diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,37485 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.8003589053387169,
+  "eval_steps": 500,
+  "global_step": 5352,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00014954389113204725,
+      "grad_norm": 35.95169344376715,
+      "learning_rate": 4.975124378109453e-08,
+      "loss": 1.1911,
+      "step": 1
+    },
+    {
+      "epoch": 0.0002990877822640945,
+      "grad_norm": 29.047342527504238,
+      "learning_rate": 9.950248756218906e-08,
+      "loss": 1.4707,
+      "step": 2
+    },
+    {
+      "epoch": 0.00044863167339614175,
+      "grad_norm": 24.718727160032117,
+      "learning_rate": 1.4925373134328358e-07,
+      "loss": 0.9534,
+      "step": 3
+    },
+    {
+      "epoch": 0.000598175564528189,
+      "grad_norm": 32.87218994198639,
+      "learning_rate": 1.9900497512437812e-07,
+      "loss": 1.2192,
+      "step": 4
+    },
+    {
+      "epoch": 0.0007477194556602363,
+      "grad_norm": 25.398344980222138,
+      "learning_rate": 2.4875621890547267e-07,
+      "loss": 1.1835,
+      "step": 5
+    },
+    {
+      "epoch": 0.0008972633467922835,
+      "grad_norm": 30.48079389364258,
+      "learning_rate": 2.9850746268656716e-07,
+      "loss": 1.0024,
+      "step": 6
+    },
+    {
+      "epoch": 0.0010468072379243307,
+      "grad_norm": 27.780032565686206,
+      "learning_rate": 3.4825870646766175e-07,
+      "loss": 1.1796,
+      "step": 7
+    },
+    {
+      "epoch": 0.001196351129056378,
+      "grad_norm": 33.19634259772052,
+      "learning_rate": 3.9800995024875624e-07,
+      "loss": 0.9585,
+      "step": 8
+    },
+    {
+      "epoch": 0.0013458950201884253,
+      "grad_norm": 32.92097675417938,
+      "learning_rate": 4.4776119402985074e-07,
+      "loss": 1.1831,
+      "step": 9
+    },
+    {
+      "epoch": 0.0014954389113204726,
+      "grad_norm": 31.267461918177617,
+      "learning_rate": 4.975124378109453e-07,
+      "loss": 0.9208,
+      "step": 10
+    },
+    {
+      "epoch": 0.0016449828024525197,
+      "grad_norm": 31.652990928454088,
+      "learning_rate": 5.472636815920398e-07,
+      "loss": 0.8882,
+      "step": 11
+    },
+    {
+      "epoch": 0.001794526693584567,
+      "grad_norm": 33.800482625732165,
+      "learning_rate": 5.970149253731343e-07,
+      "loss": 1.2138,
+      "step": 12
+    },
+    {
+      "epoch": 0.0019440705847166143,
+      "grad_norm": 30.753216086819556,
+      "learning_rate": 6.467661691542289e-07,
+      "loss": 0.9896,
+      "step": 13
+    },
+    {
+      "epoch": 0.0020936144758486614,
+      "grad_norm": 32.57679525538582,
+      "learning_rate": 6.965174129353235e-07,
+      "loss": 0.9195,
+      "step": 14
+    },
+    {
+      "epoch": 0.0022431583669807087,
+      "grad_norm": 25.334089702892793,
+      "learning_rate": 7.462686567164179e-07,
+      "loss": 0.7515,
+      "step": 15
+    },
+    {
+      "epoch": 0.002392702258112756,
+      "grad_norm": 22.2961872211284,
+      "learning_rate": 7.960199004975125e-07,
+      "loss": 0.6638,
+      "step": 16
+    },
+    {
+      "epoch": 0.0025422461492448033,
+      "grad_norm": 24.245556768411276,
+      "learning_rate": 8.457711442786071e-07,
+      "loss": 0.7704,
+      "step": 17
+    },
+    {
+      "epoch": 0.0026917900403768506,
+      "grad_norm": 19.23412917202397,
