.0228,
+      "step": 2106
+    },
+    {
+      "epoch": 0.7318513372698854,
+      "grad_norm": 1.3778446243347249,
+      "learning_rate": 1.7707639339647015e-06,
+      "loss": 0.0168,
+      "step": 2107
+    },
+    {
+      "epoch": 0.732198680097256,
+      "grad_norm": 0.5276235706829822,
+      "learning_rate": 1.766470668747403e-06,
+      "loss": 0.0175,
+      "step": 2108
+    },
+    {
+      "epoch": 0.7325460229246266,
+      "grad_norm": 0.45943667910861274,
+      "learning_rate": 1.7621814975097274e-06,
+      "loss": 0.023,
+      "step": 2109
+    },
+    {
+      "epoch": 0.7328933657519973,
+      "grad_norm": 0.3538644823362223,
+      "learning_rate": 1.7578964256822018e-06,
+      "loss": 0.0146,
+      "step": 2110
+    },
+    {
+      "epoch": 0.7332407085793679,
+      "grad_norm": 0.7312777836477884,
+      "learning_rate": 1.753615458690166e-06,
+      "loss": 0.0172,
+      "step": 2111
+    },
+    {
+      "epoch": 0.7335880514067384,
+      "grad_norm": 0.9654485421097003,
+      "learning_rate": 1.7493386019537645e-06,
+      "loss": 0.0279,
+      "step": 2112
+    },
+    {
+      "epoch": 0.733935394234109,
+      "grad_norm": 0.4641407414319063,
+      "learning_rate": 1.7450658608879384e-06,
+      "loss": 0.0216,
+      "step": 2113
+    },
+    {
+      "epoch": 0.7342827370614797,
+      "grad_norm": 0.6553590581631387,
+      "learning_rate": 1.7407972409024133e-06,
+      "loss": 0.0106,
+      "step": 2114
+    },
+    {
+      "epoch": 0.7346300798888503,
+      "grad_norm": 0.7171981611687974,
+      "learning_rate": 1.7365327474016979e-06,
+      "loss": 0.0132,
+      "step": 2115
+    },
+    {
+      "epoch": 0.7349774227162209,
+      "grad_norm": 0.6691806372000761,
+      "learning_rate": 1.7322723857850816e-06,
+      "loss": 0.0125,
+      "step": 2116
+    },
+    {
+      "epoch": 0.7353247655435915,
+      "grad_norm": 0.8491625521965726,
+      "learning_rate": 1.7280161614466185e-06,
+      "loss": 0.0237,
+      "step": 2117
+    },
+    {
+      "epoch": 0.7356721083709621,
+      "grad_norm": 0.7559623864731797,
+      "learning_rate": 1.7237640797751249e-06,
+      "loss": 0.0169,
+      "step": 2118
+    },
+    {
+      "epoch": 0.7360194511983328,
+      "grad_norm": 0.9441548462460033,
+      "learning_rate": 1.7195161461541692e-06,
+      "loss": 0.0278,
+      "step": 2119
+    },
+    {
+      "epoch": 0.7363667940257034,
+      "grad_norm": 0.5416094909605166,
+      "learning_rate": 1.7152723659620735e-06,
+      "loss": 0.0143,
+      "step": 2120
+    },
+    {
+      "epoch": 0.736714136853074,
+      "grad_norm": 0.5132478425979125,
+      "learning_rate": 1.7110327445718995e-06,
+      "loss": 0.0145,
+      "step": 2121
+    },
+    {
+      "epoch": 0.7370614796804446,
+      "grad_norm": 0.9110313850628509,
+      "learning_rate": 1.706797287351441e-06,
+      "loss": 0.0255,
+      "step": 2122
+    },
+    {
+      "epoch": 0.7374088225078153,
+      "grad_norm": 0.9146610099417635,
+      "learning_rate": 1.7025659996632198e-06,
+      "loss": 0.0152,
+      "step": 2123
+    },
+    {
+      "epoch": 0.7377561653351858,
+      "grad_norm": 0.6272856149050274,
+      "learning_rate": 1.6983388868644834e-06,
+      "loss": 0.0161,
+      "step": 2124
+    },
+    {
+      "epoch": 0.7381035081625564,
+      "grad_norm": 0.4677956955658168,
+      "learning_rate": 1.6941159543071855e-06,
+      "loss": 0.0159,
+      "step": 2125
+    },
+    {
+      "epoch": 0.738450850989927,
+      "grad_norm": 0.4336551332394349,
+      "learning_rate": 1.689897207337996e-06,
+      "loss": 0.0199,
+      "step": 2126
+    },
+    {
+      "epoch": 0.7387981938172977,
+      "grad_norm": 0.5590648865175777,
+      "learning_rate": 1.6856826512982772e-06,
+      "loss": 0.0224,
+      "step": 2127
+    },
+    {
+      "epoch": 0.7391455366446683,
+      "grad_norm": 0.6332517012226541,
+      "learning_rate": 1.6814722915240922e-06,
+      "loss": 0.0177,
+      "step": 2128
+    },
+    {
+      "epoch": 0.7394928794720389,
+      "grad_norm": 0.5184386784352486,
+      "learning_rate": 1.6772661333461858e-06,
+      "loss": 0.0118,
+      "step": 2129
+    },
+    {
+      "epoch": 0.7398402222994095,
+      "grad_norm": 0.8189234197423481,
+      "learning_rate": 1.673064182089988e-06,
+      "loss": 0.0147,
+      "step": 2130
+    },
+    {
+      "epoch": 0.7401875651267801,
+      "grad_norm": 0.6863762658701025,
+      "learning_rate": 1.6688664430755964e-06,
+      "loss": 0.0227,
+      "step": 2131
+    },
+    {
+      "epoch": 0.7405349079541508,
+      "grad_norm": 0.5772056540395054,
+      "learning_rate": 1.6646729216177827e-06,
+      "loss": 0.0181,
+      "step": 2132
+    },
+    {
+      "epoch": 0.7408822507815214,
+      "grad_norm": 0.5758071456756751,
+      "learning_rate": 1.6604836230259713e-06,
+      "loss": 0.0143,
+      "step": 2133
+    },
+    {
+      "epoch": 0.741229593608892,
+      "grad_norm": 0.5432719143241201,
+      "learning_rate": 1.6562985526042474e-06,
+      "loss": 0.0192,
+      "step": 2134
+    },
+    {
+      "epoch": 0.7415769364362625,
+      "grad_norm": 0.4669709741954846,
+      "learning_rate": 1.6521177156513351e-06,
+      "loss": 0.0186,
+      "step": 2135
+    },
+    {
+      "epoch": 0.7419242792636332,
+      "grad_norm": 0.6288408754669941,
+      "learning_rate": 1.6479411174606069e-06,
+      "loss": 0.0164,
+      "step": 2136
+    },
+    {
+      "epoch": 0.7422716220910038,
+      "grad_norm": 0.5089242192772048,
+      "learning_rate": 1.6437687633200604e-06,
+      "loss": 0.0207,
+      "step": 2137
+    },
+    {
+      "epoch": 0.7426189649183744,
+      "grad_norm": 0.6965392337483026,
+      "learning_rate": 1.639600658512327e-06,
+      "loss": 0.0147,
+      "step": 2138
+    },
+    {
+      "epoch": 0.742966307745745,
+      "grad_norm": 0.2523207249854437,
+      "learning_rate": 1.6354368083146532e-06,
+      "loss": 0.0099,
+      "step": 2139
+    },
+    {
+      "epoch": 0.7433136505731157,
+      "grad_norm": 0.9510970817995196,
+      "learning_rate": 1.6312772179988983e-06,
+      "loss": 0.0257,
+      "step": 2140
+    },
+    {
+      "epoch": 0.7436609934004863,
+      "grad_norm": 0.7060524417767109,
+      "learning_rate": 1.6271218928315325e-06,
+      "loss": 0.0199,
+      "step": 2141
+    },
+    {
+      "epoch": 0.7440083362278569,
+      "grad_norm": 0.3884044654632346,
+      "learning_rate": 1.6229708380736237e-06,
+      "loss": 0.0153,
+      "step": 2142
+    },
+    {
+      "epoch": 0.7443556790552275,
+      "grad_norm": 0.5770229127268146,
+      "learning_rate": 1.6188240589808325e-06,
+      "loss": 0.0227,
+      "step": 2143
+    },
+    {
+      "epoch": 0.7447030218825981,
+      "grad_norm": 0.5435182445456208,
+      "learning_rate": 1.6146815608034033e-06,
+      "loss": 0.0142,
+      "step": 2144
+    },
+    {
+      "epoch": 0.7450503647099688,
+      "grad_norm": 0.40026444675826955,
+      "learning_rate": 1.6105433487861666e-06,
+      "loss": 0.0172,
+      "step": 2145
+    },
+    {
+      "epoch": 0.7453977075373394,
+      "grad_norm": 0.4362346680833222,
+      "learning_rate": 1.6064094281685239e-06,
+      "loss": 0.0113,
+      "step": 2146
+    },
+    {
+      "epoch": 0.74574505036471,
+      "grad_norm": 1.0318993306154467,
+      "learning_rate": 1.6022798041844407e-06,
+      "loss": 0.0174,
+      "step": 2147
+    },
+    {
+      "epoch": 0.7460923931920805,
+      "grad_norm": 0.5504252591968342,
+      "learning_rate": 1.598154482062443e-06,
+      "loss": 0.0142,
+      "step": 2148
+    },
+    {
+      "epoch": 0.7464397360194512,
+      "grad_norm": 0.5110175837775568,
+      "learning_rate": 1.594033467025613e-06,
+      "loss": 0.0114,
+      "step": 2149
+    },
+    {
+      "epoch": 0.7467870788468218,
+      "grad_norm": 0.9348497562735546,
+      "learning_rate": 1.5899167642915803e-06,
+      "loss": 0.0244,
+      "step": 2150
+    },
+    {
+      "epoch": 0.7471344216741924,
+      "grad_norm": 0.7812495568758424,
+      "learning_rate": 1.5858043790725096e-06,
+      "loss": 0.0246,
+      "step": 2151
+    },
+    {
+      "epoch": 0.747481764501563,
+      "grad_norm": 0.9116632509191444,
+      "learning_rate": 1.5816963165751026e-06,
+      "loss": 0.0135,
+      "step": 2152
+    },
+    {
+      "epoch": 0.7478291073289337,
+      "grad_norm": 0.5280524057084818,
+      "learning_rate": 1.5775925820005878e-06,
+      "loss": 0.0111,
+      "step": 2153
+    },
+    {
+      "epoch": 0.7481764501563043,
+      "grad_norm": 1.0020440686822643,
+      "learning_rate": 1.5734931805447151e-06,
+      "loss": 0.0312,
+      "step": 2154
+    },
+    {
+      "epoch": 0.7485237929836749,
+      "grad_norm": 0.5849497270719394,
+      "learning_rate": 1.5693981173977468e-06,
+      "loss": 0.0202,
+      "step": 2155
+    },
+    {
+      "epoch": 0.7488711358110455,
+      "grad_norm": 0.4259160011756788,
+      "learning_rate": 1.56530739774445e-06,
+      "loss": 0.0211,
+      "step": 2156
+    },
+    {
+      "epoch": 0.7492184786384161,
+      "grad_norm": 0.7638640166062615,
+      "learning_rate": 1.5612210267640987e-06,
+      "loss": 0.0113,
+      "step": 2157
+    },
+    {
+      "epoch": 0.7495658214657868,
+      "grad_norm": 0.43127591153772443,
+      "learning_rate": 1.5571390096304545e-06,
+      "loss": 0.0171,
+      "step": 2158
+    },
+    {
+      "epoch": 0.7499131642931574,
+      "grad_norm": 0.7943011759564156,
+      "learning_rate": 1.5530613515117721e-06,
+      "loss": 0.0187,
+      "step": 2159
+    },
+    {
+      "epoch": 0.7502605071205279,
+      "grad_norm": 1.4178672936850725,
+      "learning_rate": 1.5489880575707821e-06,
+      "loss": 0.0226,
+      "step": 2160
+    },
+    {
+      "epoch": 0.7506078499478985,
+      "grad_norm": 0.42032498283687775,
+      "learning_rate": 1.5449191329646951e-06,
+      "loss": 0.0152,
+      "step": 2161
+    },
+    {
+      "epoch": 0.7509551927752692,
+      "grad_norm": 0.4736721695112062,
+      "learning_rate": 1.5408545828451838e-06,
+      "loss": 0.0163,
+      "step": 2162
+    },
+    {
+      "epoch": 0.7513025356026398,
+      "grad_norm": 0.44347628963881275,
+      "learning_rate": 1.5367944123583884e-06,
+      "loss": 0.0177,
+      "step": 2163
+    },
+    {
+      "epoch": 0.7516498784300104,
+      "grad_norm": 0.4305512192092599,
+      "learning_rate": 1.5327386266448973e-06,
+      "loss": 0.0197,
+      "step": 2164
+    },
+    {
+      "epoch": 0.751997221257381,
+      "grad_norm": 0.304280894510165,
+      "learning_rate": 1.528687230839755e-06,
+      "loss": 0.0144,
+      "step": 2165
+    },
+    {
+      "epoch": 0.7523445640847517,
+      "grad_norm": 0.3806340910734509,
+      "learning_rate": 1.5246402300724406e-06,
+      "loss": 0.0186,
+      "step": 2166
+    },
+    {
+      "epoch": 0.7526919069121223,
+      "grad_norm": 0.49033443578842467,
+      "learning_rate": 1.5205976294668745e-06,
+      "loss": 0.0112,
+      "step": 2167
+    },
+    {
+      "epoch": 0.7530392497394929,
+      "grad_norm": 0.6297931905210379,
+      "learning_rate": 1.5165594341414014e-06,
+      "loss": 0.0139,
+      "step": 2168
+    },
+    {
+      "epoch": 0.7533865925668635,
+      "grad_norm": 0.3703644340810331,
+      "learning_rate": 1.5125256492087925e-06,
+      "loss": 0.0108,
+      "step": 2169
+    },
+    {
+      "epoch": 0.7537339353942341,
+      "grad_norm": 0.9633658417879537,
+      "learning_rate": 1.5084962797762303e-06,
+      "loss": 0.0187,
+      "step": 2170
+    },
+    {
+      "epoch": 0.7540812782216048,
+      "grad_norm": 0.3787687052461065,
+      "learning_rate": 1.5044713309453135e-06,
+      "loss": 0.0139,
+      "step": 2171
+    },
+    {
+      "epoch": 0.7544286210489753,
+      "grad_norm": 1.1842179559226889,
+      "learning_rate": 1.5004508078120378e-06,
+      "loss": 0.0211,
+      "step": 2172
+    },
+    {
+      "epoch": 0.7547759638763459,
+      "grad_norm": 0.617894991221659,
+      "learning_rate": 1.4964347154667959e-06,
+      "loss": 0.0235,
+      "step": 2173
+    },
+    {
+      "epoch": 0.7551233067037165,
+      "grad_norm": 1.2108002626639653,
+      "learning_rate": 1.4924230589943738e-06,
+      "loss": 0.025,
+      "step": 2174
+    },
+    {
+      "epoch": 0.7554706495310872,
+      "grad_norm": 0.5389764197428066,
+      "learning_rate": 1.488415843473942e-06,
+      "loss": 0.0222,
+      "step": 2175
+    },
+    {
+      "epoch": 0.7558179923584578,
+      "grad_norm": 0.5678337368948643,
+      "learning_rate": 1.4844130739790441e-06,
+      "loss": 0.0121,
+      "step": 2176
+    },
+    {
+      "epoch": 0.7561653351858284,
+      "grad_norm": 0.5096074221105189,
+      "learning_rate": 1.4804147555775955e-06,
+      "loss": 0.0162,
+      "step": 2177
+    },
+    {
+      "epoch": 0.756512678013199,
+      "grad_norm": 0.3945803267973934,
+      "learning_rate": 1.4764208933318786e-06,
+      "loss": 0.0148,
+      "step": 2178
+    },
+    {
+      "epoch": 0.7568600208405697,
+      "grad_norm": 0.7056464507505118,
+      "learning_rate": 1.472431492298534e-06,
+      "loss": 0.0277,
+      "step": 2179
+    },
+    {
+      "epoch": 0.7572073636679403,
+      "grad_norm": 0.4702850476473721,
+      "learning_rate": 1.4684465575285507e-06,
+      "loss": 0.0227,
+      "step": 2180
+    },
+    {
+      "epoch": 0.7575547064953109,
+      "grad_norm": 0.7774483178392437,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.0165,
+      "step": 2181
+    },
+    {
+      "epoch": 0.7579020493226815,
+      "grad_norm": 0.63743536927089,
+      "learning_rate": 1.4604901069543475e-06,
+      "loss": 0.0147,
+      "step": 2182
+    },
+    {
+      "epoch": 0.758249392150052,
+      "grad_norm": 1.2057859393359813,
+      "learning_rate": 1.4565186012238126e-06,
+      "loss": 0.02,
+      "step": 2183
+    },
+    {
+      "epoch": 0.7585967349774227,
+      "grad_norm": 0.6058695506497026,
+      "learning_rate": 1.452551581903991e-06,
+      "loss": 0.0171,
+      "step": 2184
+    },
+    {
+      "epoch": 0.7589440778047933,
+      "grad_norm": 1.0470651557089286,
+      "learning_rate": 1.4485890540175335e-06,
+      "loss": 0.0211,
+      "step": 2185
+    },
+    {
+      "epoch": 0.7592914206321639,
+      "grad_norm": 0.5419134679450275,
+      "learning_rate": 1.4446310225814087e-06,
+      "loss": 0.017,
+      "step": 2186
+    },
+    {
+      "epoch": 0.7596387634595345,
+      "grad_norm": 0.6607230911522339,
+      "learning_rate": 1.4406774926068912e-06,
+      "loss": 0.0097,
+      "step": 2187
+    },
+    {
+      "epoch": 0.7599861062869052,
+      "grad_norm": 0.3575619358712941,
+      "learning_rate": 1.4367284690995543e-06,
+      "loss": 0.0156,
+      "step": 2188
+    },
+    {
+      "epoch": 0.7603334491142758,
+      "grad_norm": 0.36101367127865347,
+      "learning_rate": 1.4327839570592644e-06,
+      "loss": 0.0132,
+      "step": 2189
+    },
+    {
+      "epoch": 0.7606807919416464,
+      "grad_norm": 0.5040284990352497,
+      "learning_rate": 1.4288439614801803e-06,
+      "loss": 0.0162,
+      "step": 2190
+    },
+    {
+      "epoch": 0.761028134769017,
+      "grad_norm": 0.9760018252254876,
+      "learning_rate": 1.4249084873507412e-06,
+      "loss": 0.0233,
+      "step": 2191
+    },
+    {
+      "epoch": 0.7613754775963877,
+      "grad_norm": 0.6260035980310652,
+      "learning_rate": 1.4209775396536595e-06,
+      "loss": 0.0247,
+      "step": 2192
+    },
+    {
+      "epoch": 0.7617228204237583,
+      "grad_norm": 0.723938835995245,
+      "learning_rate": 1.4170511233659167e-06,
+      "loss": 0.0232,
+      "step": 2193
+    },
+    {
+      "epoch": 0.7620701632511289,
+      "grad_norm": 0.5719862112874641,
+      "learning_rate": 1.4131292434587613e-06,
+      "loss": 0.0187,
+      "step": 2194
+    },
+    {
+      "epoch": 0.7624175060784995,
+      "grad_norm": 0.6344034231978065,
+      "learning_rate": 1.409211904897692e-06,
+      "loss": 0.0128,
+      "step": 2195
+    },
+    {
+      "epoch": 0.76276484890587,
+      "grad_norm": 1.0837290646085158,
+      "learning_rate": 1.4052991126424642e-06,
+      "loss": 0.03,
+      "step": 2196
+    },
+    {
+      "epoch": 0.7631121917332407,
+      "grad_norm": 0.5566631869774307,
+      "learning_rate": 1.4013908716470714e-06,
+      "loss": 0.0188,
+      "step": 2197
+    },
+    {
+      "epoch": 0.7634595345606113,
+      "grad_norm": 0.9291107450950402,
+      "learning_rate": 1.3974871868597495e-06,
+      "loss": 0.0212,
+      "step": 2198
+    },
+    {
+      "epoch": 0.7638068773879819,
+      "grad_norm": 0.6433035528020483,
+      "learning_rate": 1.3935880632229614e-06,
+      "loss": 0.0148,
+      "step": 2199
+    },
+    {
+      "epoch": 0.7641542202153525,
+      "grad_norm": 0.5318264623442518,
+      "learning_rate": 1.3896935056734001e-06,
+      "loss": 0.0143,
+      "step": 2200
+    },
+    {
+      "epoch": 0.7645015630427232,
+      "grad_norm": 0.3392416546423096,
+      "learning_rate": 1.385803519141971e-06,
+      "loss": 0.01,
+      "step": 2201
+    },
+    {
+      "epoch": 0.7648489058700938,
+      "grad_norm": 0.4357906997308162,
+      "learning_rate": 1.3819181085538002e-06,
+      "loss": 0.0158,
+      "step": 2202
+    },
+    {
+      "epoch": 0.7651962486974644,
+      "grad_norm": 0.8938906147298566,