+      "learning_rate": 8.955223880597015e-07,
+      "loss": 0.7354,
+      "step": 18
+    },
+    {
+      "epoch": 0.002841333931508898,
+      "grad_norm": 18.58051317424024,
+      "learning_rate": 9.452736318407961e-07,
+      "loss": 0.5749,
+      "step": 19
+    },
+    {
+      "epoch": 0.0029908778226409452,
+      "grad_norm": 11.242228896944281,
+      "learning_rate": 9.950248756218907e-07,
+      "loss": 0.4914,
+      "step": 20
+    },
+    {
+      "epoch": 0.0031404217137729925,
+      "grad_norm": 11.163527479225325,
+      "learning_rate": 1.044776119402985e-06,
+      "loss": 0.5823,
+      "step": 21
+    },
+    {
+      "epoch": 0.0032899656049050394,
+      "grad_norm": 9.100766388616314,
+      "learning_rate": 1.0945273631840796e-06,
+      "loss": 0.6887,
+      "step": 22
+    },
+    {
+      "epoch": 0.0034395094960370867,
+      "grad_norm": 9.371427313022828,
+      "learning_rate": 1.1442786069651742e-06,
+      "loss": 0.3365,
+      "step": 23
+    },
+    {
+      "epoch": 0.003589053387169134,
+      "grad_norm": 6.591365654298028,
+      "learning_rate": 1.1940298507462686e-06,
+      "loss": 0.4092,
+      "step": 24
+    },
+    {
+      "epoch": 0.0037385972783011813,
+      "grad_norm": 6.692920733889971,
+      "learning_rate": 1.2437810945273632e-06,
+      "loss": 0.4459,
+      "step": 25
+    },
+    {
+      "epoch": 0.0038881411694332286,
+      "grad_norm": 6.609492289627464,
+      "learning_rate": 1.2935323383084578e-06,
+      "loss": 0.4577,
+      "step": 26
+    },
+    {
+      "epoch": 0.004037685060565276,
+      "grad_norm": 4.9115623336358,
+      "learning_rate": 1.3432835820895524e-06,
+      "loss": 0.5349,
+      "step": 27
+    },
+    {
+      "epoch": 0.004187228951697323,
+      "grad_norm": 5.117676678055004,
+      "learning_rate": 1.393034825870647e-06,
+      "loss": 0.5483,
+      "step": 28
+    },
+    {
+      "epoch": 0.0043367728428293706,
+      "grad_norm": 5.263481949191207,
+      "learning_rate": 1.4427860696517414e-06,
+      "loss": 0.5991,
+      "step": 29
+    },
+    {
+      "epoch": 0.004486316733961417,
+      "grad_norm": 6.131569220022702,
+      "learning_rate": 1.4925373134328358e-06,
+      "loss": 0.3908,
+      "step": 30
+    },
+    {
+      "epoch": 0.004635860625093465,
+      "grad_norm": 5.928579435490833,
+      "learning_rate": 1.5422885572139304e-06,
+      "loss": 0.2084,
+      "step": 31
+    },
+    {
+      "epoch": 0.004785404516225512,
+      "grad_norm": 5.916757088180695,
+      "learning_rate": 1.592039800995025e-06,
+      "loss": 0.3858,
+      "step": 32
+    },
+    {
+      "epoch": 0.00493494840735756,
+      "grad_norm": 8.20423570651997,
+      "learning_rate": 1.6417910447761196e-06,
+      "loss": 0.2901,
+      "step": 33
+    },
+    {
+      "epoch": 0.005084492298489607,
+      "grad_norm": 8.219360009824356,
+      "learning_rate": 1.6915422885572142e-06,
+      "loss": 0.3919,
+      "step": 34
+    },
+    {
+      "epoch": 0.005234036189621654,
+      "grad_norm": 5.998450714995048,
+      "learning_rate": 1.7412935323383088e-06,
+      "loss": 0.2445,
+      "step": 35
+    },
+    {
+      "epoch": 0.005383580080753701,
+      "grad_norm": 4.267389037528284,
+      "learning_rate": 1.791044776119403e-06,
+      "loss": 0.2062,
+      "step": 36
+    },
+    {
+      "epoch": 0.005533123971885748,
+      "grad_norm": 5.463746992191978,
+      "learning_rate": 1.8407960199004975e-06,
+      "loss": 0.5357,
+      "step": 37
+    },
+    {
+      "epoch": 0.005682667863017796,
+      "grad_norm": 4.306281637510176,
+      "learning_rate": 1.8905472636815921e-06,
+      "loss": 0.1867,
+      "step": 38
+    },
+    {
+      "epoch": 0.005832211754149843,
+      "grad_norm": 6.551059942168939,
+      "learning_rate": 1.9402985074626867e-06,
+      "loss": 0.5944,
+      "step": 39
+    },
+    {
+      "epoch": 0.0059817556452818905,
+      "grad_norm": 6.110559490141819,
+      "learning_rate": 1.9900497512437813e-06,
+      "loss": 0.6173,
+      "step": 40
+    },
+    {
+      "epoch": 0.006131299536413937,
+      "grad_norm": 4.577457366278138,
+      "learning_rate": 2.0398009950248755e-06,
+      "loss": 0.3634,
+      "step": 41
+    },
+    {
+      "epoch": 0.006280843427545985,
+      "grad_norm": 6.020057986889502,
+      "learning_rate": 2.08955223880597e-06,
+      "loss": 0.5398,
+      "step": 42
+    },
+    {
+      "epoch": 0.006430387318678032,
+      "grad_norm": 12.119213807947853,
+      "learning_rate": 2.1393034825870647e-06,
+      "loss": 0.2376,
+      "step": 43
+    },
+    {
+      "epoch": 0.006579931209810079,
+      "grad_norm": 4.977979102095054,
+      "learning_rate": 2.1890547263681593e-06,
+      "loss": 0.2455,
+      "step": 44
+    },
+    {
+      "epoch": 0.006729475100942127,
+      "grad_norm": 3.4274663141099166,
+      "learning_rate": 2.238805970149254e-06,
+      "loss": 0.2356,
+      "step": 45
+    },
+    {
+      "epoch": 0.0068790189920741734,
+      "grad_norm": 4.552279062958819,
+      "learning_rate": 2.2885572139303485e-06,
+      "loss": 0.1681,
+      "step": 46
+    },
+    {
+      "epoch": 0.007028562883206221,
+      "grad_norm": 2.9323320786902496,
+      "learning_rate": 2.338308457711443e-06,
+      "loss": 0.2303,
+      "step": 47
+    },
+    {
+      "epoch": 0.007178106774338268,
+      "grad_norm": 4.623033466327724,
+      "learning_rate": 2.3880597014925373e-06,
+      "loss": 0.2404,
+      "step": 48
+    },
+    {
+      "epoch": 0.007327650665470316,
+      "grad_norm": 5.05007020882628,
+      "learning_rate": 2.437810945273632e-06,
+      "loss": 0.4128,
+      "step": 49
+    },
+    {
+      "epoch": 0.007477194556602363,
+      "grad_norm": 2.5237349934200273,
+      "learning_rate": 2.4875621890547264e-06,
+      "loss": 0.2196,
+      "step": 50
+    },
+    {
+      "epoch": 0.00762673844773441,
+      "grad_norm": 3.7483142878646594,
+      "learning_rate": 2.537313432835821e-06,
+      "loss": 0.1725,
+      "step": 51
+    },
+    {
+      "epoch": 0.007776282338866457,
+      "grad_norm": 4.032155563605261,
+      "learning_rate": 2.5870646766169156e-06,
+      "loss": 0.3821,
+      "step": 52
+    },
+    {
+      "epoch": 0.007925826229998505,
+      "grad_norm": 3.7782327104964333,
+      "learning_rate": 2.6368159203980102e-06,
+      "loss": 0.2207,
+      "step": 53
+    },
+    {
+      "epoch": 0.008075370121130552,
+      "grad_norm": 4.816720331969929,
+      "learning_rate": 2.686567164179105e-06,
+      "loss": 0.2265,
+      "step": 54
+    },
+    {
+      "epoch": 0.008224914012262599,
+      "grad_norm": 2.8481845548797478,
+      "learning_rate": 2.736318407960199e-06,
+      "loss": 0.2174,
+      "step": 55
+    },
+    {