+      "learning_rate": 1.378037278828212e-06,
+      "loss": 0.0157,
+      "step": 2203
+    },
+    {
+      "epoch": 0.765543591524835,
+      "grad_norm": 0.812151390132714,
+      "learning_rate": 1.3741610348787382e-06,
+      "loss": 0.0201,
+      "step": 2204
+    },
+    {
+      "epoch": 0.7658909343522057,
+      "grad_norm": 1.046728570840569,
+      "learning_rate": 1.3702893816130968e-06,
+      "loss": 0.0138,
+      "step": 2205
+    },
+    {
+      "epoch": 0.7662382771795763,
+      "grad_norm": 0.8470882919333848,
+      "learning_rate": 1.366422323933202e-06,
+      "loss": 0.0217,
+      "step": 2206
+    },
+    {
+      "epoch": 0.7665856200069469,
+      "grad_norm": 0.6419335535775957,
+      "learning_rate": 1.362559866735142e-06,
+      "loss": 0.0188,
+      "step": 2207
+    },
+    {
+      "epoch": 0.7669329628343174,
+      "grad_norm": 1.0195274448965153,
+      "learning_rate": 1.3587020149091856e-06,
+      "loss": 0.0188,
+      "step": 2208
+    },
+    {
+      "epoch": 0.767280305661688,
+      "grad_norm": 0.8491806303868048,
+      "learning_rate": 1.3548487733397686e-06,
+      "loss": 0.0136,
+      "step": 2209
+    },
+    {
+      "epoch": 0.7676276484890587,
+      "grad_norm": 0.496099824248827,
+      "learning_rate": 1.351000146905488e-06,
+      "loss": 0.0159,
+      "step": 2210
+    },
+    {
+      "epoch": 0.7679749913164293,
+      "grad_norm": 0.36784724782559614,
+      "learning_rate": 1.3471561404791e-06,
+      "loss": 0.0099,
+      "step": 2211
+    },
+    {
+      "epoch": 0.7683223341437999,
+      "grad_norm": 0.853519257078698,
+      "learning_rate": 1.343316758927513e-06,
+      "loss": 0.0152,
+      "step": 2212
+    },
+    {
+      "epoch": 0.7686696769711705,
+      "grad_norm": 0.4794009903353391,
+      "learning_rate": 1.3394820071117765e-06,
+      "loss": 0.0178,
+      "step": 2213
+    },
+    {
+      "epoch": 0.7690170197985412,
+      "grad_norm": 0.5638174933594083,
+      "learning_rate": 1.3356518898870773e-06,
+      "loss": 0.0226,
+      "step": 2214
+    },
+    {
+      "epoch": 0.7693643626259118,
+      "grad_norm": 0.7867218167416455,
+      "learning_rate": 1.331826412102738e-06,
+      "loss": 0.02,
+      "step": 2215
+    },
+    {
+      "epoch": 0.7697117054532824,
+      "grad_norm": 1.2101224505178252,
+      "learning_rate": 1.3280055786022078e-06,
+      "loss": 0.0339,
+      "step": 2216
+    },
+    {
+      "epoch": 0.770059048280653,
+      "grad_norm": 0.4560784481807581,
+      "learning_rate": 1.3241893942230511e-06,
+      "loss": 0.0153,
+      "step": 2217
+    },
+    {
+      "epoch": 0.7704063911080237,
+      "grad_norm": 1.089455223796198,
+      "learning_rate": 1.3203778637969478e-06,
+      "loss": 0.02,
+      "step": 2218
+    },
+    {
+      "epoch": 0.7707537339353943,
+      "grad_norm": 0.8102592696114593,
+      "learning_rate": 1.3165709921496873e-06,
+      "loss": 0.0146,
+      "step": 2219
+    },
+    {
+      "epoch": 0.7711010767627648,
+      "grad_norm": 0.2248352987503352,
+      "learning_rate": 1.312768784101161e-06,
+      "loss": 0.0096,
+      "step": 2220
+    },
+    {
+      "epoch": 0.7714484195901354,
+      "grad_norm": 0.6361836378952002,
+      "learning_rate": 1.3089712444653525e-06,
+      "loss": 0.0282,
+      "step": 2221
+    },
+    {
+      "epoch": 0.771795762417506,
+      "grad_norm": 0.6975087143669682,
+      "learning_rate": 1.3051783780503353e-06,
+      "loss": 0.0123,
+      "step": 2222
+    },
+    {
+      "epoch": 0.7721431052448767,
+      "grad_norm": 1.2464836526873286,
+      "learning_rate": 1.3013901896582677e-06,
+      "loss": 0.027,
+      "step": 2223
+    },
+    {
+      "epoch": 0.7724904480722473,
+      "grad_norm": 0.4282700872207025,
+      "learning_rate": 1.2976066840853862e-06,
+      "loss": 0.0198,
+      "step": 2224
+    },
+    {
+      "epoch": 0.7728377908996179,
+      "grad_norm": 0.4954085912972942,
+      "learning_rate": 1.2938278661219961e-06,
+      "loss": 0.011,
+      "step": 2225
+    },
+    {
+      "epoch": 0.7731851337269885,
+      "grad_norm": 0.64840741080104,
+      "learning_rate": 1.290053740552466e-06,
+      "loss": 0.0198,
+      "step": 2226
+    },
+    {
+      "epoch": 0.7735324765543592,
+      "grad_norm": 0.582268992296493,
+      "learning_rate": 1.2862843121552293e-06,
+      "loss": 0.0121,
+      "step": 2227
+    },
+    {
+      "epoch": 0.7738798193817298,
+      "grad_norm": 0.5788123624597133,
+      "learning_rate": 1.282519585702765e-06,
+      "loss": 0.0181,
+      "step": 2228
+    },
+    {
+      "epoch": 0.7742271622091004,
+      "grad_norm": 0.42036499722806164,
+      "learning_rate": 1.2787595659616063e-06,
+      "loss": 0.019,
+      "step": 2229
+    },
+    {
+      "epoch": 0.774574505036471,
+      "grad_norm": 1.2403906461674592,
+      "learning_rate": 1.275004257692321e-06,
+      "loss": 0.0192,
+      "step": 2230
+    },
+    {
+      "epoch": 0.7749218478638417,
+      "grad_norm": 0.33842093190437655,
+      "learning_rate": 1.2712536656495167e-06,
+      "loss": 0.0132,
+      "step": 2231
+    },
+    {
+      "epoch": 0.7752691906912123,
+      "grad_norm": 0.9723522981189887,
+      "learning_rate": 1.2675077945818249e-06,
+      "loss": 0.0194,
+      "step": 2232
+    },
+    {
+      "epoch": 0.7756165335185828,
+      "grad_norm": 0.6452976079476872,
+      "learning_rate": 1.263766649231905e-06,
+      "loss": 0.0235,
+      "step": 2233
+    },
+    {
+      "epoch": 0.7759638763459534,
+      "grad_norm": 0.5932675671658317,
+      "learning_rate": 1.260030234336428e-06,
+      "loss": 0.016,
+      "step": 2234
+    },
+    {
+      "epoch": 0.776311219173324,
+      "grad_norm": 0.76806867313472,
+      "learning_rate": 1.2562985546260804e-06,
+      "loss": 0.0187,
+      "step": 2235
+    },
+    {
+      "epoch": 0.7766585620006947,
+      "grad_norm": 0.5650815303570309,
+      "learning_rate": 1.252571614825549e-06,
+      "loss": 0.0169,
+      "step": 2236
+    },
+    {
+      "epoch": 0.7770059048280653,
+      "grad_norm": 0.48992636927446587,
+      "learning_rate": 1.2488494196535238e-06,
+      "loss": 0.0198,
+      "step": 2237
+    },
+    {
+      "epoch": 0.7773532476554359,
+      "grad_norm": 0.47434548335024923,
+      "learning_rate": 1.2451319738226835e-06,
+      "loss": 0.0148,
+      "step": 2238
+    },
+    {
+      "epoch": 0.7777005904828065,
+      "grad_norm": 1.587581496347539,
+      "learning_rate": 1.2414192820396987e-06,
+      "loss": 0.016,
+      "step": 2239
+    },
+    {
+      "epoch": 0.7780479333101772,
+      "grad_norm": 0.5941570896375952,
+      "learning_rate": 1.237711349005214e-06,
+      "loss": 0.0224,
+      "step": 2240
+    },
+    {
+      "epoch": 0.7783952761375478,
+      "grad_norm": 0.45192769841614433,
+      "learning_rate": 1.234008179413856e-06,
+      "loss": 0.018,
+      "step": 2241
+    },
+    {
+      "epoch": 0.7787426189649184,
+      "grad_norm": 0.5359045238517698,
+      "learning_rate": 1.2303097779542151e-06,
+      "loss": 0.015,
+      "step": 2242
+    },
+    {
+      "epoch": 0.779089961792289,
+      "grad_norm": 0.3170868716065785,
+      "learning_rate": 1.2266161493088463e-06,
+      "loss": 0.0157,
+      "step": 2243
+    },
+    {
+      "epoch": 0.7794373046196597,
+      "grad_norm": 0.5339325400301773,
+      "learning_rate": 1.2229272981542628e-06,
+      "loss": 0.017,
+      "step": 2244
+    },
+    {
+      "epoch": 0.7797846474470302,
+      "grad_norm": 0.5749594088375014,
+      "learning_rate": 1.2192432291609296e-06,
+      "loss": 0.0191,
+      "step": 2245
+    },
+    {
+      "epoch": 0.7801319902744008,
+      "grad_norm": 0.8657654364305364,
+      "learning_rate": 1.2155639469932551e-06,
+      "loss": 0.0157,
+      "step": 2246
+    },
+    {
+      "epoch": 0.7804793331017714,
+      "grad_norm": 0.6662929638257185,
+      "learning_rate": 1.2118894563095857e-06,
+      "loss": 0.0103,
+      "step": 2247
+    },
+    {
+      "epoch": 0.780826675929142,
+      "grad_norm": 0.6185784496325207,
+      "learning_rate": 1.2082197617622049e-06,
+      "loss": 0.0271,
+      "step": 2248
+    },
+    {
+      "epoch": 0.7811740187565127,
+      "grad_norm": 0.6728910637327504,
+      "learning_rate": 1.2045548679973234e-06,
+      "loss": 0.0204,
+      "step": 2249
+    },
+    {
+      "epoch": 0.7815213615838833,
+      "grad_norm": 0.503622779557992,
+      "learning_rate": 1.2008947796550714e-06,
+      "loss": 0.0168,
+      "step": 2250
+    },
+    {
+      "epoch": 0.7818687044112539,
+      "grad_norm": 0.7057212026990866,
+      "learning_rate": 1.1972395013694944e-06,
+      "loss": 0.0178,
+      "step": 2251
+    },
+    {
+      "epoch": 0.7822160472386245,
+      "grad_norm": 0.9263129101479496,
+      "learning_rate": 1.1935890377685499e-06,
+      "loss": 0.0177,
+      "step": 2252
+    },
+    {
+      "epoch": 0.7825633900659952,
+      "grad_norm": 0.5928517673288792,
+      "learning_rate": 1.1899433934741023e-06,
+      "loss": 0.0242,
+      "step": 2253
+    },
+    {
+      "epoch": 0.7829107328933658,
+      "grad_norm": 0.286499819754983,
+      "learning_rate": 1.186302573101908e-06,
+      "loss": 0.011,
+      "step": 2254
+    },
+    {
+      "epoch": 0.7832580757207364,
+      "grad_norm": 1.0908468375007774,
+      "learning_rate": 1.1826665812616183e-06,
+      "loss": 0.0343,
+      "step": 2255
+    },
+    {
+      "epoch": 0.783605418548107,
+      "grad_norm": 0.3623784367226302,
+      "learning_rate": 1.1790354225567724e-06,
+      "loss": 0.0112,
+      "step": 2256
+    },
+    {
+      "epoch": 0.7839527613754776,
+      "grad_norm": 0.3445252725726735,
+      "learning_rate": 1.175409101584793e-06,
+      "loss": 0.0093,
+      "step": 2257
+    },
+    {
+      "epoch": 0.7843001042028482,
+      "grad_norm": 0.5433146108276392,
+      "learning_rate": 1.1717876229369679e-06,
+      "loss": 0.0151,
+      "step": 2258
+    },
+    {
+      "epoch": 0.7846474470302188,
+      "grad_norm": 0.63386269312144,
+      "learning_rate": 1.168170991198464e-06,
+      "loss": 0.0211,
+      "step": 2259
+    },
+    {
+      "epoch": 0.7849947898575894,
+      "grad_norm": 0.5237366655341864,
+      "learning_rate": 1.1645592109483083e-06,
+      "loss": 0.0106,
+      "step": 2260
+    },
+    {
+      "epoch": 0.78534213268496,
+      "grad_norm": 0.4926580735814283,
+      "learning_rate": 1.1609522867593825e-06,
+      "loss": 0.0164,
+      "step": 2261
+    },
+    {
+      "epoch": 0.7856894755123307,
+      "grad_norm": 0.618563652888975,
+      "learning_rate": 1.1573502231984252e-06,
+      "loss": 0.0171,
+      "step": 2262
+    },
+    {
+      "epoch": 0.7860368183397013,
+      "grad_norm": 0.42754249596279326,
+      "learning_rate": 1.1537530248260154e-06,
+      "loss": 0.0094,
+      "step": 2263
+    },
+    {
+      "epoch": 0.7863841611670719,
+      "grad_norm": 0.4479479051275309,
+      "learning_rate": 1.1501606961965772e-06,
+      "loss": 0.0164,
+      "step": 2264
+    },
+    {
+      "epoch": 0.7867315039944425,
+      "grad_norm": 1.0791717141203099,
+      "learning_rate": 1.1465732418583652e-06,
+      "loss": 0.0216,
+      "step": 2265
+    },
+    {
+      "epoch": 0.7870788468218132,
+      "grad_norm": 0.7585334147960321,
+      "learning_rate": 1.1429906663534661e-06,
+      "loss": 0.0199,
+      "step": 2266
+    },
+    {
+      "epoch": 0.7874261896491838,
+      "grad_norm": 0.5631229523842227,
+      "learning_rate": 1.1394129742177856e-06,
+      "loss": 0.0155,
+      "step": 2267
+    },
+    {
+      "epoch": 0.7877735324765544,
+      "grad_norm": 0.7372418292459169,
+      "learning_rate": 1.1358401699810513e-06,
+      "loss": 0.0162,
+      "step": 2268
+    },
+    {
+      "epoch": 0.7881208753039249,
+      "grad_norm": 0.3814232843076739,
+      "learning_rate": 1.1322722581667972e-06,
+      "loss": 0.012,
+      "step": 2269
+    },
+    {
+      "epoch": 0.7884682181312956,
+      "grad_norm": 0.3900814023948877,
+      "learning_rate": 1.1287092432923675e-06,
+      "loss": 0.0165,
+      "step": 2270
+    },
+    {
+      "epoch": 0.7888155609586662,
+      "grad_norm": 0.3684093471163428,
+      "learning_rate": 1.1251511298689015e-06,
+      "loss": 0.0129,
+      "step": 2271
+    },
+    {
+      "epoch": 0.7891629037860368,
+      "grad_norm": 0.41541558092287234,
+      "learning_rate": 1.1215979224013395e-06,
+      "loss": 0.0073,
+      "step": 2272
+    },
+    {
+      "epoch": 0.7895102466134074,
+      "grad_norm": 0.45431105796049687,
+      "learning_rate": 1.1180496253884028e-06,
+      "loss": 0.0135,
+      "step": 2273
+    },
+    {
+      "epoch": 0.789857589440778,
+      "grad_norm": 0.8135773753424956,
+      "learning_rate": 1.1145062433226018e-06,
+      "loss": 0.0125,
+      "step": 2274
+    },
+    {
+      "epoch": 0.7902049322681487,
+      "grad_norm": 0.6545808136769573,
+      "learning_rate": 1.1109677806902203e-06,
+      "loss": 0.0136,
+      "step": 2275
+    },
+    {
+      "epoch": 0.7905522750955193,
+      "grad_norm": 0.5330476620089217,
+      "learning_rate": 1.107434241971313e-06,
+      "loss": 0.0079,
+      "step": 2276
+    },
+    {
+      "epoch": 0.7908996179228899,
+      "grad_norm": 0.6003919050746848,
+      "learning_rate": 1.1039056316397046e-06,
+      "loss": 0.014,
+      "step": 2277
+    },
+    {
+      "epoch": 0.7912469607502605,
+      "grad_norm": 0.4179604464910238,
+      "learning_rate": 1.1003819541629772e-06,
+      "loss": 0.01,
+      "step": 2278
+    },
+    {
+      "epoch": 0.7915943035776312,
+      "grad_norm": 0.7808363334276369,
+      "learning_rate": 1.0968632140024683e-06,
+      "loss": 0.0152,
+      "step": 2279
+    },
+    {
+      "epoch": 0.7919416464050018,
+      "grad_norm": 0.4167603863620862,
+      "learning_rate": 1.0933494156132607e-06,
+      "loss": 0.0169,
+      "step": 2280
+    },
+    {
+      "epoch": 0.7922889892323723,
+      "grad_norm": 1.012723748239046,
+      "learning_rate": 1.0898405634441856e-06,
+      "loss": 0.0185,
+      "step": 2281
+    },
+    {
+      "epoch": 0.7926363320597429,
+      "grad_norm": 0.6056013118255872,
+      "learning_rate": 1.0863366619378107e-06,
+      "loss": 0.016,
+      "step": 2282
+    },
+    {
+      "epoch": 0.7929836748871136,
+      "grad_norm": 0.7450736652041653,
+      "learning_rate": 1.0828377155304332e-06,
+      "loss": 0.0224,
+      "step": 2283
+    },
+    {
+      "epoch": 0.7933310177144842,
+      "grad_norm": 1.0221605650577856,
+      "learning_rate": 1.0793437286520765e-06,
+      "loss": 0.0195,
+      "step": 2284
+    },
+    {
+      "epoch": 0.7936783605418548,
+      "grad_norm": 0.6636410309989438,
+      "learning_rate": 1.0758547057264873e-06,
+      "loss": 0.0199,
+      "step": 2285
+    },
+    {
+      "epoch": 0.7940257033692254,
+      "grad_norm": 1.1325176009144873,
+      "learning_rate": 1.072370651171128e-06,
+      "loss": 0.0233,
+      "step": 2286
+    },
+    {
+      "epoch": 0.794373046196596,
+      "grad_norm": 0.5341830366952061,
+      "learning_rate": 1.0688915693971675e-06,
+      "loss": 0.012,
+      "step": 2287
+    },
+    {
+      "epoch": 0.7947203890239667,
+      "grad_norm": 0.6153155573713803,
+      "learning_rate": 1.0654174648094783e-06,
+      "loss": 0.0198,
+      "step": 2288
+    },
+    {
+      "epoch": 0.7950677318513373,
+      "grad_norm": 0.9623001317484534,
+      "learning_rate": 1.0619483418066346e-06,
+      "loss": 0.0169,
+      "step": 2289
+    },
+    {
+      "epoch": 0.7954150746787079,
+      "grad_norm": 0.7223221290889841,
+      "learning_rate": 1.0584842047809047e-06,
+      "loss": 0.0155,
+      "step": 2290
+    },
+    {
+      "epoch": 0.7957624175060785,
+      "grad_norm": 0.6504553961602982,
+      "learning_rate": 1.0550250581182353e-06,
+      "loss": 0.0224,
+      "step": 2291
+    },
+    {
+      "epoch": 0.7961097603334492,
+      "grad_norm": 1.0998554308853323,
+      "learning_rate": 1.0515709061982632e-06,
+      "loss": 0.0255,
+      "step": 2292
+    },
+    {
+      "epoch": 0.7964571031608197,
+      "grad_norm": 0.4931210569774452,
+      "learning_rate": 1.048121753394301e-06,
+      "loss": 0.0246,
+      "step": 2293
+    },
+    {
+      "epoch": 0.7968044459881903,
+      "grad_norm": 0.8526641555443321,
+      "learning_rate": 1.044677604073328e-06,
+      "loss": 0.028,
+      "step": 2294
+    },
+    {
+      "epoch": 0.7971517888155609,
+      "grad_norm": 0.6692895381848059,
+      "learning_rate": 1.0412384625959887e-06,
+      "loss": 0.02,
+      "step": 2295
+    },
+    {
+      "epoch": 0.7974991316429316,
+      "grad_norm": 0.419733606925217,
+      "learning_rate": 1.037804333316591e-06,
+      "loss": 0.0143,
+      "step": 2296
+    },
+    {
+      "epoch": 0.7978464744703022,
+      "grad_norm": 0.4921795286660977,
+      "learning_rate": 1.0343752205830948e-06,
+      "loss": 0.0166,
+      "step": 2297
+    },
+    {
+      "epoch": 0.7981938172976728,
+      "grad_norm": 0.7772755035303158,
+      "learning_rate": 1.030951128737106e-06,
+      "loss": 0.0218,
+      "step": 2298
+    },
+    {
+      "epoch": 0.7985411601250434,
+      "grad_norm": 0.5358777594915758,
+      "learning_rate": 1.027532062113879e-06,
+      "loss": 0.0237,
+      "step": 2299
+    },
+    {
+      "epoch": 0.798888502952414,
+      "grad_norm": 0.5660888423523717,