+      "epoch": 0.008374457903394646,
+      "grad_norm": 4.501151176073331,
+      "learning_rate": 2.786069651741294e-06,
+      "loss": 0.2306,
+      "step": 56
+    },
+    {
+      "epoch": 0.008524001794526694,
+      "grad_norm": 4.326693136186164,
+      "learning_rate": 2.835820895522388e-06,
+      "loss": 0.4023,
+      "step": 57
+    },
+    {
+      "epoch": 0.008673545685658741,
+      "grad_norm": 4.061925818141106,
+      "learning_rate": 2.885572139303483e-06,
+      "loss": 0.7602,
+      "step": 58
+    },
+    {
+      "epoch": 0.008823089576790788,
+      "grad_norm": 6.144988240043741,
+      "learning_rate": 2.9353233830845774e-06,
+      "loss": 0.4451,
+      "step": 59
+    },
+    {
+      "epoch": 0.008972633467922835,
+      "grad_norm": 4.985549166627373,
+      "learning_rate": 2.9850746268656716e-06,
+      "loss": 0.4621,
+      "step": 60
+    },
+    {
+      "epoch": 0.009122177359054883,
+      "grad_norm": 3.192079125281125,
+      "learning_rate": 3.0348258706467666e-06,
+      "loss": 0.3694,
+      "step": 61
+    },
+    {
+      "epoch": 0.00927172125018693,
+      "grad_norm": 4.653619400771914,
+      "learning_rate": 3.0845771144278608e-06,
+      "loss": 0.2416,
+      "step": 62
+    },
+    {
+      "epoch": 0.009421265141318977,
+      "grad_norm": 3.4214006556775156,
+      "learning_rate": 3.1343283582089558e-06,
+      "loss": 0.4755,
+      "step": 63
+    },
+    {
+      "epoch": 0.009570809032451024,
+      "grad_norm": 3.0809019894250613,
+      "learning_rate": 3.18407960199005e-06,
+      "loss": 0.4154,
+      "step": 64
+    },
+    {
+      "epoch": 0.009720352923583071,
+      "grad_norm": 4.190290076677796,
+      "learning_rate": 3.233830845771145e-06,
+      "loss": 0.4362,
+      "step": 65
+    },
+    {
+      "epoch": 0.00986989681471512,
+      "grad_norm": 3.1777725686355356,
+      "learning_rate": 3.283582089552239e-06,
+      "loss": 0.3635,
+      "step": 66
+    },
+    {
+      "epoch": 0.010019440705847166,
+      "grad_norm": 2.592442539170553,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.1739,
+      "step": 67
+    },
+    {
+      "epoch": 0.010168984596979213,
+      "grad_norm": 4.610893839801018,
+      "learning_rate": 3.3830845771144283e-06,
+      "loss": 0.3845,
+      "step": 68
+    },
+    {
+      "epoch": 0.01031852848811126,
+      "grad_norm": 2.941030939381248,
+      "learning_rate": 3.4328358208955225e-06,
+      "loss": 0.226,
+      "step": 69
+    },
+    {
+      "epoch": 0.010468072379243309,
+      "grad_norm": 2.641062959772403,
+      "learning_rate": 3.4825870646766175e-06,
+      "loss": 0.2083,
+      "step": 70
+    },
+    {
+      "epoch": 0.010617616270375356,
+      "grad_norm": 4.573399002022637,
+      "learning_rate": 3.5323383084577117e-06,
+      "loss": 0.3639,
+      "step": 71
+    },
+    {
+      "epoch": 0.010767160161507403,
+      "grad_norm": 3.811597787697304,
+      "learning_rate": 3.582089552238806e-06,
+      "loss": 0.2046,
+      "step": 72
+    },
+    {
+      "epoch": 0.01091670405263945,
+      "grad_norm": 7.593654702612937,
+      "learning_rate": 3.631840796019901e-06,
+      "loss": 0.3831,
+      "step": 73
+    },
+    {
+      "epoch": 0.011066247943771496,
+      "grad_norm": 2.6372126137968013,
+      "learning_rate": 3.681592039800995e-06,
+      "loss": 0.2155,
+      "step": 74
+    },
+    {
+      "epoch": 0.011215791834903545,
+      "grad_norm": 3.401033168780161,
+      "learning_rate": 3.73134328358209e-06,
+      "loss": 0.2439,
+      "step": 75
+    },
+    {
+      "epoch": 0.011365335726035592,
+      "grad_norm": 2.8172647382036047,
+      "learning_rate": 3.7810945273631843e-06,
+      "loss": 0.1614,
+      "step": 76
+    },
+    {
+      "epoch": 0.011514879617167639,
+      "grad_norm": 3.525793180439174,
+      "learning_rate": 3.8308457711442784e-06,
+      "loss": 0.2176,
+      "step": 77
+    },
+    {
+      "epoch": 0.011664423508299685,
+      "grad_norm": 2.4029805525684527,
+      "learning_rate": 3.8805970149253735e-06,
+      "loss": 0.1893,
+      "step": 78
+    },
+    {
+      "epoch": 0.011813967399431732,
+      "grad_norm": 5.727795685387504,
+      "learning_rate": 3.930348258706468e-06,
+      "loss": 0.5702,
+      "step": 79
+    },
+    {
+      "epoch": 0.011963511290563781,
+      "grad_norm": 4.021893784746645,
+      "learning_rate": 3.980099502487563e-06,
+      "loss": 0.4027,
+      "step": 80
+    },
+    {
+      "epoch": 0.012113055181695828,
+      "grad_norm": 2.7773808558650535,
+      "learning_rate": 4.029850746268657e-06,
+      "loss": 0.2963,
+      "step": 81
+    },
+    {
+      "epoch": 0.012262599072827875,
+      "grad_norm": 3.4349426033049992,
+      "learning_rate": 4.079601990049751e-06,
+      "loss": 0.2211,
+      "step": 82
+    },
+    {
+      "epoch": 0.012412142963959922,
+      "grad_norm": 4.127258766074891,
+      "learning_rate": 4.129353233830846e-06,
+      "loss": 0.2516,
+      "step": 83
+    },
+    {
+      "epoch": 0.01256168685509197,
+      "grad_norm": 3.551977981988865,
+      "learning_rate": 4.17910447761194e-06,
+      "loss": 0.2206,
+      "step": 84
+    },
+    {
+      "epoch": 0.012711230746224017,
+      "grad_norm": 2.988554589230421,
+      "learning_rate": 4.228855721393035e-06,
+      "loss": 0.366,
+      "step": 85
+    },
+    {
+      "epoch": 0.012860774637356064,
+      "grad_norm": 3.256233912334862,
+      "learning_rate": 4.278606965174129e-06,
+      "loss": 0.341,
+      "step": 86
+    },
+    {
+      "epoch": 0.01301031852848811,
+      "grad_norm": 3.917242635149468,
+      "learning_rate": 4.3283582089552236e-06,
+      "loss": 0.281,
+      "step": 87
+    },
+    {
+      "epoch": 0.013159862419620158,
+      "grad_norm": 3.8372869351661247,
+      "learning_rate": 4.378109452736319e-06,
+      "loss": 0.1933,
+      "step": 88
+    },
+    {
+      "epoch": 0.013309406310752206,
+      "grad_norm": 4.03192980896834,
+      "learning_rate": 4.427860696517413e-06,
+      "loss": 0.184,
+      "step": 89
+    },
+    {
+      "epoch": 0.013458950201884253,
+      "grad_norm": 4.944440623197377,
+      "learning_rate": 4.477611940298508e-06,
+      "loss": 0.2406,
+      "step": 90
+    },
+    {
+      "epoch": 0.0136084940930163,
+      "grad_norm": 3.2771345760625916,
+      "learning_rate": 4.527363184079602e-06,
+      "loss": 0.3635,
+      "step": 91
+    },
+    {
+      "epoch": 0.013758037984148347,
+      "grad_norm": 2.5552685161479913,
+      "learning_rate": 4.577114427860697e-06,
+      "loss": 0.3581,
+      "step": 92
+    },
+    {
+      "epoch": 0.013907581875280395,
+      "grad_norm": 3.825258197515859,
+      "learning_rate": 4.626865671641791e-06,
+      "loss": 0.2157,