+      "learning_rate": 1.0241180250423e-06,
+      "loss": 0.0152,
+      "step": 2300
+    },
+    {
+      "epoch": 0.7992358457797847,
+      "grad_norm": 1.1542651595626319,
+      "learning_rate": 1.0207090218448923e-06,
+      "loss": 0.0125,
+      "step": 2301
+    },
+    {
+      "epoch": 0.7995831886071553,
+      "grad_norm": 1.2171482202491108,
+      "learning_rate": 1.0173050568378002e-06,
+      "loss": 0.0215,
+      "step": 2302
+    },
+    {
+      "epoch": 0.7999305314345259,
+      "grad_norm": 1.1045425181167858,
+      "learning_rate": 1.013906134330796e-06,
+      "loss": 0.0132,
+      "step": 2303
+    },
+    {
+      "epoch": 0.8002778742618964,
+      "grad_norm": 0.8205260763175557,
+      "learning_rate": 1.0105122586272615e-06,
+      "loss": 0.0168,
+      "step": 2304
+    },
+    {
+      "epoch": 0.8006252170892671,
+      "grad_norm": 0.6460529317376611,
+      "learning_rate": 1.0071234340241925e-06,
+      "loss": 0.0146,
+      "step": 2305
+    },
+    {
+      "epoch": 0.8009725599166377,
+      "grad_norm": 1.1028425848766075,
+      "learning_rate": 1.0037396648121872e-06,
+      "loss": 0.0203,
+      "step": 2306
+    },
+    {
+      "epoch": 0.8013199027440083,
+      "grad_norm": 0.6040772659608722,
+      "learning_rate": 1.0003609552754468e-06,
+      "loss": 0.0131,
+      "step": 2307
+    },
+    {
+      "epoch": 0.8016672455713789,
+      "grad_norm": 0.42232669891432706,
+      "learning_rate": 9.969873096917614e-07,
+      "loss": 0.0113,
+      "step": 2308
+    },
+    {
+      "epoch": 0.8020145883987496,
+      "grad_norm": 0.700969940979702,
+      "learning_rate": 9.93618732332512e-07,
+      "loss": 0.0191,
+      "step": 2309
+    },
+    {
+      "epoch": 0.8023619312261202,
+      "grad_norm": 0.8526190683813358,
+      "learning_rate": 9.902552274626638e-07,
+      "loss": 0.0206,
+      "step": 2310
+    },
+    {
+      "epoch": 0.8027092740534908,
+      "grad_norm": 0.5131311092706733,
+      "learning_rate": 9.868967993407603e-07,
+      "loss": 0.0139,
+      "step": 2311
+    },
+    {
+      "epoch": 0.8030566168808614,
+      "grad_norm": 0.5377387868045655,
+      "learning_rate": 9.83543452218914e-07,
+      "loss": 0.0208,
+      "step": 2312
+    },
+    {
+      "epoch": 0.803403959708232,
+      "grad_norm": 0.6207978174819988,
+      "learning_rate": 9.801951903428053e-07,
+      "loss": 0.0169,
+      "step": 2313
+    },
+    {
+      "epoch": 0.8037513025356027,
+      "grad_norm": 1.0623394552894256,
+      "learning_rate": 9.768520179516782e-07,
+      "loss": 0.033,
+      "step": 2314
+    },
+    {
+      "epoch": 0.8040986453629733,
+      "grad_norm": 0.29785651145166425,
+      "learning_rate": 9.735139392783326e-07,
+      "loss": 0.0109,
+      "step": 2315
+    },
+    {
+      "epoch": 0.8044459881903439,
+      "grad_norm": 2.62736664673775,
+      "learning_rate": 9.70180958549118e-07,
+      "loss": 0.025,
+      "step": 2316
+    },
+    {
+      "epoch": 0.8047933310177144,
+      "grad_norm": 0.5896374439097164,
+      "learning_rate": 9.66853079983927e-07,
+      "loss": 0.0225,
+      "step": 2317
+    },
+    {
+      "epoch": 0.8051406738450851,
+      "grad_norm": 0.7960436391597935,
+      "learning_rate": 9.63530307796197e-07,
+      "loss": 0.0112,
+      "step": 2318
+    },
+    {
+      "epoch": 0.8054880166724557,
+      "grad_norm": 0.615407883572327,
+      "learning_rate": 9.602126461929002e-07,
+      "loss": 0.0224,
+      "step": 2319
+    },
+    {
+      "epoch": 0.8058353594998263,
+      "grad_norm": 0.48130754370985307,
+      "learning_rate": 9.569000993745336e-07,
+      "loss": 0.0161,
+      "step": 2320
+    },
+    {
+      "epoch": 0.8061827023271969,
+      "grad_norm": 0.41555455314622086,
+      "learning_rate": 9.535926715351207e-07,
+      "loss": 0.0124,
+      "step": 2321
+    },
+    {
+      "epoch": 0.8065300451545676,
+      "grad_norm": 0.8444184620512747,
+      "learning_rate": 9.502903668622055e-07,
+      "loss": 0.0166,
+      "step": 2322
+    },
+    {
+      "epoch": 0.8068773879819382,
+      "grad_norm": 0.6578650196091546,
+      "learning_rate": 9.469931895368462e-07,
+      "loss": 0.0193,
+      "step": 2323
+    },
+    {
+      "epoch": 0.8072247308093088,
+      "grad_norm": 0.8813060676876311,
+      "learning_rate": 9.43701143733603e-07,
+      "loss": 0.017,
+      "step": 2324
+    },
+    {
+      "epoch": 0.8075720736366794,
+      "grad_norm": 0.48427747281839323,
+      "learning_rate": 9.404142336205452e-07,
+      "loss": 0.018,
+      "step": 2325
+    },
+    {
+      "epoch": 0.80791941646405,
+      "grad_norm": 0.5218223865429743,
+      "learning_rate": 9.371324633592399e-07,
+      "loss": 0.0165,
+      "step": 2326
+    },
+    {
+      "epoch": 0.8082667592914207,
+      "grad_norm": 0.5497878546670253,
+      "learning_rate": 9.338558371047429e-07,
+      "loss": 0.0155,
+      "step": 2327
+    },
+    {
+      "epoch": 0.8086141021187913,
+      "grad_norm": 0.5828845938065368,
+      "learning_rate": 9.30584359005598e-07,
+      "loss": 0.0226,
+      "step": 2328
+    },
+    {
+      "epoch": 0.8089614449461618,
+      "grad_norm": 0.45607595995290867,
+      "learning_rate": 9.273180332038328e-07,
+      "loss": 0.012,
+      "step": 2329
+    },
+    {
+      "epoch": 0.8093087877735324,
+      "grad_norm": 0.37620117048721097,
+      "learning_rate": 9.240568638349523e-07,
+      "loss": 0.0131,
+      "step": 2330
+    },
+    {
+      "epoch": 0.8096561306009031,
+      "grad_norm": 0.3290671929012356,
+      "learning_rate": 9.208008550279296e-07,
+      "loss": 0.0103,
+      "step": 2331
+    },
+    {
+      "epoch": 0.8100034734282737,
+      "grad_norm": 0.623859993612422,
+      "learning_rate": 9.175500109052044e-07,
+      "loss": 0.022,
+      "step": 2332
+    },
+    {
+      "epoch": 0.8103508162556443,
+      "grad_norm": 0.38503253698459383,
+      "learning_rate": 9.143043355826802e-07,
+      "loss": 0.0169,
+      "step": 2333
+    },
+    {
+      "epoch": 0.8106981590830149,
+      "grad_norm": 0.5608051701389355,
+      "learning_rate": 9.110638331697158e-07,
+      "loss": 0.0145,
+      "step": 2334
+    },
+    {
+      "epoch": 0.8110455019103856,
+      "grad_norm": 0.8102603807769083,
+      "learning_rate": 9.078285077691179e-07,
+      "loss": 0.0205,
+      "step": 2335
+    },
+    {
+      "epoch": 0.8113928447377562,
+      "grad_norm": 0.8701051653222716,
+      "learning_rate": 9.045983634771388e-07,
+      "loss": 0.0144,
+      "step": 2336
+    },
+    {
+      "epoch": 0.8117401875651268,
+      "grad_norm": 0.6909556680271054,
+      "learning_rate": 9.013734043834743e-07,
+      "loss": 0.0159,
+      "step": 2337
+    },
+    {
+      "epoch": 0.8120875303924974,
+      "grad_norm": 1.0316061837686277,
+      "learning_rate": 8.981536345712544e-07,
+      "loss": 0.0184,
+      "step": 2338
+    },
+    {
+      "epoch": 0.812434873219868,
+      "grad_norm": 0.7754820020926894,
+      "learning_rate": 8.949390581170341e-07,
+      "loss": 0.0241,
+      "step": 2339
+    },
+    {
+      "epoch": 0.8127822160472387,
+      "grad_norm": 0.39205221031248105,
+      "learning_rate": 8.917296790908009e-07,
+      "loss": 0.0134,
+      "step": 2340
+    },
+    {
+      "epoch": 0.8131295588746092,
+      "grad_norm": 0.7321244654973654,
+      "learning_rate": 8.885255015559552e-07,
+      "loss": 0.0192,
+      "step": 2341
+    },
+    {
+      "epoch": 0.8134769017019798,
+      "grad_norm": 1.6237885505587426,
+      "learning_rate": 8.853265295693131e-07,
+      "loss": 0.0182,
+      "step": 2342
+    },
+    {
+      "epoch": 0.8138242445293504,
+      "grad_norm": 0.40790172026137156,
+      "learning_rate": 8.821327671811025e-07,
+      "loss": 0.0158,
+      "step": 2343
+    },
+    {
+      "epoch": 0.8141715873567211,
+      "grad_norm": 0.6612071453798002,
+      "learning_rate": 8.789442184349556e-07,
+      "loss": 0.0232,
+      "step": 2344
+    },
+    {
+      "epoch": 0.8145189301840917,
+      "grad_norm": 0.5315701172252235,
+      "learning_rate": 8.757608873679008e-07,
+      "loss": 0.0226,
+      "step": 2345
+    },
+    {
+      "epoch": 0.8148662730114623,
+      "grad_norm": 0.47678792802998987,
+      "learning_rate": 8.72582778010359e-07,
+      "loss": 0.0095,
+      "step": 2346
+    },
+    {
+      "epoch": 0.8152136158388329,
+      "grad_norm": 0.4744887984080288,
+      "learning_rate": 8.694098943861457e-07,
+      "loss": 0.0134,
+      "step": 2347
+    },
+    {
+      "epoch": 0.8155609586662036,
+      "grad_norm": 0.5992243344701217,
+      "learning_rate": 8.662422405124565e-07,
+      "loss": 0.0147,
+      "step": 2348
+    },
+    {
+      "epoch": 0.8159083014935742,
+      "grad_norm": 0.45940543311353116,
+      "learning_rate": 8.630798203998653e-07,
+      "loss": 0.0171,
+      "step": 2349
+    },
+    {
+      "epoch": 0.8162556443209448,
+      "grad_norm": 0.5247973512112908,
+      "learning_rate": 8.59922638052319e-07,
+      "loss": 0.0167,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8166029871483154,
+      "grad_norm": 0.5656596886450084,
+      "learning_rate": 8.567706974671353e-07,
+      "loss": 0.012,
+      "step": 2351
+    },
+    {
+      "epoch": 0.816950329975686,
+      "grad_norm": 0.8482170289088908,
+      "learning_rate": 8.536240026349951e-07,
+      "loss": 0.0219,
+      "step": 2352
+    },
+    {
+      "epoch": 0.8172976728030567,
+      "grad_norm": 0.6035908558385402,
+      "learning_rate": 8.504825575399356e-07,
+      "loss": 0.0169,
+      "step": 2353
+    },
+    {
+      "epoch": 0.8176450156304272,
+      "grad_norm": 1.3540920045468,
+      "learning_rate": 8.473463661593473e-07,
+      "loss": 0.0197,
+      "step": 2354
+    },
+    {
+      "epoch": 0.8179923584577978,
+      "grad_norm": 0.7084599678534361,
+      "learning_rate": 8.442154324639706e-07,
+      "loss": 0.0156,
+      "step": 2355
+    },
+    {
+      "epoch": 0.8183397012851684,
+      "grad_norm": 0.40515237224446915,
+      "learning_rate": 8.410897604178913e-07,
+      "loss": 0.0144,
+      "step": 2356
+    },
+    {
+      "epoch": 0.8186870441125391,
+      "grad_norm": 1.0892964020587188,
+      "learning_rate": 8.379693539785266e-07,
+      "loss": 0.0292,
+      "step": 2357
+    },
+    {
+      "epoch": 0.8190343869399097,
+      "grad_norm": 1.4771693089117002,
+      "learning_rate": 8.348542170966317e-07,
+      "loss": 0.0215,
+      "step": 2358
+    },
+    {
+      "epoch": 0.8193817297672803,
+      "grad_norm": 0.4065081151813109,
+      "learning_rate": 8.317443537162922e-07,
+      "loss": 0.0146,
+      "step": 2359
+    },
+    {
+      "epoch": 0.8197290725946509,
+      "grad_norm": 0.5743831870975237,
+      "learning_rate": 8.286397677749114e-07,
+      "loss": 0.0201,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8200764154220216,
+      "grad_norm": 0.5683746981641697,
+      "learning_rate": 8.255404632032126e-07,
+      "loss": 0.017,
+      "step": 2361
+    },
+    {
+      "epoch": 0.8204237582493922,
+      "grad_norm": 0.6228554253741776,
+      "learning_rate": 8.224464439252344e-07,
+      "loss": 0.0233,
+      "step": 2362
+    },
+    {
+      "epoch": 0.8207711010767628,
+      "grad_norm": 0.9700168786950992,
+      "learning_rate": 8.193577138583242e-07,
+      "loss": 0.0145,
+      "step": 2363
+    },
+    {
+      "epoch": 0.8211184439041334,
+      "grad_norm": 0.5794044470730007,
+      "learning_rate": 8.162742769131282e-07,
+      "loss": 0.0197,
+      "step": 2364
+    },
+    {
+      "epoch": 0.8214657867315039,
+      "grad_norm": 0.36573146405518225,
+      "learning_rate": 8.131961369935943e-07,
+      "loss": 0.0165,
+      "step": 2365
+    },
+    {
+      "epoch": 0.8218131295588746,
+      "grad_norm": 0.8465877042129827,
+      "learning_rate": 8.101232979969625e-07,
+      "loss": 0.0154,
+      "step": 2366
+    },
+    {
+      "epoch": 0.8221604723862452,
+      "grad_norm": 0.40438541583542964,
+      "learning_rate": 8.070557638137649e-07,
+      "loss": 0.0104,
+      "step": 2367
+    },
+    {
+      "epoch": 0.8225078152136158,
+      "grad_norm": 0.8024309888658482,
+      "learning_rate": 8.039935383278119e-07,
+      "loss": 0.0194,
+      "step": 2368
+    },
+    {
+      "epoch": 0.8228551580409864,
+      "grad_norm": 0.5129057990979784,
+      "learning_rate": 8.009366254161943e-07,
+      "loss": 0.0228,
+      "step": 2369
+    },
+    {
+      "epoch": 0.8232025008683571,
+      "grad_norm": 0.3255972401547693,
+      "learning_rate": 7.978850289492779e-07,
+      "loss": 0.013,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8235498436957277,
+      "grad_norm": 0.7555922278203009,
+      "learning_rate": 7.948387527906987e-07,
+      "loss": 0.0198,
+      "step": 2371
+    },
+    {
+      "epoch": 0.8238971865230983,
+      "grad_norm": 0.7146376306999533,
+      "learning_rate": 7.91797800797352e-07,
+      "loss": 0.0164,
+      "step": 2372
+    },
+    {
+      "epoch": 0.8242445293504689,
+      "grad_norm": 0.37704727475605315,
+      "learning_rate": 7.887621768193954e-07,
+      "loss": 0.0195,
+      "step": 2373
+    },
+    {
+      "epoch": 0.8245918721778396,
+      "grad_norm": 0.7235950861900677,
+      "learning_rate": 7.85731884700241e-07,
+      "loss": 0.0123,
+      "step": 2374
+    },
+    {
+      "epoch": 0.8249392150052102,
+      "grad_norm": 0.5352341742681154,
+      "learning_rate": 7.827069282765475e-07,
+      "loss": 0.0173,
+      "step": 2375
+    },
+    {
+      "epoch": 0.8252865578325808,
+      "grad_norm": 0.35228903667411676,
+      "learning_rate": 7.796873113782205e-07,
+      "loss": 0.0138,
+      "step": 2376
+    },
+    {
+      "epoch": 0.8256339006599513,
+      "grad_norm": 0.8622702937938816,
+      "learning_rate": 7.766730378284065e-07,
+      "loss": 0.0171,
+      "step": 2377
+    },
+    {
+      "epoch": 0.8259812434873219,
+      "grad_norm": 0.9369503338878897,
+      "learning_rate": 7.736641114434834e-07,
+      "loss": 0.0206,
+      "step": 2378
+    },
+    {
+      "epoch": 0.8263285863146926,
+      "grad_norm": 0.45846125634992196,
+      "learning_rate": 7.706605360330594e-07,
+      "loss": 0.0174,
+      "step": 2379
+    },
+    {
+      "epoch": 0.8266759291420632,
+      "grad_norm": 0.3955188331989464,
+      "learning_rate": 7.676623153999696e-07,
+      "loss": 0.016,
+      "step": 2380
+    },
+    {
+      "epoch": 0.8270232719694338,
+      "grad_norm": 0.7836478004454387,
+      "learning_rate": 7.646694533402699e-07,
+      "loss": 0.0261,
+      "step": 2381
+    },
+    {
+      "epoch": 0.8273706147968044,
+      "grad_norm": 0.33544601587638473,
+      "learning_rate": 7.616819536432296e-07,
+      "loss": 0.0162,
+      "step": 2382
+    },
+    {
+      "epoch": 0.8277179576241751,
+      "grad_norm": 0.7414613842115939,
+      "learning_rate": 7.586998200913282e-07,
+      "loss": 0.0209,
+      "step": 2383
+    },
+    {
+      "epoch": 0.8280653004515457,
+      "grad_norm": 0.4693067215853339,
+      "learning_rate": 7.557230564602541e-07,
+      "loss": 0.0238,
+      "step": 2384
+    },
+    {
+      "epoch": 0.8284126432789163,
+      "grad_norm": 0.43643356007744005,
+      "learning_rate": 7.527516665188956e-07,
+      "loss": 0.0112,
+      "step": 2385
+    },
+    {
+      "epoch": 0.8287599861062869,
+      "grad_norm": 0.7401366594367462,
+      "learning_rate": 7.497856540293369e-07,
+      "loss": 0.0249,
+      "step": 2386
+    },
+    {
+      "epoch": 0.8291073289336576,
+      "grad_norm": 0.6044238634021094,
+      "learning_rate": 7.468250227468515e-07,
+      "loss": 0.0167,
+      "step": 2387
+    },
+    {
+      "epoch": 0.8294546717610282,
+      "grad_norm": 0.42786196774482066,
+      "learning_rate": 7.438697764199043e-07,
+      "loss": 0.0167,
+      "step": 2388
+    },
+    {
+      "epoch": 0.8298020145883988,
+      "grad_norm": 0.2809219680220145,
+      "learning_rate": 7.409199187901417e-07,
+      "loss": 0.0103,
+      "step": 2389
+    },
+    {
+      "epoch": 0.8301493574157693,
+      "grad_norm": 0.3850053448956444,
+      "learning_rate": 7.379754535923817e-07,
+      "loss": 0.011,
+      "step": 2390
+    },
+    {
+      "epoch": 0.8304967002431399,
+      "grad_norm": 0.5129627163049738,
+      "learning_rate": 7.35036384554621e-07,
+      "loss": 0.0144,
+      "step": 2391
+    },
+    {
+      "epoch": 0.8308440430705106,
+      "grad_norm": 0.7155517690937496,
+      "learning_rate": 7.321027153980237e-07,
+      "loss": 0.0183,
+      "step": 2392
+    },
+    {
+      "epoch": 0.8311913858978812,
+      "grad_norm": 0.7899372791574505,
+      "learning_rate": 7.291744498369146e-07,
+      "loss": 0.0179,
+      "step": 2393
+    },
+    {
+      "epoch": 0.8315387287252518,
+      "grad_norm": 0.6650626503017965,
+      "learning_rate": 7.262515915787771e-07,
+      "loss": 0.0193,
+      "step": 2394
+    },
+    {
+      "epoch": 0.8318860715526224,
+      "grad_norm": 0.5901749401101161,
+      "learning_rate": 7.233341443242504e-07,
+      "loss": 0.0212,
+      "step": 2395
+    },
+    {
+      "epoch": 0.8322334143799931,
+      "grad_norm": 0.4629616119625876,
+      "learning_rate": 7.204221117671229e-07,
+      "loss": 0.0154,
+      "step": 2396
+    },
+    {
+      "epoch": 0.8325807572073637,
+      "grad_norm": 0.5047026630331894,
+      "learning_rate": 7.175154975943244e-07,
+      "loss": 0.0177,
+      "step": 2397
+    },
+    {
+      "epoch": 0.8329281000347343,
+      "grad_norm": 0.9117791956297813,
+      "learning_rate": 7.146143054859267e-07,
+      "loss": 0.025,
+      "step": 2398
+    },
+    {
+      "epoch": 0.8332754428621049,
+      "grad_norm": 0.9994693023274762,
+      "learning_rate": 7.117185391151371e-07,
+      "loss": 0.0238,
+      "step": 2399
+    },
+    {
+      "epoch": 0.8336227856894756,
+      "grad_norm": 0.2469147876407477,
+      "learning_rate": 7.088282021482934e-07,
+      "loss": 0.0069,
+      "step": 2400
+    },
+    {
+      "epoch": 0.8339701285168462,
+      "grad_norm": 0.4503780485766356,
+      "learning_rate": 7.059432982448571e-07,
+      "loss": 0.0235,
+      "step": 2401
+    },
+    {
+      "epoch": 0.8343174713442167,
+      "grad_norm": 0.39895909831018983,
+      "learning_rate": 7.030638310574123e-07,
+      "loss": 0.0139,
+      "step": 2402
+    },
+    {
+      "epoch": 0.8346648141715873,
+      "grad_norm": 0.8121294300478942,
+      "learning_rate": 7.001898042316602e-07,
+      "loss": 0.0143,
+      "step": 2403
+    },
+    {
+      "epoch": 0.8350121569989579,
+      "grad_norm": 0.5682479563314492,
+      "learning_rate": 6.97321221406414e-07,
+      "loss": 0.0254,
+      "step": 2404
+    },
+    {
+      "epoch": 0.8353594998263286,
+      "grad_norm": 0.5858957903067712,
+      "learning_rate": 6.944580862135935e-07,
+      "loss": 0.0164,
+      "step": 2405
+    },
+    {
+      "epoch": 0.8357068426536992,
+      "grad_norm": 0.8591367268668065,
+      "learning_rate": 6.916004022782191e-07,
+      "loss": 0.012,
+      "step": 2406
+    },
+    {
+      "epoch": 0.8360541854810698,
+      "grad_norm": 0.3931982527826898,
+      "learning_rate": 6.887481732184148e-07,
+      "loss": 0.0185,
+      "step": 2407
+    },
+    {
+      "epoch": 0.8364015283084404,
+      "grad_norm": 0.7603616933042572,
+      "learning_rate": 6.859014026453925e-07,
+      "loss": 0.0149,
+      "step": 2408
+    },
+    {
+      "epoch": 0.8367488711358111,
+      "grad_norm": 0.40064016500672056,
+      "learning_rate": 6.830600941634579e-07,
+      "loss": 0.0212,
+      "step": 2409
+    },
+    {
+      "epoch": 0.8370962139631817,
+      "grad_norm": 0.7197529488413131,
+      "learning_rate": 6.802242513699963e-07,
+      "loss": 0.0187,
+      "step": 2410
+    },
+    {
+      "epoch": 0.8374435567905523,
+      "grad_norm": 1.1480554116820771,
+      "learning_rate": 6.773938778554773e-07,
+      "loss": 0.015,
+      "step": 2411
+    },
+    {
+      "epoch": 0.8377908996179229,
+      "grad_norm": 0.45140844123391066,
+      "learning_rate": 6.745689772034425e-07,
+      "loss": 0.02,
+      "step": 2412
+    },
+    {
+      "epoch": 0.8381382424452936,
+      "grad_norm": 0.26189833795464385,
+      "learning_rate": 6.717495529905077e-07,
+      "loss": 0.0118,
+      "step": 2413
+    },
+    {
+      "epoch": 0.8384855852726641,
+      "grad_norm": 0.42657880248327595,
+      "learning_rate": 6.689356087863508e-07,
+      "loss": 0.0166,
+      "step": 2414
+    },
+    {
+      "epoch": 0.8388329281000347,
+      "grad_norm": 0.6022359413113759,
+      "learning_rate": 6.661271481537157e-07,
+      "loss": 0.0171,
+      "step": 2415
+    },
+    {
+      "epoch": 0.8391802709274053,
+      "grad_norm": 0.4458461340423978,
+      "learning_rate": 6.633241746483993e-07,
+      "loss": 0.0193,
+      "step": 2416
+    },
+    {
+      "epoch": 0.8395276137547759,
+      "grad_norm": 0.4109499161581392,
+      "learning_rate": 6.605266918192543e-07,
+      "loss": 0.0158,
+      "step": 2417
+    },
+    {
+      "epoch": 0.8398749565821466,
+      "grad_norm": 0.8284740415704146,
+      "learning_rate": 6.577347032081816e-07,
+      "loss": 0.0196,
+      "step": 2418
+    },
+    {
+      "epoch": 0.8402222994095172,
+      "grad_norm": 0.5823268709121688,
+      "learning_rate": 6.549482123501249e-07,
+      "loss": 0.0259,
+      "step": 2419
+    },
+    {
+      "epoch": 0.8405696422368878,
+      "grad_norm": 0.4351391472723503,
+      "learning_rate": 6.521672227730658e-07,
+      "loss": 0.0119,
+      "step": 2420
+    },
+    {
+      "epoch": 0.8409169850642584,
+      "grad_norm": 0.7478210142980828,
+      "learning_rate": 6.49391737998023e-07,
+      "loss": 0.0179,
+      "step": 2421
+    },
+    {
+      "epoch": 0.8412643278916291,
+      "grad_norm": 0.714684958708105,
+      "learning_rate": 6.466217615390468e-07,
+      "loss": 0.0152,
+      "step": 2422
+    },
+    {
+      "epoch": 0.8416116707189997,
+      "grad_norm": 0.47658348097908554,
+      "learning_rate": 6.438572969032075e-07,
+      "loss": 0.0247,
+      "step": 2423
+    },
+    {
+      "epoch": 0.8419590135463703,
+      "grad_norm": 0.5726724916032655,
+      "learning_rate": 6.410983475906024e-07,
+      "loss": 0.0213,
+      "step": 2424
+    },
+    {
+      "epoch": 0.8423063563737408,
+      "grad_norm": 0.4699141551972462,
+      "learning_rate": 6.383449170943457e-07,
+      "loss": 0.0073,
+      "step": 2425
+    },
+    {
+      "epoch": 0.8426536992011116,
+      "grad_norm": 0.48804476100038013,
+      "learning_rate": 6.355970089005615e-07,
+      "loss": 0.0189,
+      "step": 2426
+    },
+    {
+      "epoch": 0.8430010420284821,
+      "grad_norm": 0.5447605632942257,
+      "learning_rate": 6.328546264883822e-07,
+      "loss": 0.0241,
+      "step": 2427
+    },
+    {
+      "epoch": 0.8433483848558527,
+      "grad_norm": 1.1943754155338666,
+      "learning_rate": 6.301177733299457e-07,
+      "loss": 0.0231,
+      "step": 2428
+    },
+    {
+      "epoch": 0.8436957276832233,
+      "grad_norm": 0.4781400798615543,
+      "learning_rate": 6.273864528903906e-07,
+      "loss": 0.0119,
+      "step": 2429
+    },
+    {
+      "epoch": 0.8440430705105939,
+      "grad_norm": 0.6541732578812876,
+      "learning_rate": 6.246606686278467e-07,
+      "loss": 0.0275,
+      "step": 2430
+    },
+    {
+      "epoch": 0.8443904133379646,
+      "grad_norm": 0.40455279688964085,
+      "learning_rate": 6.219404239934357e-07,
+      "loss": 0.0115,
+      "step": 2431
+    },
+    {
+      "epoch": 0.8447377561653352,
+      "grad_norm": 0.6875625753644061,
+      "learning_rate": 6.19225722431267e-07,
+      "loss": 0.0187,
+      "step": 2432
+    },
+    {
+      "epoch": 0.8450850989927058,
+      "grad_norm": 0.7921616555433921,
+      "learning_rate": 6.165165673784318e-07,
+      "loss": 0.0171,
+      "step": 2433
+    },
+    {
+      "epoch": 0.8454324418200764,
+      "grad_norm": 0.8867189304336157,
+      "learning_rate": 6.13812962264998e-07,
+      "loss": 0.0236,
+      "step": 2434
+    },
+    {
+      "epoch": 0.8457797846474471,
+      "grad_norm": 1.115479157352609,
+      "learning_rate": 6.111149105140052e-07,
+      "loss": 0.0177,
+      "step": 2435
+    },
+    {
+      "epoch": 0.8461271274748177,
+      "grad_norm": 0.5543931361704746,
+      "learning_rate": 6.084224155414647e-07,
+      "loss": 0.0105,
+      "step": 2436
+    },
+    {
+      "epoch": 0.8464744703021883,
+      "grad_norm": 0.48262329533304676,
+      "learning_rate": 6.057354807563526e-07,
+      "loss": 0.0109,
+      "step": 2437
+    },
+    {
+      "epoch": 0.8468218131295588,
+      "grad_norm": 0.41747137999997436,
+      "learning_rate": 6.030541095606018e-07,
+      "loss": 0.011,
+      "step": 2438
+    },
+    {
+      "epoch": 0.8471691559569295,
+      "grad_norm": 0.49664059634379965,
+      "learning_rate": 6.003783053491025e-07,
+      "loss": 0.013,
+      "step": 2439
+    },
+    {
+      "epoch": 0.8475164987843001,
+      "grad_norm": 0.4877777397413885,
+      "learning_rate": 5.977080715096995e-07,
+      "loss": 0.0232,
+      "step": 2440
+    },
+    {
+      "epoch": 0.8478638416116707,
+      "grad_norm": 1.2489307888455528,
+      "learning_rate": 5.950434114231801e-07,
+      "loss": 0.0314,
+      "step": 2441
+    },
+    {
+      "epoch": 0.8482111844390413,
+      "grad_norm": 0.4900829237700161,
+      "learning_rate": 5.923843284632796e-07,
+      "loss": 0.0184,
+      "step": 2442
+    },
+    {
+      "epoch": 0.8485585272664119,
+      "grad_norm": 0.4575837010051548,
+      "learning_rate": 5.897308259966672e-07,
+      "loss": 0.0176,
+      "step": 2443
+    },
+    {
+      "epoch": 0.8489058700937826,
+      "grad_norm": 0.38584962250720306,
+      "learning_rate": 5.870829073829515e-07,
+      "loss": 0.0179,
+      "step": 2444
+    },
+    {
+      "epoch": 0.8492532129211532,
+      "grad_norm": 0.4177353453219893,
+      "learning_rate": 5.844405759746663e-07,
+      "loss": 0.013,
+      "step": 2445
+    },
+    {
+      "epoch": 0.8496005557485238,
+      "grad_norm": 0.4869716570678539,
+      "learning_rate": 5.818038351172767e-07,
+      "loss": 0.0191,
+      "step": 2446
+    },
+    {
+      "epoch": 0.8499478985758944,
+      "grad_norm": 0.6783571708436903,
+      "learning_rate": 5.791726881491644e-07,
+      "loss": 0.0213,
+      "step": 2447
+    },
+    {
+      "epoch": 0.8502952414032651,
+      "grad_norm": 0.6897537136000237,
+      "learning_rate": 5.765471384016341e-07,
+      "loss": 0.0162,
+      "step": 2448
+    },
+    {
+      "epoch": 0.8506425842306357,
+      "grad_norm": 0.44891761410994796,
+      "learning_rate": 5.739271891988974e-07,
+      "loss": 0.0214,
+      "step": 2449
+    },
+    {
+      "epoch": 0.8509899270580062,
+      "grad_norm": 1.0985759896890122,
+      "learning_rate": 5.713128438580823e-07,
+      "loss": 0.0229,
+      "step": 2450
+    },
+    {
+      "epoch": 0.8513372698853768,
+      "grad_norm": 0.4352898494688454,
+      "learning_rate": 5.687041056892145e-07,
+      "loss": 0.0143,
+      "step": 2451
+    },
+    {
+      "epoch": 0.8516846127127475,
+      "grad_norm": 0.5789110156147614,
+      "learning_rate": 5.66100977995227e-07,
+      "loss": 0.0201,
+      "step": 2452
+    },
+    {
+      "epoch": 0.8520319555401181,
+      "grad_norm": 0.6518621410478981,
+      "learning_rate": 5.635034640719433e-07,
+      "loss": 0.0216,
+      "step": 2453
+    },
+    {
+      "epoch": 0.8523792983674887,
+      "grad_norm": 0.9846895298135566,
+      "learning_rate": 5.609115672080845e-07,
+      "loss": 0.018,
+      "step": 2454
+    },
+    {
+      "epoch": 0.8527266411948593,
+      "grad_norm": 0.8407928319843598,
+      "learning_rate": 5.583252906852594e-07,
+      "loss": 0.0144,
+      "step": 2455
+    },
+    {
+      "epoch": 0.8530739840222299,
+      "grad_norm": 0.5601226706008987,
+      "learning_rate": 5.557446377779546e-07,
+      "loss": 0.0215,
+      "step": 2456
+    },
+    {
+      "epoch": 0.8534213268496006,
+      "grad_norm": 0.5001353143572673,
+      "learning_rate": 5.53169611753544e-07,
+      "loss": 0.0183,
+      "step": 2457
+    },
+    {
+      "epoch": 0.8537686696769712,
+      "grad_norm": 0.5236537168132929,
+      "learning_rate": 5.506002158722751e-07,
+      "loss": 0.0218,
+      "step": 2458
+    },
+    {
+      "epoch": 0.8541160125043418,
+      "grad_norm": 0.36795187111448785,
+      "learning_rate": 5.48036453387265e-07,
+      "loss": 0.0129,
+      "step": 2459
+    },
+    {
+      "epoch": 0.8544633553317124,
+      "grad_norm": 0.7797517640305994,
+      "learning_rate": 5.454783275445003e-07,
+      "loss": 0.0167,
+      "step": 2460
+    },
+    {
+      "epoch": 0.8548106981590831,
+      "grad_norm": 0.40671604163293146,
+      "learning_rate": 5.429258415828298e-07,
+      "loss": 0.0179,
+      "step": 2461
+    },
+    {
+      "epoch": 0.8551580409864536,
+      "grad_norm": 0.8392548489754016,
+      "learning_rate": 5.403789987339647e-07,
+      "loss": 0.0164,
+      "step": 2462
+    },
+    {
+      "epoch": 0.8555053838138242,
+      "grad_norm": 1.8042310161892017,
+      "learning_rate": 5.378378022224679e-07,
+      "loss": 0.0182,
+      "step": 2463
+    },
+    {
+      "epoch": 0.8558527266411948,
+      "grad_norm": 0.9597712322684139,
+      "learning_rate": 5.353022552657533e-07,
+      "loss": 0.0242,
+      "step": 2464
+    },
+    {
+      "epoch": 0.8562000694685655,
+      "grad_norm": 1.2902699361504208,
+      "learning_rate": 5.327723610740843e-07,
+      "loss": 0.0194,
+      "step": 2465
+    },
+    {
+      "epoch": 0.8565474122959361,
+      "grad_norm": 0.9025915486754845,
+      "learning_rate": 5.302481228505674e-07,
+      "loss": 0.0159,
+      "step": 2466
+    },
+    {
+      "epoch": 0.8568947551233067,
+      "grad_norm": 0.8935041671025581,
+      "learning_rate": 5.277295437911462e-07,
+      "loss": 0.0206,
+      "step": 2467
+    },
+    {
+      "epoch": 0.8572420979506773,
+      "grad_norm": 0.8398011042782978,
+      "learning_rate": 5.252166270845994e-07,
+      "loss": 0.0175,
+      "step": 2468
+    },
+    {
+      "epoch": 0.8575894407780479,
+      "grad_norm": 0.6273784672353577,
+      "learning_rate": 5.227093759125368e-07,
+      "loss": 0.013,
+      "step": 2469
+    },
+    {
+      "epoch": 0.8579367836054186,
+      "grad_norm": 0.7599839379620519,
+      "learning_rate": 5.20207793449397e-07,
+      "loss": 0.0148,
+      "step": 2470
+    },
+    {
+      "epoch": 0.8582841264327892,
+      "grad_norm": 0.9137884963550477,
+      "learning_rate": 5.177118828624395e-07,
+      "loss": 0.0177,
+      "step": 2471
+    },
+    {
+      "epoch": 0.8586314692601598,
+      "grad_norm": 0.4027490941272288,
+      "learning_rate": 5.152216473117416e-07,
+      "loss": 0.0115,
+      "step": 2472
+    },
+    {
+      "epoch": 0.8589788120875304,
+      "grad_norm": 0.4067939554578114,
+      "learning_rate": 5.127370899501988e-07,
+      "loss": 0.0132,
+      "step": 2473
+    },
+    {
+      "epoch": 0.859326154914901,
+      "grad_norm": 0.39233639414180854,
+      "learning_rate": 5.10258213923513e-07,
+      "loss": 0.0134,
+      "step": 2474
+    },
+    {
+      "epoch": 0.8596734977422716,
+      "grad_norm": 0.5636819438029829,
+      "learning_rate": 5.07785022370198e-07,
+      "loss": 0.0162,
+      "step": 2475
+    },
+    {
+      "epoch": 0.8600208405696422,
+      "grad_norm": 0.45051474131353364,
+      "learning_rate": 5.053175184215653e-07,
+      "loss": 0.019,
+      "step": 2476
+    },
+    {
+      "epoch": 0.8603681833970128,
+      "grad_norm": 0.504428208604678,
+      "learning_rate": 5.028557052017302e-07,
+      "loss": 0.0128,
+      "step": 2477
+    },
+    {
+      "epoch": 0.8607155262243835,
+      "grad_norm": 0.5002145933355393,
+      "learning_rate": 5.003995858275984e-07,
+      "loss": 0.0148,
+      "step": 2478
+    },
+    {
+      "epoch": 0.8610628690517541,
+      "grad_norm": 0.7066450698383614,
+      "learning_rate": 4.979491634088712e-07,
+      "loss": 0.0216,
+      "step": 2479
+    },
+    {
+      "epoch": 0.8614102118791247,
+      "grad_norm": 0.7209554287180902,
+      "learning_rate": 4.955044410480326e-07,
+      "loss": 0.0102,
+      "step": 2480
+    },
+    {
+      "epoch": 0.8617575547064953,
+      "grad_norm": 0.37457104151587356,
+      "learning_rate": 4.93065421840353e-07,
+      "loss": 0.0129,
+      "step": 2481
+    },
+    {
+      "epoch": 0.8621048975338659,
+      "grad_norm": 0.35405427472628026,
+      "learning_rate": 4.906321088738791e-07,
+      "loss": 0.0145,
+      "step": 2482
+    },
+    {
+      "epoch": 0.8624522403612366,
+      "grad_norm": 0.9145616386018184,
+      "learning_rate": 4.882045052294371e-07,
+      "loss": 0.0154,
+      "step": 2483
+    },
+    {
+      "epoch": 0.8627995831886072,
+      "grad_norm": 0.45595222296109655,
+      "learning_rate": 4.857826139806194e-07,
+      "loss": 0.0159,
+      "step": 2484
+    },
+    {
+      "epoch": 0.8631469260159778,
+      "grad_norm": 0.6853468761771402,
+      "learning_rate": 4.833664381937908e-07,
+      "loss": 0.015,
+      "step": 2485
+    },
+    {
+      "epoch": 0.8634942688433483,
+      "grad_norm": 0.9539717763403811,
+      "learning_rate": 4.809559809280756e-07,
+      "loss": 0.0173,
+      "step": 2486
+    },
+    {
+      "epoch": 0.863841611670719,
+      "grad_norm": 0.8437730803188537,
+      "learning_rate": 4.785512452353619e-07,
+      "loss": 0.0244,
+      "step": 2487
+    },
+    {
+      "epoch": 0.8641889544980896,
+      "grad_norm": 0.7966421861571771,
+      "learning_rate": 4.7615223416029086e-07,
+      "loss": 0.0232,
+      "step": 2488
+    },
+    {
+      "epoch": 0.8645362973254602,
+      "grad_norm": 0.5829026692062281,
+      "learning_rate": 4.737589507402546e-07,
+      "loss": 0.0191,
+      "step": 2489
+    },
+    {
+      "epoch": 0.8648836401528308,
+      "grad_norm": 0.6971309820400614,
+      "learning_rate": 4.7137139800539746e-07,
+      "loss": 0.0147,
+      "step": 2490
+    },
+    {
+      "epoch": 0.8652309829802015,
+      "grad_norm": 0.9730338109684425,
+      "learning_rate": 4.689895789786059e-07,
+      "loss": 0.027,
+      "step": 2491
+    },
+    {
+      "epoch": 0.8655783258075721,
+      "grad_norm": 0.46701122458001026,
+      "learning_rate": 4.666134966755059e-07,
+      "loss": 0.0179,
+      "step": 2492
+    },
+    {
+      "epoch": 0.8659256686349427,
+      "grad_norm": 0.7261076983049326,
+      "learning_rate": 4.6424315410446117e-07,
+      "loss": 0.0231,
+      "step": 2493
+    },
+    {
+      "epoch": 0.8662730114623133,
+      "grad_norm": 0.4333800390573832,
+      "learning_rate": 4.618785542665688e-07,