+      "step": 93
+    },
+    {
+      "epoch": 0.014057125766412442,
+      "grad_norm": 3.820006828326968,
+      "learning_rate": 4.676616915422886e-06,
+      "loss": 0.401,
+      "step": 94
+    },
+    {
+      "epoch": 0.01420666965754449,
+      "grad_norm": 3.4269639891084056,
+      "learning_rate": 4.72636815920398e-06,
+      "loss": 0.21,
+      "step": 95
+    },
+    {
+      "epoch": 0.014356213548676536,
+      "grad_norm": 3.614177044324435,
+      "learning_rate": 4.7761194029850745e-06,
+      "loss": 0.2305,
+      "step": 96
+    },
+    {
+      "epoch": 0.014505757439808583,
+      "grad_norm": 2.8474787904051633,
+      "learning_rate": 4.8258706467661695e-06,
+      "loss": 0.2002,
+      "step": 97
+    },
+    {
+      "epoch": 0.014655301330940632,
+      "grad_norm": 3.1529185682156333,
+      "learning_rate": 4.875621890547264e-06,
+      "loss": 0.3126,
+      "step": 98
+    },
+    {
+      "epoch": 0.014804845222072678,
+      "grad_norm": 2.805579699726101,
+      "learning_rate": 4.925373134328359e-06,
+      "loss": 0.3977,
+      "step": 99
+    },
+    {
+      "epoch": 0.014954389113204725,
+      "grad_norm": 2.5072872378288134,
+      "learning_rate": 4.975124378109453e-06,
+      "loss": 0.1986,
+      "step": 100
+    },
+    {
+      "epoch": 0.015103933004336772,
+      "grad_norm": 2.8773082972301816,
+      "learning_rate": 5.024875621890548e-06,
+      "loss": 0.2421,
+      "step": 101
+    },
+    {
+      "epoch": 0.01525347689546882,
+      "grad_norm": 2.3650776175631765,
+      "learning_rate": 5.074626865671642e-06,
+      "loss": 0.1864,
+      "step": 102
+    },
+    {
+      "epoch": 0.015403020786600868,
+      "grad_norm": 4.721891286027898,
+      "learning_rate": 5.124378109452737e-06,
+      "loss": 0.2939,
+      "step": 103
+    },
+    {
+      "epoch": 0.015552564677732915,
+      "grad_norm": 2.6753396233648705,
+      "learning_rate": 5.174129353233831e-06,
+      "loss": 0.2558,
+      "step": 104
+    },
+    {
+      "epoch": 0.01570210856886496,
+      "grad_norm": 3.149876968312327,
+      "learning_rate": 5.2238805970149255e-06,
+      "loss": 0.3405,
+      "step": 105
+    },
+    {
+      "epoch": 0.01585165245999701,
+      "grad_norm": 1.6322197066205648,
+      "learning_rate": 5.2736318407960205e-06,
+      "loss": 0.1453,
+      "step": 106
+    },
+    {
+      "epoch": 0.016001196351129055,
+      "grad_norm": 3.3492234789043236,
+      "learning_rate": 5.323383084577115e-06,
+      "loss": 0.404,
+      "step": 107
+    },
+    {
+      "epoch": 0.016150740242261104,
+      "grad_norm": 2.2518951047915157,
+      "learning_rate": 5.37313432835821e-06,
+      "loss": 0.2278,
+      "step": 108
+    },
+    {
+      "epoch": 0.016300284133393152,
+      "grad_norm": 3.0471913491370404,
+      "learning_rate": 5.422885572139304e-06,
+      "loss": 0.265,
+      "step": 109
+    },
+    {
+      "epoch": 0.016449828024525198,
+      "grad_norm": 1.6928519222295142,
+      "learning_rate": 5.472636815920398e-06,
+      "loss": 0.2169,
+      "step": 110
+    },
+    {
+      "epoch": 0.016599371915657246,
+      "grad_norm": 3.265018826674296,
+      "learning_rate": 5.522388059701493e-06,
+      "loss": 0.429,
+      "step": 111
+    },
+    {
+      "epoch": 0.01674891580678929,
+      "grad_norm": 2.637671664378066,