+      "loss": 0.0171,
+      "step": 2494
+    },
+    {
+      "epoch": 0.8666203542896839,
+      "grad_norm": 0.7645732325430994,
+      "learning_rate": 4.5951970015565617e-07,
+      "loss": 0.0152,
+      "step": 2495
+    },
+    {
+      "epoch": 0.8669676971170546,
+      "grad_norm": 1.7725027824937925,
+      "learning_rate": 4.571665947582726e-07,
+      "loss": 0.0164,
+      "step": 2496
+    },
+    {
+      "epoch": 0.8673150399444252,
+      "grad_norm": 0.5820223340143823,
+      "learning_rate": 4.5481924105369e-07,
+      "loss": 0.0147,
+      "step": 2497
+    },
+    {
+      "epoch": 0.8676623827717957,
+      "grad_norm": 0.9426845874078936,
+      "learning_rate": 4.5247764201390045e-07,
+      "loss": 0.0181,
+      "step": 2498
+    },
+    {
+      "epoch": 0.8680097255991663,
+      "grad_norm": 0.40736158785553966,
+      "learning_rate": 4.5014180060360843e-07,
+      "loss": 0.0161,
+      "step": 2499
+    },
+    {
+      "epoch": 0.868357068426537,
+      "grad_norm": 0.35291016663434527,
+      "learning_rate": 4.4781171978022786e-07,
+      "loss": 0.0136,
+      "step": 2500
+    },
+    {
+      "epoch": 0.8687044112539076,
+      "grad_norm": 0.7083995120920576,
+      "learning_rate": 4.4548740249387934e-07,
+      "loss": 0.0287,
+      "step": 2501
+    },
+    {
+      "epoch": 0.8690517540812782,
+      "grad_norm": 0.6893058334285791,
+      "learning_rate": 4.4316885168738776e-07,
+      "loss": 0.0191,
+      "step": 2502
+    },
+    {
+      "epoch": 0.8693990969086488,
+      "grad_norm": 0.4921717937501951,
+      "learning_rate": 4.4085607029627717e-07,
+      "loss": 0.0112,
+      "step": 2503
+    },
+    {
+      "epoch": 0.8697464397360195,
+      "grad_norm": 0.5902281066197675,
+      "learning_rate": 4.3854906124876415e-07,
+      "loss": 0.0157,
+      "step": 2504
+    },
+    {
+      "epoch": 0.8700937825633901,
+      "grad_norm": 0.5650017177846125,
+      "learning_rate": 4.3624782746575886e-07,
+      "loss": 0.0221,
+      "step": 2505
+    },
+    {
+      "epoch": 0.8704411253907607,
+      "grad_norm": 0.5938641449620281,
+      "learning_rate": 4.3395237186086014e-07,
+      "loss": 0.0119,
+      "step": 2506
+    },
+    {
+      "epoch": 0.8707884682181313,
+      "grad_norm": 0.44786371675122244,
+      "learning_rate": 4.316626973403487e-07,
+      "loss": 0.0138,
+      "step": 2507
+    },
+    {
+      "epoch": 0.8711358110455019,
+      "grad_norm": 0.529485412783005,
+      "learning_rate": 4.2937880680318846e-07,
+      "loss": 0.0203,
+      "step": 2508
+    },
+    {
+      "epoch": 0.8714831538728726,
+      "grad_norm": 0.4695178384130346,
+      "learning_rate": 4.2710070314101845e-07,
+      "loss": 0.0171,
+      "step": 2509
+    },
+    {
+      "epoch": 0.8718304967002432,
+      "grad_norm": 0.6865782270212439,
+      "learning_rate": 4.2482838923815163e-07,
+      "loss": 0.0221,
+      "step": 2510
+    },
+    {
+      "epoch": 0.8721778395276137,
+      "grad_norm": 1.008599802575518,
+      "learning_rate": 4.2256186797156986e-07,
+      "loss": 0.0163,
+      "step": 2511
+    },
+    {
+      "epoch": 0.8725251823549843,
+      "grad_norm": 0.8277810535713074,
+      "learning_rate": 4.203011422109227e-07,
+      "loss": 0.0131,
+      "step": 2512
+    },
+    {
+      "epoch": 0.872872525182355,
+      "grad_norm": 0.47209426725218306,
+      "learning_rate": 4.180462148185188e-07,
+      "loss": 0.0115,
+      "step": 2513
+    },
+    {
+      "epoch": 0.8732198680097256,
+      "grad_norm": 1.0335741947460342,
+      "learning_rate": 4.1579708864932956e-07,
+      "loss": 0.0202,
+      "step": 2514
+    },
+    {
+      "epoch": 0.8735672108370962,
+      "grad_norm": 0.4371396632860345,
+      "learning_rate": 4.1355376655097704e-07,
+      "loss": 0.0187,
+      "step": 2515
+    },
+    {
+      "epoch": 0.8739145536644668,
+      "grad_norm": 0.4841174212281427,
+      "learning_rate": 4.113162513637392e-07,
+      "loss": 0.0146,
+      "step": 2516
+    },
+    {
+      "epoch": 0.8742618964918375,
+      "grad_norm": 0.5418422212985032,
+      "learning_rate": 4.090845459205378e-07,
+      "loss": 0.0205,
+      "step": 2517
+    },
+    {
+      "epoch": 0.8746092393192081,
+      "grad_norm": 0.6175530773760641,
+      "learning_rate": 4.0685865304694205e-07,
+      "loss": 0.0257,
+      "step": 2518
+    },
+    {
+      "epoch": 0.8749565821465787,
+      "grad_norm": 0.6600694577668837,
+      "learning_rate": 4.0463857556115924e-07,
+      "loss": 0.0126,
+      "step": 2519
+    },
+    {
+      "epoch": 0.8753039249739493,
+      "grad_norm": 0.45151100408746436,
+      "learning_rate": 4.0242431627403656e-07,
+      "loss": 0.0108,
+      "step": 2520
+    },
+    {
+      "epoch": 0.8756512678013199,
+      "grad_norm": 0.7267112922014047,
+      "learning_rate": 4.0021587798905247e-07,
+      "loss": 0.0165,
+      "step": 2521
+    },
+    {
+      "epoch": 0.8759986106286906,
+      "grad_norm": 0.7989326256489759,
+      "learning_rate": 3.980132635023154e-07,
+      "loss": 0.022,
+      "step": 2522
+    },
+    {
+      "epoch": 0.8763459534560611,
+      "grad_norm": 1.297364986774049,
+      "learning_rate": 3.9581647560256175e-07,
+      "loss": 0.0218,
+      "step": 2523
+    },
+    {
+      "epoch": 0.8766932962834317,
+      "grad_norm": 0.5435363177683816,
+      "learning_rate": 3.9362551707115114e-07,
+      "loss": 0.0247,
+      "step": 2524
+    },
+    {
+      "epoch": 0.8770406391108023,
+      "grad_norm": 0.38578321310369523,
+      "learning_rate": 3.914403906820613e-07,
+      "loss": 0.0133,
+      "step": 2525
+    },
+    {
+      "epoch": 0.877387981938173,
+      "grad_norm": 0.4932282625829498,
+      "learning_rate": 3.892610992018847e-07,
+      "loss": 0.016,
+      "step": 2526
+    },
+    {
+      "epoch": 0.8777353247655436,
+      "grad_norm": 0.3669442115959157,
+      "learning_rate": 3.870876453898292e-07,
+      "loss": 0.0166,
+      "step": 2527
+    },
+    {
+      "epoch": 0.8780826675929142,
+      "grad_norm": 0.9342073340109053,
+      "learning_rate": 3.849200319977109e-07,
+      "loss": 0.0228,
+      "step": 2528
+    },
+    {
+      "epoch": 0.8784300104202848,
+      "grad_norm": 0.25770986108617316,
+      "learning_rate": 3.8275826176994936e-07,
+      "loss": 0.0092,
+      "step": 2529
+    },
+    {
+      "epoch": 0.8787773532476555,
+      "grad_norm": 0.4339581883005602,
+      "learning_rate": 3.8060233744356634e-07,
+      "loss": 0.0168,
+      "step": 2530
+    },
+    {
+      "epoch": 0.8791246960750261,
+      "grad_norm": 0.9289029914869522,
+      "learning_rate": 3.784522617481845e-07,
+      "loss": 0.0215,
+      "step": 2531
+    },
+    {
+      "epoch": 0.8794720389023967,
+      "grad_norm": 0.5849896866294249,
+      "learning_rate": 3.7630803740602073e-07,
+      "loss": 0.0224,
+      "step": 2532
+    },
+    {
+      "epoch": 0.8798193817297673,
+      "grad_norm": 0.7327664241177156,
+      "learning_rate": 3.7416966713188174e-07,
+      "loss": 0.0106,
+      "step": 2533
+    },
+    {
+      "epoch": 0.8801667245571378,
+      "grad_norm": 0.3696886941163735,
+      "learning_rate": 3.7203715363316294e-07,
+      "loss": 0.0089,
+      "step": 2534
+    },
+    {
+      "epoch": 0.8805140673845085,
+      "grad_norm": 0.49024463460685797,
+      "learning_rate": 3.699104996098457e-07,
+      "loss": 0.0151,
+      "step": 2535
+    },
+    {
+      "epoch": 0.8808614102118791,
+      "grad_norm": 0.7947374546207403,
+      "learning_rate": 3.6778970775449283e-07,
+      "loss": 0.0142,
+      "step": 2536
+    },
+    {
+      "epoch": 0.8812087530392497,
+      "grad_norm": 0.4625649202944896,
+      "learning_rate": 3.656747807522437e-07,
+      "loss": 0.0109,
+      "step": 2537
+    },
+    {
+      "epoch": 0.8815560958666203,
+      "grad_norm": 0.5558918378768948,
+      "learning_rate": 3.6356572128081134e-07,
+      "loss": 0.0238,
+      "step": 2538
+    },
+    {
+      "epoch": 0.881903438693991,
+      "grad_norm": 0.42918432489972863,
+      "learning_rate": 3.614625320104831e-07,
+      "loss": 0.0153,
+      "step": 2539
+    },
+    {
+      "epoch": 0.8822507815213616,
+      "grad_norm": 1.027886704845079,
+      "learning_rate": 3.593652156041122e-07,
+      "loss": 0.0195,
+      "step": 2540
+    },
+    {
+      "epoch": 0.8825981243487322,
+      "grad_norm": 0.33562683798972165,
+      "learning_rate": 3.572737747171151e-07,
+      "loss": 0.0125,
+      "step": 2541
+    },
+    {
+      "epoch": 0.8829454671761028,
+      "grad_norm": 0.5951105843622697,
+      "learning_rate": 3.5518821199747035e-07,
+      "loss": 0.0175,
+      "step": 2542
+    },
+    {
+      "epoch": 0.8832928100034735,
+      "grad_norm": 0.6577768489226572,
+      "learning_rate": 3.531085300857151e-07,
+      "loss": 0.0252,
+      "step": 2543
+    },
+    {
+      "epoch": 0.8836401528308441,
+      "grad_norm": 1.1447374774363468,
+      "learning_rate": 3.510347316149393e-07,
+      "loss": 0.0253,
+      "step": 2544
+    },
+    {
+      "epoch": 0.8839874956582147,
+      "grad_norm": 1.075748573924704,
+      "learning_rate": 3.4896681921078477e-07,
+      "loss": 0.0138,
+      "step": 2545
+    },
+    {
+      "epoch": 0.8843348384855853,
+      "grad_norm": 0.5241854815133766,
+      "learning_rate": 3.469047954914395e-07,
+      "loss": 0.0178,
+      "step": 2546
+    },
+    {
+      "epoch": 0.8846821813129558,
+      "grad_norm": 0.48865176944956945,
+      "learning_rate": 3.4484866306763896e-07,
+      "loss": 0.0123,
+      "step": 2547
+    },
+    {
+      "epoch": 0.8850295241403265,
+      "grad_norm": 0.48077206825883056,
+      "learning_rate": 3.4279842454265523e-07,
+      "loss": 0.0114,
+      "step": 2548
+    },
+    {
+      "epoch": 0.8853768669676971,
+      "grad_norm": 0.7919381272023467,
+      "learning_rate": 3.407540825123024e-07,
+      "loss": 0.015,
+      "step": 2549
+    },
+    {
+      "epoch": 0.8857242097950677,
+      "grad_norm": 0.9115719414888127,
+      "learning_rate": 3.3871563956492546e-07,
+      "loss": 0.0215,
+      "step": 2550
+    },
+    {
+      "epoch": 0.8860715526224383,
+      "grad_norm": 0.9428448234794181,
+      "learning_rate": 3.36683098281404e-07,
+      "loss": 0.019,
+      "step": 2551
+    },
+    {
+      "epoch": 0.886418895449809,
+      "grad_norm": 0.5725716009079287,
+      "learning_rate": 3.346564612351416e-07,
+      "loss": 0.0221,
+      "step": 2552
+    },
+    {
+      "epoch": 0.8867662382771796,
+      "grad_norm": 0.5339812873256146,
+      "learning_rate": 3.3263573099207025e-07,
+      "loss": 0.0156,
+      "step": 2553
+    },
+    {
+      "epoch": 0.8871135811045502,
+      "grad_norm": 0.7200318917306987,
+      "learning_rate": 3.3062091011064e-07,
+      "loss": 0.0189,
+      "step": 2554
+    },
+    {
+      "epoch": 0.8874609239319208,
+      "grad_norm": 0.8289865754088781,
+      "learning_rate": 3.2861200114182257e-07,
+      "loss": 0.0191,
+      "step": 2555
+    },
+    {
+      "epoch": 0.8878082667592915,
+      "grad_norm": 0.41087275655115824,
+      "learning_rate": 3.2660900662910056e-07,
+      "loss": 0.014,
+      "step": 2556
+    },
+    {
+      "epoch": 0.8881556095866621,
+      "grad_norm": 0.4168171217347719,
+      "learning_rate": 3.2461192910847263e-07,
+      "loss": 0.0127,
+      "step": 2557
+    },
+    {
+      "epoch": 0.8885029524140327,
+      "grad_norm": 0.3718290081453453,
+      "learning_rate": 3.2262077110844224e-07,
+      "loss": 0.0094,
+      "step": 2558
+    },
+    {
+      "epoch": 0.8888502952414032,
+      "grad_norm": 0.32435438005787876,
+      "learning_rate": 3.206355351500184e-07,
+      "loss": 0.0123,
+      "step": 2559
+    },
+    {
+      "epoch": 0.8891976380687738,
+      "grad_norm": 0.4727940603046291,
+      "learning_rate": 3.186562237467156e-07,
+      "loss": 0.014,
+      "step": 2560
+    },
+    {
+      "epoch": 0.8895449808961445,
+      "grad_norm": 0.3199322989762899,
+      "learning_rate": 3.16682839404544e-07,
+      "loss": 0.0156,
+      "step": 2561
+    },
+    {
+      "epoch": 0.8898923237235151,
+      "grad_norm": 0.5052862572229199,
+      "learning_rate": 3.147153846220108e-07,
+      "loss": 0.0162,
+      "step": 2562
+    },
+    {
+      "epoch": 0.8902396665508857,
+      "grad_norm": 0.4935045434697062,
+      "learning_rate": 3.127538618901144e-07,
+      "loss": 0.0201,
+      "step": 2563
+    },
+    {
+      "epoch": 0.8905870093782563,
+      "grad_norm": 0.45484641731771275,
+      "learning_rate": 3.107982736923448e-07,
+      "loss": 0.012,
+      "step": 2564
+    },
+    {
+      "epoch": 0.890934352205627,
+      "grad_norm": 0.4589036891401816,
+      "learning_rate": 3.0884862250467715e-07,
+      "loss": 0.0185,
+      "step": 2565
+    },
+    {
+      "epoch": 0.8912816950329976,
+      "grad_norm": 0.4845147421338059,
+      "learning_rate": 3.069049107955696e-07,
+      "loss": 0.0122,
+      "step": 2566
+    },
+    {
+      "epoch": 0.8916290378603682,
+      "grad_norm": 0.4953532759104348,
+      "learning_rate": 3.0496714102595914e-07,
+      "loss": 0.0105,
+      "step": 2567
+    },
+    {
+      "epoch": 0.8919763806877388,
+      "grad_norm": 1.1581053470310272,
+      "learning_rate": 3.030353156492627e-07,
+      "loss": 0.0217,
+      "step": 2568
+    },
+    {
+      "epoch": 0.8923237235151095,
+      "grad_norm": 0.5167371489599756,
+      "learning_rate": 3.0110943711136874e-07,
+      "loss": 0.0119,
+      "step": 2569
+    },
+    {
+      "epoch": 0.8926710663424801,
+      "grad_norm": 0.6648490939529547,
+      "learning_rate": 2.9918950785063684e-07,
+      "loss": 0.015,
+      "step": 2570
+    },
+    {
+      "epoch": 0.8930184091698506,
+      "grad_norm": 0.7019645201421016,
+      "learning_rate": 2.9727553029789303e-07,
+      "loss": 0.019,
+      "step": 2571
+    },
+    {
+      "epoch": 0.8933657519972212,
+      "grad_norm": 0.3570823917219566,
+      "learning_rate": 2.953675068764311e-07,
+      "loss": 0.0123,
+      "step": 2572
+    },
+    {
+      "epoch": 0.8937130948245918,
+      "grad_norm": 0.44958846952146136,
+      "learning_rate": 2.9346544000200373e-07,
+      "loss": 0.0175,
+      "step": 2573
+    },
+    {
+      "epoch": 0.8940604376519625,
+      "grad_norm": 0.8033961501217101,
+      "learning_rate": 2.915693320828222e-07,
+      "loss": 0.0245,
+      "step": 2574
+    },
+    {
+      "epoch": 0.8944077804793331,
+      "grad_norm": 0.6221815609588339,
+      "learning_rate": 2.89679185519553e-07,
+      "loss": 0.0164,
+      "step": 2575
+    },
+    {
+      "epoch": 0.8947551233067037,
+      "grad_norm": 0.8670700629752854,
+      "learning_rate": 2.877950027053167e-07,
+      "loss": 0.0251,
+      "step": 2576
+    },
+    {
+      "epoch": 0.8951024661340743,
+      "grad_norm": 0.33526691551846566,
+      "learning_rate": 2.859167860256801e-07,
+      "loss": 0.0145,
+      "step": 2577
+    },
+    {
+      "epoch": 0.895449808961445,
+      "grad_norm": 0.3738576749343047,
+      "learning_rate": 2.8404453785866037e-07,
+      "loss": 0.0126,
+      "step": 2578
+    },
+    {
+      "epoch": 0.8957971517888156,
+      "grad_norm": 0.6834813763355805,
+      "learning_rate": 2.8217826057471423e-07,
+      "loss": 0.0273,
+      "step": 2579
+    },
+    {
+      "epoch": 0.8961444946161862,
+      "grad_norm": 0.47554504790974056,
+      "learning_rate": 2.8031795653674033e-07,
+      "loss": 0.0161,
+      "step": 2580
+    },
+    {
+      "epoch": 0.8964918374435568,
+      "grad_norm": 0.3690282253450739,
+      "learning_rate": 2.7846362810007355e-07,
+      "loss": 0.0151,
+      "step": 2581
+    },
+    {
+      "epoch": 0.8968391802709275,
+      "grad_norm": 0.7307835358362869,
+      "learning_rate": 2.766152776124853e-07,
+      "loss": 0.0207,
+      "step": 2582
+    },
+    {
+      "epoch": 0.897186523098298,
+      "grad_norm": 0.755714402406654,
+      "learning_rate": 2.7477290741417526e-07,
+      "loss": 0.0172,
+      "step": 2583
+    },
+    {
+      "epoch": 0.8975338659256686,
+      "grad_norm": 0.43973806244842545,
+      "learning_rate": 2.729365198377748e-07,
+      "loss": 0.0147,
+      "step": 2584
+    },
+    {
+      "epoch": 0.8978812087530392,
+      "grad_norm": 0.9009144598568006,
+      "learning_rate": 2.711061172083368e-07,
+      "loss": 0.0192,
+      "step": 2585
+    },
+    {
+      "epoch": 0.8982285515804098,
+      "grad_norm": 0.8203741129744253,
+      "learning_rate": 2.692817018433397e-07,
+      "loss": 0.0147,
+      "step": 2586
+    },
+    {
+      "epoch": 0.8985758944077805,
+      "grad_norm": 0.8664237660996313,
+      "learning_rate": 2.6746327605268017e-07,
+      "loss": 0.018,
+      "step": 2587
+    },
+    {
+      "epoch": 0.8989232372351511,
+      "grad_norm": 0.49174214473254213,
+      "learning_rate": 2.656508421386722e-07,
+      "loss": 0.019,
+      "step": 2588
+    },
+    {
+      "epoch": 0.8992705800625217,
+      "grad_norm": 0.41984596348114944,
+      "learning_rate": 2.638444023960418e-07,
+      "loss": 0.0147,
+      "step": 2589
+    },
+    {
+      "epoch": 0.8996179228898923,
+      "grad_norm": 0.6425656015149634,
+      "learning_rate": 2.6204395911192836e-07,
+      "loss": 0.0142,
+      "step": 2590
+    },
+    {
+      "epoch": 0.899965265717263,
+      "grad_norm": 0.6500871093127806,
+      "learning_rate": 2.6024951456587677e-07,
+      "loss": 0.0146,
+      "step": 2591
+    },
+    {
+      "epoch": 0.9003126085446336,
+      "grad_norm": 0.7086875286298896,
+      "learning_rate": 2.5846107102983744e-07,
+      "loss": 0.0243,
+      "step": 2592
+    },
+    {
+      "epoch": 0.9006599513720042,
+      "grad_norm": 0.38145659806209425,
+      "learning_rate": 2.566786307681635e-07,
+      "loss": 0.0131,
+      "step": 2593
+    },
+    {
+      "epoch": 0.9010072941993748,
+      "grad_norm": 0.6205010960638877,
+      "learning_rate": 2.549021960376075e-07,
+      "loss": 0.015,
+      "step": 2594
+    },
+    {
+      "epoch": 0.9013546370267453,
+      "grad_norm": 0.5706136061121067,
+      "learning_rate": 2.531317690873181e-07,
+      "loss": 0.0197,
+      "step": 2595
+    },
+    {
+      "epoch": 0.901701979854116,
+      "grad_norm": 0.3901107728118566,
+      "learning_rate": 2.5136735215883613e-07,
+      "loss": 0.0095,
+      "step": 2596
+    },
+    {
+      "epoch": 0.9020493226814866,
+      "grad_norm": 0.5881915928887412,
+      "learning_rate": 2.4960894748609465e-07,
+      "loss": 0.0202,
+      "step": 2597
+    },
+    {
+      "epoch": 0.9023966655088572,
+      "grad_norm": 0.42911566159079534,
+      "learning_rate": 2.4785655729541555e-07,
+      "loss": 0.0159,
+      "step": 2598
+    },
+    {
+      "epoch": 0.9027440083362278,
+      "grad_norm": 0.27083599132726704,
+      "learning_rate": 2.46110183805503e-07,
+      "loss": 0.0074,
+      "step": 2599
+    },
+    {
+      "epoch": 0.9030913511635985,
+      "grad_norm": 0.6530377955588946,
+      "learning_rate": 2.4436982922744547e-07,
+      "loss": 0.0168,
+      "step": 2600
+    },
+    {
+      "epoch": 0.9034386939909691,
+      "grad_norm": 0.8639097600309369,
+      "learning_rate": 2.426354957647098e-07,
+      "loss": 0.0139,
+      "step": 2601
+    },
+    {
+      "epoch": 0.9037860368183397,
+      "grad_norm": 0.5654469170463523,
+      "learning_rate": 2.4090718561314064e-07,
+      "loss": 0.0218,
+      "step": 2602
+    },
+    {
+      "epoch": 0.9041333796457103,
+      "grad_norm": 0.8583056813584015,
+      "learning_rate": 2.391849009609559e-07,
+      "loss": 0.0109,
+      "step": 2603
+    },
+    {
+      "epoch": 0.904480722473081,
+      "grad_norm": 0.6257033656528606,
+      "learning_rate": 2.374686439887436e-07,
+      "loss": 0.0149,
+      "step": 2604
+    },
+    {
+      "epoch": 0.9048280653004516,
+      "grad_norm": 0.5398376002785664,
+      "learning_rate": 2.3575841686946155e-07,
+      "loss": 0.0164,
+      "step": 2605
+    },
+    {
+      "epoch": 0.9051754081278222,
+      "grad_norm": 0.6221925809468902,
+      "learning_rate": 2.3405422176843329e-07,
+      "loss": 0.0256,
+      "step": 2606
+    },
+    {
+      "epoch": 0.9055227509551927,
+      "grad_norm": 0.5713859427829016,
+      "learning_rate": 2.3235606084334285e-07,
+      "loss": 0.0226,
+      "step": 2607
+    },
+    {
+      "epoch": 0.9058700937825633,
+      "grad_norm": 1.3040906101475984,
+      "learning_rate": 2.3066393624423754e-07,
+      "loss": 0.0166,
+      "step": 2608
+    },
+    {
+      "epoch": 0.906217436609934,
+      "grad_norm": 0.7255451058697687,
+      "learning_rate": 2.2897785011351982e-07,
+      "loss": 0.0209,
+      "step": 2609
+    },
+    {
+      "epoch": 0.9065647794373046,
+      "grad_norm": 0.6422064768024528,
+      "learning_rate": 2.2729780458594751e-07,
+      "loss": 0.0109,
+      "step": 2610
+    },
+    {
+      "epoch": 0.9069121222646752,
+      "grad_norm": 0.3372897791705386,
+      "learning_rate": 2.256238017886314e-07,
+      "loss": 0.0187,
+      "step": 2611
+    },
+    {
+      "epoch": 0.9072594650920458,
+      "grad_norm": 0.6097453676252612,
+      "learning_rate": 2.2395584384102943e-07,
+      "loss": 0.0207,
+      "step": 2612
+    },
+    {
+      "epoch": 0.9076068079194165,
+      "grad_norm": 0.6884196263626761,
+      "learning_rate": 2.2229393285494893e-07,
+      "loss": 0.0144,
+      "step": 2613
+    },
+    {
+      "epoch": 0.9079541507467871,
+      "grad_norm": 1.0714957544176478,
+      "learning_rate": 2.2063807093453736e-07,
+      "loss": 0.0182,
+      "step": 2614
+    },
+    {
+      "epoch": 0.9083014935741577,
+      "grad_norm": 0.8267148959880947,
+      "learning_rate": 2.1898826017628772e-07,
+      "loss": 0.0239,
+      "step": 2615
+    },
+    {
+      "epoch": 0.9086488364015283,
+      "grad_norm": 0.7458685484740719,
+      "learning_rate": 2.173445026690285e-07,
+      "loss": 0.0211,
+      "step": 2616
+    },
+    {
+      "epoch": 0.908996179228899,
+      "grad_norm": 0.5686980841248143,
+      "learning_rate": 2.1570680049392556e-07,
+      "loss": 0.0125,
+      "step": 2617
+    },
+    {
+      "epoch": 0.9093435220562696,
+      "grad_norm": 0.6449900442523265,
+      "learning_rate": 2.1407515572447747e-07,
+      "loss": 0.0241,
+      "step": 2618
+    },
+    {
+      "epoch": 0.9096908648836401,
+      "grad_norm": 0.4341128255677446,
+      "learning_rate": 2.1244957042651394e-07,
+      "loss": 0.0169,
+      "step": 2619
+    },
+    {
+      "epoch": 0.9100382077110107,
+      "grad_norm": 0.6600049909611397,
+      "learning_rate": 2.108300466581925e-07,
+      "loss": 0.0152,
+      "step": 2620
+    },
+    {
+      "epoch": 0.9103855505383813,
+      "grad_norm": 0.6245210505743247,
+      "learning_rate": 2.0921658646999687e-07,
+      "loss": 0.025,
+      "step": 2621
+    },
+    {
+      "epoch": 0.910732893365752,
+      "grad_norm": 0.5922289848107282,
+      "learning_rate": 2.0760919190473182e-07,
+      "loss": 0.0109,
+      "step": 2622
+    },
+    {
+      "epoch": 0.9110802361931226,
+      "grad_norm": 0.39775195287639503,
+      "learning_rate": 2.06007864997525e-07,
+      "loss": 0.015,
+      "step": 2623
+    },
+    {
+      "epoch": 0.9114275790204932,
+      "grad_norm": 0.7009054983235024,
+      "learning_rate": 2.0441260777582018e-07,
+      "loss": 0.0231,
+      "step": 2624
+    },
+    {
+      "epoch": 0.9117749218478638,
+      "grad_norm": 0.6973007401473146,
+      "learning_rate": 2.0282342225937503e-07,
+      "loss": 0.021,
+      "step": 2625
+    },
+    {
+      "epoch": 0.9121222646752345,
+      "grad_norm": 0.5883004994975823,
+      "learning_rate": 2.0124031046026283e-07,
+      "loss": 0.0159,
+      "step": 2626
+    },
+    {
+      "epoch": 0.9124696075026051,
+      "grad_norm": 0.454183134142274,
+      "learning_rate": 1.9966327438286582e-07,
+      "loss": 0.0169,
+      "step": 2627
+    },
+    {
+      "epoch": 0.9128169503299757,
+      "grad_norm": 0.384092493102016,
+      "learning_rate": 1.9809231602387236e-07,
+      "loss": 0.0132,
+      "step": 2628
+    },
+    {
+      "epoch": 0.9131642931573463,
+      "grad_norm": 0.4491174239681359,
+      "learning_rate": 1.9652743737227643e-07,
+      "loss": 0.0199,
+      "step": 2629
+    },
+    {
+      "epoch": 0.913511635984717,
+      "grad_norm": 0.59277378736531,
+      "learning_rate": 1.9496864040937536e-07,
+      "loss": 0.0196,
+      "step": 2630
+    },
+    {
+      "epoch": 0.9138589788120876,
+      "grad_norm": 0.7391457567820585,
+      "learning_rate": 1.9341592710876656e-07,
+      "loss": 0.0164,
+      "step": 2631
+    },
+    {
+      "epoch": 0.9142063216394581,
+      "grad_norm": 0.667688347659246,
+      "learning_rate": 1.9186929943634358e-07,
+      "loss": 0.026,
+      "step": 2632
+    },
+    {
+      "epoch": 0.9145536644668287,
+      "grad_norm": 0.4364968802938354,
+      "learning_rate": 1.9032875935029504e-07,
+      "loss": 0.0128,
+      "step": 2633
+    },
+    {
+      "epoch": 0.9149010072941993,
+      "grad_norm": 0.6408274013619847,
+      "learning_rate": 1.887943088011035e-07,
+      "loss": 0.0157,
+      "step": 2634
+    },
+    {
+      "epoch": 0.91524835012157,
+      "grad_norm": 0.5188280320693144,
+      "learning_rate": 1.87265949731541e-07,
+      "loss": 0.0108,
+      "step": 2635
+    },
+    {
+      "epoch": 0.9155956929489406,
+      "grad_norm": 0.8443286605185657,
+      "learning_rate": 1.8574368407666576e-07,
+      "loss": 0.0234,
+      "step": 2636
+    },
+    {
+      "epoch": 0.9159430357763112,
+      "grad_norm": 0.4426808473145947,
+      "learning_rate": 1.8422751376382274e-07,
+      "loss": 0.0067,
+      "step": 2637
+    },
+    {
+      "epoch": 0.9162903786036818,
+      "grad_norm": 0.47113894581099847,
+      "learning_rate": 1.8271744071263808e-07,
+      "loss": 0.0206,
+      "step": 2638
+    },
+    {
+      "epoch": 0.9166377214310525,
+      "grad_norm": 0.8457616638367563,
+      "learning_rate": 1.8121346683502183e-07,
+      "loss": 0.021,
+      "step": 2639
+    },
+    {
+      "epoch": 0.9169850642584231,
+      "grad_norm": 0.6588573901077968,
+      "learning_rate": 1.7971559403515526e-07,
+      "loss": 0.0191,
+      "step": 2640
+    },
+    {
+      "epoch": 0.9173324070857937,
+      "grad_norm": 0.45622450404307274,
+      "learning_rate": 1.782238242095008e-07,
+      "loss": 0.014,
+      "step": 2641
+    },
+    {
+      "epoch": 0.9176797499131643,
+      "grad_norm": 0.4489208891532178,
+      "learning_rate": 1.767381592467926e-07,
+      "loss": 0.0163,
+      "step": 2642
+    },
+    {
+      "epoch": 0.918027092740535,
+      "grad_norm": 0.52794845468743,
+      "learning_rate": 1.7525860102803438e-07,
+      "loss": 0.0214,
+      "step": 2643
+    },
+    {
+      "epoch": 0.9183744355679055,
+      "grad_norm": 0.33511808267174975,
+      "learning_rate": 1.7378515142649767e-07,
+      "loss": 0.0113,
+      "step": 2644
+    },
+    {
+      "epoch": 0.9187217783952761,
+      "grad_norm": 0.8011767384068549,
+      "learning_rate": 1.7231781230772127e-07,
+      "loss": 0.0294,
+      "step": 2645
+    },
+    {
+      "epoch": 0.9190691212226467,
+      "grad_norm": 0.3396441784353466,
+      "learning_rate": 1.7085658552950746e-07,
+      "loss": 0.0134,
+      "step": 2646
+    },
+    {
+      "epoch": 0.9194164640500173,
+      "grad_norm": 0.30267544503409916,
+      "learning_rate": 1.69401472941918e-07,
+      "loss": 0.0108,
+      "step": 2647
+    },
+    {
+      "epoch": 0.919763806877388,
+      "grad_norm": 0.7104785989934016,
+      "learning_rate": 1.6795247638727585e-07,
+      "loss": 0.0232,
+      "step": 2648
+    },
+    {
+      "epoch": 0.9201111497047586,
+      "grad_norm": 0.27352163599926305,
+      "learning_rate": 1.6650959770015796e-07,
+      "loss": 0.0135,
+      "step": 2649
+    },
+    {
+      "epoch": 0.9204584925321292,
+      "grad_norm": 0.494764482965323,
+      "learning_rate": 1.6507283870739798e-07,
+      "loss": 0.0164,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9208058353594998,
+      "grad_norm": 0.6063381987250709,
+      "learning_rate": 1.6364220122807862e-07,
+      "loss": 0.014,
+      "step": 2651
+    },
+    {
+      "epoch": 0.9211531781868705,
+      "grad_norm": 0.3484396467979152,
+      "learning_rate": 1.6221768707353536e-07,
+      "loss": 0.0146,
+      "step": 2652
+    },
+    {
+      "epoch": 0.9215005210142411,
+      "grad_norm": 0.5811657495793892,
+      "learning_rate": 1.6079929804734716e-07,
+      "loss": 0.0186,
+      "step": 2653
+    },
+    {
+      "epoch": 0.9218478638416117,
+      "grad_norm": 0.38993789351226965,
+      "learning_rate": 1.59387035945342e-07,
+      "loss": 0.0104,
+      "step": 2654
+    },
+    {
+      "epoch": 0.9221952066689822,
+      "grad_norm": 0.2970396986761084,
+      "learning_rate": 1.5798090255558617e-07,
+      "loss": 0.0094,
+      "step": 2655
+    },
+    {
+      "epoch": 0.922542549496353,
+      "grad_norm": 0.7453340514328095,
+      "learning_rate": 1.565808996583912e-07,
+      "loss": 0.0236,
+      "step": 2656
+    },
+    {
+      "epoch": 0.9228898923237235,
+      "grad_norm": 0.6111888530371641,
+      "learning_rate": 1.55187029026303e-07,
+      "loss": 0.0112,
+      "step": 2657
+    },
+    {
+      "epoch": 0.9232372351510941,
+      "grad_norm": 0.8674798788983026,
+      "learning_rate": 1.5379929242410385e-07,
+      "loss": 0.0263,
+      "step": 2658
+    },
+    {
+      "epoch": 0.9235845779784647,
+      "grad_norm": 0.527019561763946,
+      "learning_rate": 1.5241769160881104e-07,
+      "loss": 0.0142,
+      "step": 2659
+    },
+    {
+      "epoch": 0.9239319208058353,
+      "grad_norm": 0.7830231346293527,
+      "learning_rate": 1.5104222832967419e-07,
+      "loss": 0.0209,
+      "step": 2660
+    },
+    {
+      "epoch": 0.924279263633206,
+      "grad_norm": 0.5153735628692967,
+      "learning_rate": 1.4967290432817028e-07,
+      "loss": 0.0184,
+      "step": 2661
+    },
+    {
+      "epoch": 0.9246266064605766,
+      "grad_norm": 0.2844914040623888,
+      "learning_rate": 1.4830972133800247e-07,
+      "loss": 0.0093,
+      "step": 2662
+    },
+    {
+      "epoch": 0.9249739492879472,
+      "grad_norm": 0.43842457587592126,
+      "learning_rate": 1.4695268108510075e-07,
+      "loss": 0.0146,
+      "step": 2663
+    },
+    {
+      "epoch": 0.9253212921153178,
+      "grad_norm": 0.7398002758095784,
+      "learning_rate": 1.4560178528761848e-07,
+      "loss": 0.0177,
+      "step": 2664
+    },
+    {
+      "epoch": 0.9256686349426885,
+      "grad_norm": 0.35972295634537466,
+      "learning_rate": 1.4425703565592753e-07,
+      "loss": 0.0108,
+      "step": 2665
+    },
+    {
+      "epoch": 0.9260159777700591,
+      "grad_norm": 0.7120797307812906,
+      "learning_rate": 1.429184338926176e-07,
+      "loss": 0.0157,
+      "step": 2666
+    },
+    {
+      "epoch": 0.9263633205974297,
+      "grad_norm": 0.45611345610981685,
+      "learning_rate": 1.415859816924975e-07,
+      "loss": 0.0236,
+      "step": 2667
+    },
+    {
+      "epoch": 0.9267106634248002,
+      "grad_norm": 0.6910514224473907,
+      "learning_rate": 1.402596807425871e-07,
+      "loss": 0.0259,
+      "step": 2668
+    },
+    {
+      "epoch": 0.9270580062521709,
+      "grad_norm": 0.3565369326923046,
+      "learning_rate": 1.3893953272212046e-07,
+      "loss": 0.0147,
+      "step": 2669
+    },
+    {
+      "epoch": 0.9274053490795415,
+      "grad_norm": 0.4838529238212452,
+      "learning_rate": 1.3762553930253885e-07,
+      "loss": 0.0196,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9277526919069121,
+      "grad_norm": 0.32075955939895817,
+      "learning_rate": 1.3631770214749374e-07,
+      "loss": 0.016,
+      "step": 2671
+    },
+    {
+      "epoch": 0.9281000347342827,
+      "grad_norm": 0.5031442718892309,
+      "learning_rate": 1.3501602291284166e-07,
+      "loss": 0.0201,
+      "step": 2672
+    },
+    {
+      "epoch": 0.9284473775616533,
+      "grad_norm": 0.3459439042987737,
+      "learning_rate": 1.3372050324663988e-07,
+      "loss": 0.0214,
+      "step": 2673
+    },
+    {
+      "epoch": 0.928794720389024,
+      "grad_norm": 0.24515986422727065,
+      "learning_rate": 1.3243114478915076e-07,
+      "loss": 0.0098,
+      "step": 2674
+    },
+    {
+      "epoch": 0.9291420632163946,
+      "grad_norm": 0.5721382453019332,
+      "learning_rate": 1.3114794917283403e-07,
+      "loss": 0.0195,
+      "step": 2675
+    },
+    {
+      "epoch": 0.9294894060437652,
+      "grad_norm": 1.008562948955966,
+      "learning_rate": 1.2987091802234675e-07,
+      "loss": 0.0146,
+      "step": 2676
+    },
+    {
+      "epoch": 0.9298367488711358,
+      "grad_norm": 0.3165525870137379,
+      "learning_rate": 1.286000529545406e-07,
+      "loss": 0.0125,
+      "step": 2677
+    },
+    {
+      "epoch": 0.9301840916985065,
+      "grad_norm": 0.3519330637522624,
+      "learning_rate": 1.2733535557846176e-07,
+      "loss": 0.0132,
+      "step": 2678
+    },
+    {
+      "epoch": 0.930531434525877,
+      "grad_norm": 0.609989019077514,
+      "learning_rate": 1.2607682749534723e-07,
+      "loss": 0.0227,
+      "step": 2679
+    },
+    {
+      "epoch": 0.9308787773532476,
+      "grad_norm": 0.9719297935968579,
+      "learning_rate": 1.2482447029862177e-07,
+      "loss": 0.0175,
+      "step": 2680
+    },
+    {
+      "epoch": 0.9312261201806182,
+      "grad_norm": 0.5004324439145527,
+      "learning_rate": 1.2357828557389762e-07,
+      "loss": 0.0198,
+      "step": 2681
+    },
+    {
+      "epoch": 0.9315734630079889,
+      "grad_norm": 0.6042127292832795,
+      "learning_rate": 1.223382748989732e-07,
+      "loss": 0.0156,
+      "step": 2682
+    },
+    {
+      "epoch": 0.9319208058353595,
+      "grad_norm": 0.6442994121709944,
+      "learning_rate": 1.2110443984382936e-07,
+      "loss": 0.0178,
+      "step": 2683
+    },
+    {
+      "epoch": 0.9322681486627301,
+      "grad_norm": 0.4373568494058291,
+      "learning_rate": 1.1987678197062758e-07,
+      "loss": 0.0116,
+      "step": 2684
+    },
+    {
+      "epoch": 0.9326154914901007,
+      "grad_norm": 0.4862790006331638,
+      "learning_rate": 1.186553028337073e-07,
+      "loss": 0.0133,
+      "step": 2685
+    },
+    {
+      "epoch": 0.9329628343174713,
+      "grad_norm": 0.4434022690215712,
+      "learning_rate": 1.1744000397958755e-07,
+      "loss": 0.0176,
+      "step": 2686
+    },
+    {
+      "epoch": 0.933310177144842,
+      "grad_norm": 0.8580640939743742,
+      "learning_rate": 1.1623088694696194e-07,
+      "loss": 0.0198,
+      "step": 2687
+    },
+    {
+      "epoch": 0.9336575199722126,