+      "learning_rate": 5.572139303482588e-06,
+      "loss": 0.2762,
+      "step": 112
+    },
+    {
+      "epoch": 0.01689845969792134,
+      "grad_norm": 3.1617986987096134,
+      "learning_rate": 5.621890547263682e-06,
+      "loss": 0.4272,
+      "step": 113
+    },
+    {
+      "epoch": 0.01704800358905339,
+      "grad_norm": 3.0132316717807175,
+      "learning_rate": 5.671641791044776e-06,
+      "loss": 0.3644,
+      "step": 114
+    },
+    {
+      "epoch": 0.017197547480185434,
+      "grad_norm": 2.2850314864309813,
+      "learning_rate": 5.721393034825871e-06,
+      "loss": 0.1967,
+      "step": 115
+    },
+    {
+      "epoch": 0.017347091371317482,
+      "grad_norm": 3.0835871860462314,
+      "learning_rate": 5.771144278606966e-06,
+      "loss": 0.2322,
+      "step": 116
+    },
+    {
+      "epoch": 0.017496635262449527,
+      "grad_norm": 3.5275796788122893,
+      "learning_rate": 5.820895522388061e-06,
+      "loss": 0.3543,
+      "step": 117
+    },
+    {
+      "epoch": 0.017646179153581576,
+      "grad_norm": 3.1301356173345494,
+      "learning_rate": 5.870646766169155e-06,
+      "loss": 0.5064,
+      "step": 118
+    },
+    {
+      "epoch": 0.017795723044713625,
+      "grad_norm": 3.9689250366780313,
+      "learning_rate": 5.920398009950249e-06,
+      "loss": 0.8428,
+      "step": 119
+    },
+    {
+      "epoch": 0.01794526693584567,
+      "grad_norm": 2.6992548320472984,
+      "learning_rate": 5.970149253731343e-06,
+      "loss": 0.2727,
+      "step": 120
+    },
+    {
+      "epoch": 0.01809481082697772,
+      "grad_norm": 2.8823271138601414,
+      "learning_rate": 6.019900497512439e-06,
+      "loss": 0.3301,
+      "step": 121
+    },
+    {
+      "epoch": 0.018244354718109767,
+      "grad_norm": 2.652199321292131,
+      "learning_rate": 6.069651741293533e-06,
+      "loss": 0.234,
+      "step": 122
+    },
+    {
+      "epoch": 0.018393898609241812,
+      "grad_norm": 4.008459949806747,
+      "learning_rate": 6.119402985074627e-06,
+      "loss": 0.5713,
+      "step": 123
+    },
+    {
+      "epoch": 0.01854344250037386,
+      "grad_norm": 2.8867543983581236,
+      "learning_rate": 6.1691542288557215e-06,
+      "loss": 0.2146,
+      "step": 124
+    },
+    {
+      "epoch": 0.018692986391505906,
+      "grad_norm": 2.379666412119815,
+      "learning_rate": 6.218905472636816e-06,
+      "loss": 0.3812,
+      "step": 125
+    },
+    {
+      "epoch": 0.018842530282637954,
+      "grad_norm": 2.8364015730213716,
+      "learning_rate": 6.2686567164179116e-06,
+      "loss": 0.3729,
+      "step": 126
+    },
+    {
+      "epoch": 0.018992074173770003,
+      "grad_norm": 2.9731590306978957,
+      "learning_rate": 6.318407960199006e-06,
+      "loss": 0.3922,
+      "step": 127
+    },
+    {
+      "epoch": 0.019141618064902048,
+      "grad_norm": 2.431931443805707,
+      "learning_rate": 6.3681592039801e-06,
+      "loss": 0.2316,
+      "step": 128
+    },
+    {
+      "epoch": 0.019291161956034097,
+      "grad_norm": 2.5964092588685594,
+      "learning_rate": 6.417910447761194e-06,
+      "loss": 0.2129,
+      "step": 129
+    },
+    {
+      "epoch": 0.019440705847166142,
+      "grad_norm": 4.241711858566103,
+      "learning_rate": 6.46766169154229e-06,
+      "loss": 0.2677,
+      "step": 130
+    },
+    {
+      "epoch": 0.01959024973829819,