+      "grad_norm": 0.4912841363664674,
+      "learning_rate": 1.150279532666948e-07,
+      "loss": 0.0137,
+      "step": 2688
+    },
+    {
+      "epoch": 0.9340048627995832,
+      "grad_norm": 0.6768885959717444,
+      "learning_rate": 1.1383120446182505e-07,
+      "loss": 0.023,
+      "step": 2689
+    },
+    {
+      "epoch": 0.9343522056269538,
+      "grad_norm": 0.4122917358139824,
+      "learning_rate": 1.1264064204755898e-07,
+      "loss": 0.0146,
+      "step": 2690
+    },
+    {
+      "epoch": 0.9346995484543245,
+      "grad_norm": 0.2906332176742442,
+      "learning_rate": 1.114562675312697e-07,
+      "loss": 0.0091,
+      "step": 2691
+    },
+    {
+      "epoch": 0.935046891281695,
+      "grad_norm": 0.788414928922905,
+      "learning_rate": 1.1027808241249715e-07,
+      "loss": 0.0234,
+      "step": 2692
+    },
+    {
+      "epoch": 0.9353942341090656,
+      "grad_norm": 0.7000778534595465,
+      "learning_rate": 1.0910608818294588e-07,
+      "loss": 0.0171,
+      "step": 2693
+    },
+    {
+      "epoch": 0.9357415769364362,
+      "grad_norm": 0.3365753374263883,
+      "learning_rate": 1.079402863264789e-07,
+      "loss": 0.011,
+      "step": 2694
+    },
+    {
+      "epoch": 0.9360889197638069,
+      "grad_norm": 0.500493153669285,
+      "learning_rate": 1.0678067831912164e-07,
+      "loss": 0.0246,
+      "step": 2695
+    },
+    {
+      "epoch": 0.9364362625911775,
+      "grad_norm": 0.49634500860470243,
+      "learning_rate": 1.056272656290569e-07,
+      "loss": 0.0151,
+      "step": 2696
+    },
+    {
+      "epoch": 0.9367836054185481,
+      "grad_norm": 0.2959869869958872,
+      "learning_rate": 1.0448004971662317e-07,
+      "loss": 0.0144,
+      "step": 2697
+    },
+    {
+      "epoch": 0.9371309482459187,
+      "grad_norm": 0.5721150903591803,
+      "learning_rate": 1.0333903203431362e-07,
+      "loss": 0.0158,
+      "step": 2698
+    },
+    {
+      "epoch": 0.9374782910732893,
+      "grad_norm": 0.5174080846934402,
+      "learning_rate": 1.0220421402677261e-07,
+      "loss": 0.0172,
+      "step": 2699
+    },
+    {
+      "epoch": 0.93782563390066,
+      "grad_norm": 0.6111422252957059,
+      "learning_rate": 1.0107559713079751e-07,
+      "loss": 0.0159,
+      "step": 2700
+    },
+    {
+      "epoch": 0.9381729767280306,
+      "grad_norm": 0.3308102163778448,
+      "learning_rate": 9.99531827753325e-08,
+      "loss": 0.0116,
+      "step": 2701
+    },
+    {
+      "epoch": 0.9385203195554012,
+      "grad_norm": 0.4117220886730794,
+      "learning_rate": 9.883697238146917e-08,
+      "loss": 0.0159,
+      "step": 2702
+    },
+    {
+      "epoch": 0.9388676623827718,
+      "grad_norm": 0.6253135827117662,
+      "learning_rate": 9.772696736244369e-08,
+      "loss": 0.0161,
+      "step": 2703
+    },
+    {
+      "epoch": 0.9392150052101425,
+      "grad_norm": 0.7918100091195154,
+      "learning_rate": 9.662316912363634e-08,
+      "loss": 0.0202,
+      "step": 2704
+    },
+    {
+      "epoch": 0.939562348037513,
+      "grad_norm": 0.8027328036502868,
+      "learning_rate": 9.552557906257032e-08,
+      "loss": 0.018,
+      "step": 2705
+    },
+    {
+      "epoch": 0.9399096908648836,
+      "grad_norm": 0.399436967336629,
+      "learning_rate": 9.443419856890568e-08,
+      "loss": 0.0154,
+      "step": 2706
+    },
+    {
+      "epoch": 0.9402570336922542,
+      "grad_norm": 0.5291507460704015,
+      "learning_rate": 9.33490290244421e-08,
+      "loss": 0.0147,
+      "step": 2707
+    },
+    {
+      "epoch": 0.9406043765196249,
+      "grad_norm": 0.9303566715405308,
+      "learning_rate": 9.227007180311609e-08,
+      "loss": 0.013,
+      "step": 2708
+    },
+    {
+      "epoch": 0.9409517193469955,
+      "grad_norm": 0.48135730059288107,
+      "learning_rate": 9.119732827099826e-08,
+      "loss": 0.0178,
+      "step": 2709
+    },
+    {
+      "epoch": 0.9412990621743661,
+      "grad_norm": 0.289263017584217,
+      "learning_rate": 9.013079978629047e-08,
+      "loss": 0.0139,
+      "step": 2710
+    },
+    {
+      "epoch": 0.9416464050017367,
+      "grad_norm": 1.2281515688429252,
+      "learning_rate": 8.907048769932813e-08,
+      "loss": 0.0232,
+      "step": 2711
+    },
+    {
+      "epoch": 0.9419937478291073,
+      "grad_norm": 0.45867978328141407,
+      "learning_rate": 8.801639335257573e-08,
+      "loss": 0.018,
+      "step": 2712
+    },
+    {
+      "epoch": 0.942341090656478,
+      "grad_norm": 0.8116884665077297,
+      "learning_rate": 8.696851808062401e-08,
+      "loss": 0.0166,
+      "step": 2713
+    },
+    {
+      "epoch": 0.9426884334838486,
+      "grad_norm": 0.629444316419227,
+      "learning_rate": 8.592686321019005e-08,
+      "loss": 0.0187,
+      "step": 2714
+    },
+    {
+      "epoch": 0.9430357763112192,
+      "grad_norm": 0.467448231376131,
+      "learning_rate": 8.489143006011613e-08,
+      "loss": 0.0191,
+      "step": 2715
+    },
+    {
+      "epoch": 0.9433831191385897,
+      "grad_norm": 1.540746884539893,
+      "learning_rate": 8.38622199413669e-08,
+      "loss": 0.0276,
+      "step": 2716
+    },
+    {
+      "epoch": 0.9437304619659604,
+      "grad_norm": 0.9960497105684367,
+      "learning_rate": 8.283923415702832e-08,
+      "loss": 0.0185,
+      "step": 2717
+    },
+    {
+      "epoch": 0.944077804793331,
+      "grad_norm": 0.6697810432465229,
+      "learning_rate": 8.182247400230381e-08,
+      "loss": 0.0239,
+      "step": 2718
+    },
+    {
+      "epoch": 0.9444251476207016,
+      "grad_norm": 0.5928459085259657,
+      "learning_rate": 8.081194076451749e-08,
+      "loss": 0.0155,
+      "step": 2719
+    },
+    {
+      "epoch": 0.9447724904480722,
+      "grad_norm": 0.9024042874062206,
+      "learning_rate": 7.980763572310702e-08,
+      "loss": 0.0175,
+      "step": 2720
+    },
+    {
+      "epoch": 0.9451198332754429,
+      "grad_norm": 0.9579838145780024,
+      "learning_rate": 7.880956014962694e-08,
+      "loss": 0.0271,
+      "step": 2721
+    },
+    {
+      "epoch": 0.9454671761028135,
+      "grad_norm": 0.49204147819791966,
+      "learning_rate": 7.781771530774085e-08,
+      "loss": 0.0233,
+      "step": 2722
+    },
+    {
+      "epoch": 0.9458145189301841,
+      "grad_norm": 0.6787970584155775,
+      "learning_rate": 7.683210245322869e-08,
+      "loss": 0.025,
+      "step": 2723
+    },
+    {
+      "epoch": 0.9461618617575547,
+      "grad_norm": 0.32439298949722206,
+      "learning_rate": 7.585272283397504e-08,
+      "loss": 0.0123,
+      "step": 2724
+    },
+    {
+      "epoch": 0.9465092045849253,
+      "grad_norm": 0.8141650539318078,
+      "learning_rate": 7.487957768997633e-08,
+      "loss": 0.0142,
+      "step": 2725
+    },
+    {
+      "epoch": 0.946856547412296,
+      "grad_norm": 0.8468827577767781,
+      "learning_rate": 7.391266825333365e-08,
+      "loss": 0.0111,
+      "step": 2726
+    },
+    {
+      "epoch": 0.9472038902396666,
+      "grad_norm": 0.5687781476564356,
+      "learning_rate": 7.295199574825384e-08,
+      "loss": 0.0198,
+      "step": 2727
+    },
+    {
+      "epoch": 0.9475512330670371,
+      "grad_norm": 0.4074296259542572,
+      "learning_rate": 7.199756139104563e-08,
+      "loss": 0.0144,
+      "step": 2728
+    },
+    {
+      "epoch": 0.9478985758944077,
+      "grad_norm": 0.7996569231526766,
+      "learning_rate": 7.104936639012239e-08,
+      "loss": 0.0304,
+      "step": 2729
+    },
+    {
+      "epoch": 0.9482459187217784,
+      "grad_norm": 0.45337039410172214,
+      "learning_rate": 7.01074119459949e-08,
+      "loss": 0.0146,
+      "step": 2730
+    },
+    {
+      "epoch": 0.948593261549149,
+      "grad_norm": 0.7844203973946874,
+      "learning_rate": 6.917169925127476e-08,
+      "loss": 0.0272,
+      "step": 2731
+    },
+    {
+      "epoch": 0.9489406043765196,
+      "grad_norm": 0.4489625006480948,
+      "learning_rate": 6.824222949066983e-08,
+      "loss": 0.0195,
+      "step": 2732
+    },
+    {
+      "epoch": 0.9492879472038902,
+      "grad_norm": 1.1267855989046178,
+      "learning_rate": 6.731900384098433e-08,
+      "loss": 0.0137,
+      "step": 2733
+    },
+    {
+      "epoch": 0.9496352900312609,
+      "grad_norm": 1.0751845620346967,
+      "learning_rate": 6.640202347111657e-08,
+      "loss": 0.019,
+      "step": 2734
+    },
+    {
+      "epoch": 0.9499826328586315,
+      "grad_norm": 0.5627865416818346,
+      "learning_rate": 6.54912895420573e-08,
+      "loss": 0.0185,
+      "step": 2735
+    },
+    {
+      "epoch": 0.9503299756860021,
+      "grad_norm": 0.49730875779627454,
+      "learning_rate": 6.458680320688914e-08,
+      "loss": 0.0198,
+      "step": 2736
+    },
+    {
+      "epoch": 0.9506773185133727,
+      "grad_norm": 0.5114294099131146,
+      "learning_rate": 6.368856561078496e-08,
+      "loss": 0.0115,
+      "step": 2737
+    },
+    {
+      "epoch": 0.9510246613407433,
+      "grad_norm": 0.8871643487076939,
+      "learning_rate": 6.279657789100612e-08,
+      "loss": 0.016,
+      "step": 2738
+    },
+    {
+      "epoch": 0.951372004168114,
+      "grad_norm": 0.6824181407432921,
+      "learning_rate": 6.191084117689871e-08,
+      "loss": 0.0225,
+      "step": 2739
+    },
+    {
+      "epoch": 0.9517193469954845,
+      "grad_norm": 0.3788527139900645,
+      "learning_rate": 6.103135658989789e-08,
+      "loss": 0.0135,
+      "step": 2740
+    },
+    {
+      "epoch": 0.9520666898228551,
+      "grad_norm": 2.27805723002845,
+      "learning_rate": 6.015812524352072e-08,
+      "loss": 0.0262,
+      "step": 2741
+    },
+    {
+      "epoch": 0.9524140326502257,
+      "grad_norm": 0.5250351000200569,
+      "learning_rate": 5.9291148243367235e-08,
+      "loss": 0.017,
+      "step": 2742
+    },
+    {
+      "epoch": 0.9527613754775964,
+      "grad_norm": 0.6854351694783098,
+      "learning_rate": 5.8430426687119954e-08,
+      "loss": 0.0209,
+      "step": 2743
+    },
+    {
+      "epoch": 0.953108718304967,
+      "grad_norm": 0.8186234378831619,
+      "learning_rate": 5.7575961664539384e-08,
+      "loss": 0.0262,
+      "step": 2744
+    },
+    {
+      "epoch": 0.9534560611323376,
+      "grad_norm": 0.5783055322077533,
+      "learning_rate": 5.672775425746735e-08,
+      "loss": 0.0164,
+      "step": 2745
+    },
+    {
+      "epoch": 0.9538034039597082,
+      "grad_norm": 1.1613038468708938,
+      "learning_rate": 5.588580553982092e-08,
+      "loss": 0.0306,
+      "step": 2746
+    },
+    {
+      "epoch": 0.9541507467870789,
+      "grad_norm": 0.8318319702699184,
+      "learning_rate": 5.505011657759296e-08,
+      "loss": 0.0129,
+      "step": 2747
+    },
+    {
+      "epoch": 0.9544980896144495,
+      "grad_norm": 0.7343407926498003,
+      "learning_rate": 5.4220688428850974e-08,
+      "loss": 0.0125,
+      "step": 2748
+    },
+    {
+      "epoch": 0.9548454324418201,
+      "grad_norm": 0.5822397422630301,
+      "learning_rate": 5.3397522143737725e-08,
+      "loss": 0.0187,
+      "step": 2749
+    },
+    {
+      "epoch": 0.9551927752691907,
+      "grad_norm": 0.36611892790873857,
+      "learning_rate": 5.258061876446507e-08,
+      "loss": 0.0129,
+      "step": 2750
+    },
+    {
+      "epoch": 0.9555401180965613,
+      "grad_norm": 0.506143736687749,
+      "learning_rate": 5.176997932531569e-08,
+      "loss": 0.0111,
+      "step": 2751
+    },
+    {
+      "epoch": 0.955887460923932,
+      "grad_norm": 0.37104961585297513,
+      "learning_rate": 5.096560485264301e-08,
+      "loss": 0.0178,
+      "step": 2752
+    },
+    {
+      "epoch": 0.9562348037513025,
+      "grad_norm": 0.39087699228743555,
+      "learning_rate": 5.016749636486851e-08,
+      "loss": 0.0219,
+      "step": 2753
+    },
+    {
+      "epoch": 0.9565821465786731,
+      "grad_norm": 0.3940695373758418,
+      "learning_rate": 4.937565487247775e-08,
+      "loss": 0.0217,
+      "step": 2754
+    },
+    {
+      "epoch": 0.9569294894060437,
+      "grad_norm": 0.8139685846877176,
+      "learning_rate": 4.859008137802379e-08,
+      "loss": 0.0188,
+      "step": 2755
+    },
+    {
+      "epoch": 0.9572768322334144,
+      "grad_norm": 0.3884984470065128,
+      "learning_rate": 4.781077687612379e-08,
+      "loss": 0.0118,
+      "step": 2756
+    },
+    {
+      "epoch": 0.957624175060785,
+      "grad_norm": 0.9041840556875058,
+      "learning_rate": 4.703774235345626e-08,
+      "loss": 0.0204,
+      "step": 2757
+    },
+    {
+      "epoch": 0.9579715178881556,
+      "grad_norm": 0.46537968867455554,
+      "learning_rate": 4.627097878876274e-08,
+      "loss": 0.0156,
+      "step": 2758
+    },
+    {
+      "epoch": 0.9583188607155262,
+      "grad_norm": 0.41094808662250093,
+      "learning_rate": 4.551048715284445e-08,
+      "loss": 0.0174,
+      "step": 2759
+    },
+    {
+      "epoch": 0.9586662035428969,
+      "grad_norm": 0.3654417168577741,
+      "learning_rate": 4.4756268408561174e-08,
+      "loss": 0.0193,
+      "step": 2760
+    },
+    {
+      "epoch": 0.9590135463702675,
+      "grad_norm": 0.6891593746161022,
+      "learning_rate": 4.400832351083184e-08,
+      "loss": 0.0185,
+      "step": 2761
+    },
+    {
+      "epoch": 0.9593608891976381,
+      "grad_norm": 0.9417802736827884,
+      "learning_rate": 4.326665340663117e-08,
+      "loss": 0.0176,
+      "step": 2762
+    },
+    {
+      "epoch": 0.9597082320250087,
+      "grad_norm": 0.6092562454977165,
+      "learning_rate": 4.253125903498967e-08,
+      "loss": 0.0113,
+      "step": 2763
+    },
+    {
+      "epoch": 0.9600555748523792,
+      "grad_norm": 0.4780402820280536,
+      "learning_rate": 4.180214132699201e-08,
+      "loss": 0.0132,
+      "step": 2764
+    },
+    {
+      "epoch": 0.96040291767975,
+      "grad_norm": 0.7203305496862711,
+      "learning_rate": 4.10793012057753e-08,
+      "loss": 0.0201,
+      "step": 2765
+    },
+    {
+      "epoch": 0.9607502605071205,
+      "grad_norm": 0.78632337470775,
+      "learning_rate": 4.03627395865297e-08,
+      "loss": 0.0161,
+      "step": 2766
+    },
+    {
+      "epoch": 0.9610976033344911,
+      "grad_norm": 0.7800907681147895,
+      "learning_rate": 3.9652457376496146e-08,
+      "loss": 0.0226,
+      "step": 2767
+    },
+    {
+      "epoch": 0.9614449461618617,
+      "grad_norm": 0.7579603624294482,
+      "learning_rate": 3.894845547496418e-08,
+      "loss": 0.018,
+      "step": 2768
+    },
+    {
+      "epoch": 0.9617922889892324,
+      "grad_norm": 0.4992736774720632,
+      "learning_rate": 3.8250734773272455e-08,
+      "loss": 0.0202,
+      "step": 2769
+    },
+    {
+      "epoch": 0.962139631816603,
+      "grad_norm": 0.5447806191032949,
+      "learning_rate": 3.755929615480658e-08,
+      "loss": 0.0158,
+      "step": 2770
+    },
+    {
+      "epoch": 0.9624869746439736,
+      "grad_norm": 0.4587220635778511,
+      "learning_rate": 3.687414049500015e-08,
+      "loss": 0.0174,
+      "step": 2771
+    },
+    {
+      "epoch": 0.9628343174713442,
+      "grad_norm": 0.6017122263910978,
+      "learning_rate": 3.619526866132872e-08,
+      "loss": 0.0212,
+      "step": 2772
+    },
+    {
+      "epoch": 0.9631816602987149,
+      "grad_norm": 0.45153146666560434,
+      "learning_rate": 3.552268151331417e-08,
+      "loss": 0.0129,
+      "step": 2773
+    },
+    {
+      "epoch": 0.9635290031260855,
+      "grad_norm": 0.8050815601154685,
+      "learning_rate": 3.485637990252089e-08,
+      "loss": 0.0203,
+      "step": 2774
+    },
+    {
+      "epoch": 0.9638763459534561,
+      "grad_norm": 0.5550194035824456,
+      "learning_rate": 3.4196364672555715e-08,
+      "loss": 0.0149,
+      "step": 2775
+    },
+    {
+      "epoch": 0.9642236887808266,
+      "grad_norm": 0.6061372955316445,
+      "learning_rate": 3.3542636659064095e-08,
+      "loss": 0.017,
+      "step": 2776
+    },
+    {
+      "epoch": 0.9645710316081972,
+      "grad_norm": 0.7681878631554084,
+      "learning_rate": 3.2895196689733953e-08,
+      "loss": 0.0143,
+      "step": 2777
+    },
+    {
+      "epoch": 0.9649183744355679,
+      "grad_norm": 0.7249193630109478,
+      "learning_rate": 3.225404558429068e-08,
+      "loss": 0.0175,
+      "step": 2778
+    },
+    {
+      "epoch": 0.9652657172629385,
+      "grad_norm": 0.31018295263385637,
+      "learning_rate": 3.1619184154496605e-08,
+      "loss": 0.0096,
+      "step": 2779
+    },
+    {
+      "epoch": 0.9656130600903091,
+      "grad_norm": 0.5323520395260942,
+      "learning_rate": 3.099061320415153e-08,
+      "loss": 0.02,
+      "step": 2780
+    },
+    {
+      "epoch": 0.9659604029176797,
+      "grad_norm": 0.5460021526097322,
+      "learning_rate": 3.036833352909052e-08,
+      "loss": 0.0138,
+      "step": 2781
+    },
+    {
+      "epoch": 0.9663077457450504,
+      "grad_norm": 0.4231306425496086,
+      "learning_rate": 2.9752345917184456e-08,
+      "loss": 0.0128,
+      "step": 2782
+    },
+    {
+      "epoch": 0.966655088572421,
+      "grad_norm": 0.5104133537482568,
+      "learning_rate": 2.914265114833614e-08,
+      "loss": 0.0181,
+      "step": 2783
+    },
+    {
+      "epoch": 0.9670024313997916,
+      "grad_norm": 0.32680316277058136,
+      "learning_rate": 2.8539249994480878e-08,
+      "loss": 0.0142,
+      "step": 2784
+    },
+    {
+      "epoch": 0.9673497742271622,
+      "grad_norm": 0.6787838217925692,
+      "learning_rate": 2.7942143219587547e-08,
+      "loss": 0.0227,
+      "step": 2785
+    },
+    {
+      "epoch": 0.9676971170545329,
+      "grad_norm": 0.5393149296688037,
+      "learning_rate": 2.7351331579654194e-08,
+      "loss": 0.0199,
+      "step": 2786
+    },
+    {
+      "epoch": 0.9680444598819035,
+      "grad_norm": 0.3709030328592483,
+      "learning_rate": 2.6766815822709124e-08,
+      "loss": 0.0189,
+      "step": 2787
+    },
+    {
+      "epoch": 0.968391802709274,
+      "grad_norm": 2.1813300319236433,
+      "learning_rate": 2.618859668880869e-08,
+      "loss": 0.0233,
+      "step": 2788
+    },
+    {
+      "epoch": 0.9687391455366446,
+      "grad_norm": 0.6115166749222286,
+      "learning_rate": 2.561667491003783e-08,
+      "loss": 0.0151,
+      "step": 2789
+    },
+    {
+      "epoch": 0.9690864883640152,
+      "grad_norm": 0.7565981079039917,
+      "learning_rate": 2.5051051210508437e-08,
+      "loss": 0.0133,
+      "step": 2790
+    },
+    {
+      "epoch": 0.9694338311913859,
+      "grad_norm": 0.6494118382951497,
+      "learning_rate": 2.4491726306357656e-08,
+      "loss": 0.0129,
+      "step": 2791
+    },
+    {
+      "epoch": 0.9697811740187565,
+      "grad_norm": 0.8914095771286661,
+      "learning_rate": 2.3938700905747902e-08,
+      "loss": 0.0165,
+      "step": 2792
+    },
+    {
+      "epoch": 0.9701285168461271,
+      "grad_norm": 0.3651793620369032,
+      "learning_rate": 2.3391975708866866e-08,
+      "loss": 0.014,
+      "step": 2793
+    },
+    {
+      "epoch": 0.9704758596734977,
+      "grad_norm": 0.6009630665505777,
+      "learning_rate": 2.2851551407924166e-08,
+      "loss": 0.0126,
+      "step": 2794
+    },
+    {
+      "epoch": 0.9708232025008684,
+      "grad_norm": 0.6517704643511893,
+      "learning_rate": 2.231742868715303e-08,
+      "loss": 0.0215,
+      "step": 2795
+    },
+    {
+      "epoch": 0.971170545328239,
+      "grad_norm": 0.5374743831920767,
+      "learning_rate": 2.17896082228064e-08,
+      "loss": 0.0109,
+      "step": 2796
+    },
+    {
+      "epoch": 0.9715178881556096,
+      "grad_norm": 0.43751454745868124,
+      "learning_rate": 2.1268090683159714e-08,
+      "loss": 0.0151,
+      "step": 2797
+    },
+    {
+      "epoch": 0.9718652309829802,
+      "grad_norm": 0.9225025519084343,
+      "learning_rate": 2.075287672850812e-08,
+      "loss": 0.0185,
+      "step": 2798
+    },
+    {
+      "epoch": 0.9722125738103509,
+      "grad_norm": 0.6499562483522786,
+      "learning_rate": 2.0243967011164267e-08,
+      "loss": 0.0143,
+      "step": 2799
+    },
+    {
+      "epoch": 0.9725599166377215,
+      "grad_norm": 0.4078620285370667,
+      "learning_rate": 1.9741362175461076e-08,
+      "loss": 0.0095,
+      "step": 2800
+    },
+    {
+      "epoch": 0.972907259465092,
+      "grad_norm": 0.649018054650212,
+      "learning_rate": 1.9245062857746744e-08,
+      "loss": 0.015,
+      "step": 2801
+    },
+    {
+      "epoch": 0.9732546022924626,
+      "grad_norm": 0.8446552989012637,
+      "learning_rate": 1.8755069686388074e-08,
+      "loss": 0.0265,
+      "step": 2802
+    },
+    {
+      "epoch": 0.9736019451198332,
+      "grad_norm": 0.8836022726540815,
+      "learning_rate": 1.827138328176603e-08,
+      "loss": 0.0197,
+      "step": 2803
+    },
+    {
+      "epoch": 0.9739492879472039,
+      "grad_norm": 0.39841982862426034,
+      "learning_rate": 1.7794004256277976e-08,
+      "loss": 0.0112,
+      "step": 2804
+    },
+    {
+      "epoch": 0.9742966307745745,
+      "grad_norm": 0.6590609628914503,
+      "learning_rate": 1.732293321433376e-08,
+      "loss": 0.022,
+      "step": 2805
+    },
+    {
+      "epoch": 0.9746439736019451,
+      "grad_norm": 0.5300192127082276,
+      "learning_rate": 1.6858170752357407e-08,
+      "loss": 0.0164,
+      "step": 2806
+    },
+    {
+      "epoch": 0.9749913164293157,
+      "grad_norm": 1.0204073151035766,
+      "learning_rate": 1.63997174587871e-08,
+      "loss": 0.0123,
+      "step": 2807
+    },
+    {
+      "epoch": 0.9753386592566864,
+      "grad_norm": 0.571323310061111,
+      "learning_rate": 1.594757391407076e-08,
+      "loss": 0.0157,
+      "step": 2808
+    },
+    {
+      "epoch": 0.975686002084057,
+      "grad_norm": 0.3778707394308839,
+      "learning_rate": 1.5501740690668788e-08,
+      "loss": 0.0235,
+      "step": 2809
+    },
+    {
+      "epoch": 0.9760333449114276,
+      "grad_norm": 0.44292675404107196,
+      "learning_rate": 1.506221835305133e-08,
+      "loss": 0.0173,
+      "step": 2810
+    },
+    {
+      "epoch": 0.9763806877387982,
+      "grad_norm": 0.2763995858299458,
+      "learning_rate": 1.4629007457699906e-08,
+      "loss": 0.0093,
+      "step": 2811
+    },
+    {
+      "epoch": 0.9767280305661689,
+      "grad_norm": 0.8458253024017103,
+      "learning_rate": 1.4202108553102445e-08,
+      "loss": 0.0216,
+      "step": 2812
+    },
+    {
+      "epoch": 0.9770753733935394,
+      "grad_norm": 0.47122826092790115,
+      "learning_rate": 1.3781522179757146e-08,
+      "loss": 0.0211,
+      "step": 2813
+    },
+    {
+      "epoch": 0.97742271622091,
+      "grad_norm": 0.6437597372265773,
+      "learning_rate": 1.3367248870170269e-08,
+      "loss": 0.0253,
+      "step": 2814
+    },
+    {
+      "epoch": 0.9777700590482806,
+      "grad_norm": 0.7953386444191836,
+      "learning_rate": 1.295928914885336e-08,
+      "loss": 0.0216,
+      "step": 2815
+    },
+    {
+      "epoch": 0.9781174018756512,
+      "grad_norm": 0.7510223662917004,
+      "learning_rate": 1.255764353232547e-08,
+      "loss": 0.0162,
+      "step": 2816
+    },
+    {
+      "epoch": 0.9784647447030219,
+      "grad_norm": 0.38965761008763483,
+      "learning_rate": 1.2162312529111487e-08,
+      "loss": 0.0089,
+      "step": 2817
+    },
+    {
+      "epoch": 0.9788120875303925,
+      "grad_norm": 0.8708571004518411,
+      "learning_rate": 1.1773296639741028e-08,
+      "loss": 0.0186,
+      "step": 2818
+    },
+    {
+      "epoch": 0.9791594303577631,
+      "grad_norm": 0.4280076925155896,
+      "learning_rate": 1.139059635674733e-08,
+      "loss": 0.0125,
+      "step": 2819
+    },
+    {
+      "epoch": 0.9795067731851337,
+      "grad_norm": 0.5982212585888964,
+      "learning_rate": 1.1014212164668914e-08,
+      "loss": 0.0128,
+      "step": 2820
+    },
+    {
+      "epoch": 0.9798541160125044,
+      "grad_norm": 0.4802236337534653,
+      "learning_rate": 1.0644144540046253e-08,
+      "loss": 0.0154,
+      "step": 2821
+    },
+    {
+      "epoch": 0.980201458839875,
+      "grad_norm": 0.6709260979923914,
+      "learning_rate": 1.028039395142344e-08,
+      "loss": 0.0162,
+      "step": 2822
+    },
+    {
+      "epoch": 0.9805488016672456,
+      "grad_norm": 0.3994630255273555,
+      "learning_rate": 9.92296085934541e-09,
+      "loss": 0.0136,
+      "step": 2823
+    },
+    {
+      "epoch": 0.9808961444946162,
+      "grad_norm": 0.8512917771148851,
+      "learning_rate": 9.571845716360162e-09,
+      "loss": 0.0137,
+      "step": 2824
+    },
+    {
+      "epoch": 0.9812434873219869,
+      "grad_norm": 0.7023478964706273,
+      "learning_rate": 9.227048967013762e-09,
+      "loss": 0.0199,
+      "step": 2825
+    },
+    {
+      "epoch": 0.9815908301493574,
+      "grad_norm": 0.5923172701448587,
+      "learning_rate": 8.888571047855899e-09,
+      "loss": 0.0149,
+      "step": 2826
+    },
+    {
+      "epoch": 0.981938172976728,
+      "grad_norm": 0.18336419674101587,
+      "learning_rate": 8.55641238743321e-09,
+      "loss": 0.0045,
+      "step": 2827
+    },
+    {
+      "epoch": 0.9822855158040986,
+      "grad_norm": 0.7801052776351165,
+      "learning_rate": 8.230573406293186e-09,
+      "loss": 0.0185,
+      "step": 2828
+    },
+    {
+      "epoch": 0.9826328586314692,
+      "grad_norm": 0.4666029152584524,
+      "learning_rate": 7.911054516981376e-09,
+      "loss": 0.0092,
+      "step": 2829
+    },
+    {
+      "epoch": 0.9829802014588399,
+      "grad_norm": 0.5508063511757076,
+      "learning_rate": 7.597856124040848e-09,
+      "loss": 0.0278,
+      "step": 2830
+    },
+    {
+      "epoch": 0.9833275442862105,
+      "grad_norm": 0.47041522453276524,
+      "learning_rate": 7.290978624013289e-09,
+      "loss": 0.0131,
+      "step": 2831
+    },
+    {
+      "epoch": 0.9836748871135811,
+      "grad_norm": 0.5117867904313976,
+      "learning_rate": 6.990422405437903e-09,
+      "loss": 0.0169,
+      "step": 2832
+    },
+    {
+      "epoch": 0.9840222299409517,
+      "grad_norm": 0.4928579747341889,
+      "learning_rate": 6.696187848848068e-09,
+      "loss": 0.014,
+      "step": 2833
+    },
+    {
+      "epoch": 0.9843695727683224,
+      "grad_norm": 0.3862793465920075,
+      "learning_rate": 6.408275326775792e-09,
+      "loss": 0.0194,
+      "step": 2834
+    },
+    {
+      "epoch": 0.984716915595693,
+      "grad_norm": 0.38744597996896796,
+      "learning_rate": 6.126685203747818e-09,
+      "loss": 0.0142,
+      "step": 2835
+    },
+    {
+      "epoch": 0.9850642584230636,
+      "grad_norm": 0.36149263951244837,
+      "learning_rate": 5.851417836286177e-09,
+      "loss": 0.0152,
+      "step": 2836
+    },
+    {
+      "epoch": 0.9854116012504341,
+      "grad_norm": 0.6059984628372787,
+      "learning_rate": 5.582473572907643e-09,
+      "loss": 0.0125,
+      "step": 2837
+    },
+    {
+      "epoch": 0.9857589440778048,
+      "grad_norm": 1.0805004360514219,
+      "learning_rate": 5.319852754122612e-09,
+      "loss": 0.0188,
+      "step": 2838
+    },
+    {
+      "epoch": 0.9861062869051754,
+      "grad_norm": 0.7645678869514635,
+      "learning_rate": 5.0635557124362185e-09,
+      "loss": 0.0189,
+      "step": 2839
+    },
+    {
+      "epoch": 0.986453629732546,
+      "grad_norm": 1.4270577859101476,
+      "learning_rate": 4.813582772347225e-09,
+      "loss": 0.027,
+      "step": 2840
+    },
+    {
+      "epoch": 0.9868009725599166,
+      "grad_norm": 0.6651264047671889,
+      "learning_rate": 4.569934250346908e-09,
+      "loss": 0.0171,
+      "step": 2841
+    },
+    {
+      "epoch": 0.9871483153872872,
+      "grad_norm": 0.562975977952286,
+      "learning_rate": 4.332610454919062e-09,
+      "loss": 0.0155,
+      "step": 2842
+    },
+    {
+      "epoch": 0.9874956582146579,
+      "grad_norm": 0.5118103669731423,
+      "learning_rate": 4.101611686539442e-09,
+      "loss": 0.013,
+      "step": 2843
+    },
+    {
+      "epoch": 0.9878430010420285,
+      "grad_norm": 0.7783130801930213,
+      "learning_rate": 3.876938237676875e-09,
+      "loss": 0.0253,
+      "step": 2844
+    },
+    {
+      "epoch": 0.9881903438693991,
+      "grad_norm": 0.34801718183552194,
+      "learning_rate": 3.6585903927910395e-09,
+      "loss": 0.0118,
+      "step": 2845
+    },
+    {
+      "epoch": 0.9885376866967697,
+      "grad_norm": 0.535293498292763,
+      "learning_rate": 3.446568428332464e-09,
+      "loss": 0.014,
+      "step": 2846
+    },
+    {
+      "epoch": 0.9888850295241404,
+      "grad_norm": 0.504081454212183,
+      "learning_rate": 3.2408726127425294e-09,
+      "loss": 0.0184,
+      "step": 2847
+    },
+    {
+      "epoch": 0.989232372351511,
+      "grad_norm": 0.5720403882389867,
+      "learning_rate": 3.0415032064534668e-09,
+      "loss": 0.0243,
+      "step": 2848
+    },
+    {
+      "epoch": 0.9895797151788815,
+      "grad_norm": 0.5342584944130937,
+      "learning_rate": 2.848460461887248e-09,
+      "loss": 0.0212,
+      "step": 2849
+    },
+    {
+      "epoch": 0.9899270580062521,
+      "grad_norm": 0.37088583044080237,
+      "learning_rate": 2.6617446234555866e-09,
+      "loss": 0.0199,
+      "step": 2850
+    },
+    {
+      "epoch": 0.9902744008336228,
+      "grad_norm": 0.36631835755907144,
+      "learning_rate": 2.4813559275604914e-09,
+      "loss": 0.0147,
+      "step": 2851
+    },
+    {
+      "epoch": 0.9906217436609934,
+      "grad_norm": 0.7717204298531588,
+      "learning_rate": 2.3072946025920474e-09,
+      "loss": 0.0133,
+      "step": 2852
+    },
+    {
+      "epoch": 0.990969086488364,
+      "grad_norm": 0.4286410759035607,
+      "learning_rate": 2.1395608689295244e-09,
+      "loss": 0.0172,
+      "step": 2853
+    },
+    {
+      "epoch": 0.9913164293157346,
+      "grad_norm": 0.7876162638170817,
+      "learning_rate": 1.978154938941379e-09,
+      "loss": 0.0192,
+      "step": 2854
+    },
+    {
+      "epoch": 0.9916637721431052,
+      "grad_norm": 0.9045977670509248,
+      "learning_rate": 1.8230770169841427e-09,
+      "loss": 0.0145,
+      "step": 2855
+    },
+    {
+      "epoch": 0.9920111149704759,
+      "grad_norm": 0.34387118239672126,
+      "learning_rate": 1.674327299402423e-09,
+      "loss": 0.0089,
+      "step": 2856
+    },
+    {
+      "epoch": 0.9923584577978465,
+      "grad_norm": 0.37959493138909545,
+      "learning_rate": 1.531905974528347e-09,
+      "loss": 0.0163,
+      "step": 2857
+    },
+    {
+      "epoch": 0.9927058006252171,
+      "grad_norm": 0.7818912632587496,
+      "learning_rate": 1.3958132226821187e-09,
+      "loss": 0.02,
+      "step": 2858
+    },
+    {
+      "epoch": 0.9930531434525877,
+      "grad_norm": 0.8616967656128703,
+      "learning_rate": 1.266049216170906e-09,
+      "loss": 0.0235,
+      "step": 2859
+    },
+    {
+      "epoch": 0.9934004862799584,
+      "grad_norm": 0.8090514774003552,
+      "learning_rate": 1.142614119289398e-09,
+      "loss": 0.0208,
+      "step": 2860
+    },
+    {
+      "epoch": 0.993747829107329,
+      "grad_norm": 0.40981264825919256,
+      "learning_rate": 1.025508088318694e-09,
+      "loss": 0.0119,
+      "step": 2861
+    },
+    {
+      "epoch": 0.9940951719346995,
+      "grad_norm": 0.9853414305979534,
+      "learning_rate": 9.147312715279688e-10,
+      "loss": 0.0194,
+      "step": 2862
+    },
+    {
+      "epoch": 0.9944425147620701,
+      "grad_norm": 0.6823065427028371,
+      "learning_rate": 8.102838091705867e-10,
+      "loss": 0.0171,
+      "step": 2863
+    },
+    {
+      "epoch": 0.9947898575894408,
+      "grad_norm": 0.49217144790501227,
+      "learning_rate": 7.121658334890979e-10,
+      "loss": 0.0127,
+      "step": 2864
+    },
+    {
+      "epoch": 0.9951372004168114,
+      "grad_norm": 0.7961270955956514,
+      "learning_rate": 6.203774687102426e-10,
+      "loss": 0.0213,
+      "step": 2865
+    },
+    {
+      "epoch": 0.995484543244182,
+      "grad_norm": 0.5901951990218518,
+      "learning_rate": 5.34918831047726e-10,
+      "loss": 0.0181,
+      "step": 2866
+    },
+    {
+      "epoch": 0.9958318860715526,
+      "grad_norm": 1.0523471488146157,
+      "learning_rate": 4.5579002870110854e-10,
+      "loss": 0.0231,
+      "step": 2867
+    },
+    {
+      "epoch": 0.9961792288989232,
+      "grad_norm": 0.5550763297657431,
+      "learning_rate": 3.8299116185525066e-10,
+      "loss": 0.011,
+      "step": 2868
+    },
+    {
+      "epoch": 0.9965265717262939,
+      "grad_norm": 0.5408286201029091,
+      "learning_rate": 3.165223226808678e-10,
+      "loss": 0.0161,
+      "step": 2869
+    },
+    {
+      "epoch": 0.9968739145536645,
+      "grad_norm": 0.4991725997802592,
+      "learning_rate": 2.563835953345306e-10,
+      "loss": 0.0175,
+      "step": 2870
+    },
+    {
+      "epoch": 0.9972212573810351,
+      "grad_norm": 0.5996623257508237,
+      "learning_rate": 2.0257505595810966e-10,
+      "loss": 0.0183,
+      "step": 2871
+    },
+    {
+      "epoch": 0.9975686002084057,
+      "grad_norm": 0.7177885884770104,
+      "learning_rate": 1.5509677267877555e-10,
+      "loss": 0.0194,
+      "step": 2872
+    },
+    {
+      "epoch": 0.9979159430357764,
+      "grad_norm": 0.639091556482265,
+      "learning_rate": 1.1394880560844368e-10,
+      "loss": 0.0201,
+      "step": 2873
+    },
+    {
+      "epoch": 0.9982632858631469,
+      "grad_norm": 0.5680423116235028,
+      "learning_rate": 7.913120684488462e-11,
+      "loss": 0.0189,
+      "step": 2874
+    },
+    {
+      "epoch": 0.9986106286905175,
+      "grad_norm": 0.23508445434740197,
+      "learning_rate": 5.0644020471168894e-11,
+      "loss": 0.0098,
+      "step": 2875
+    },
+    {
+      "epoch": 0.9989579715178881,
+      "grad_norm": 0.3673594266436054,
+      "learning_rate": 2.8487282554556796e-11,
+      "loss": 0.0109,
+      "step": 2876
+    },
+    {
+      "epoch": 0.9993053143452588,
+      "grad_norm": 0.4480479527832172,
+      "learning_rate": 1.2661021148163699e-11,
+      "loss": 0.0206,
+      "step": 2877
+    },
+    {
+      "epoch": 0.9996526571726294,
+      "grad_norm": 0.48054013159980413,
+      "learning_rate": 3.1652562887396486e-12,
+      "loss": 0.0155,
+      "step": 2878
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7081858993617053,
+      "learning_rate": 0.0,
+      "loss": 0.0175,
+      "step": 2879
+    },
+    {
+      "epoch": 1.0,
+      "step": 2879,
+      "total_flos": 49962393949184.0,
+      "train_loss": 0.021164150436296568,
+      "train_runtime": 16194.0514,
+      "train_samples_per_second": 11.381,
+      "train_steps_per_second": 0.178
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2879,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 49962393949184.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}