+      "grad_norm": 3.743763522090278,
+      "learning_rate": 6.517412935323384e-06,
+      "loss": 0.7324,
+      "step": 131
+    },
+    {
+      "epoch": 0.01973979362943024,
+      "grad_norm": 2.325325226468886,
+      "learning_rate": 6.567164179104478e-06,
+      "loss": 0.2282,
+      "step": 132
+    },
+    {
+      "epoch": 0.019889337520562284,
+      "grad_norm": 2.187485810642544,
+      "learning_rate": 6.6169154228855725e-06,
+      "loss": 0.3479,
+      "step": 133
+    },
+    {
+      "epoch": 0.020038881411694333,
+      "grad_norm": 2.555235252803596,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.3084,
+      "step": 134
+    },
+    {
+      "epoch": 0.020188425302826378,
+      "grad_norm": 2.1409254211343405,
+      "learning_rate": 6.7164179104477625e-06,
+      "loss": 0.2413,
+      "step": 135
+    },
+    {
+      "epoch": 0.020337969193958427,
+      "grad_norm": 2.9475030013466292,
+      "learning_rate": 6.766169154228857e-06,
+      "loss": 0.5899,
+      "step": 136
+    },
+    {
+      "epoch": 0.020487513085090475,
+      "grad_norm": 3.161190387153201,
+      "learning_rate": 6.815920398009951e-06,
+      "loss": 0.2722,
+      "step": 137
+    },
+    {
+      "epoch": 0.02063705697622252,
+      "grad_norm": 3.4231688087143786,
+      "learning_rate": 6.865671641791045e-06,
+      "loss": 0.25,
+      "step": 138
+    },
+    {
+      "epoch": 0.02078660086735457,
+      "grad_norm": 2.891852432700459,
+      "learning_rate": 6.915422885572139e-06,
+      "loss": 0.5206,
+      "step": 139
+    },
+    {
+      "epoch": 0.020936144758486618,
+      "grad_norm": 2.4149596821734645,
+      "learning_rate": 6.965174129353235e-06,
+      "loss": 0.2792,
+      "step": 140
+    },
+    {
+      "epoch": 0.021085688649618663,
+      "grad_norm": 2.737327253049286,
+      "learning_rate": 7.014925373134329e-06,
+      "loss": 0.1785,
+      "step": 141
+    },
+    {
+      "epoch": 0.02123523254075071,
+      "grad_norm": 2.271710572333297,
+      "learning_rate": 7.064676616915423e-06,
+      "loss": 0.2216,
+      "step": 142
+    },
+    {
+      "epoch": 0.021384776431882756,
+      "grad_norm": 3.123818135886555,
+      "learning_rate": 7.114427860696518e-06,
+      "loss": 0.5292,
+      "step": 143
+    },
+    {
+      "epoch": 0.021534320323014805,
+      "grad_norm": 3.4353230085188775,
+      "learning_rate": 7.164179104477612e-06,
+      "loss": 0.257,
+      "step": 144
+    },
+    {
+      "epoch": 0.021683864214146854,
+      "grad_norm": 3.292198842322858,
+      "learning_rate": 7.213930348258708e-06,
+      "loss": 0.4413,
+      "step": 145
+    },
+    {
+      "epoch": 0.0218334081052789,
+      "grad_norm": 2.408669543365234,
+      "learning_rate": 7.263681592039802e-06,
+      "loss": 0.4034,
+      "step": 146
+    },
+    {
+      "epoch": 0.021982951996410947,
+      "grad_norm": 2.918318139010717,
+      "learning_rate": 7.313432835820896e-06,
+      "loss": 0.1789,
+      "step": 147
+    },
+    {
+      "epoch": 0.022132495887542993,
+      "grad_norm": 2.016064943310167,
+      "learning_rate": 7.36318407960199e-06,
+      "loss": 0.2454,
+      "step": 148
+    },
+    {
+      "epoch": 0.02228203977867504,
+      "grad_norm": 3.375282717272202,
+      "learning_rate": 7.412935323383084e-06,
+      "loss": 0.5047,
+      "step": 149
+    },
+    {
+      "epoch": 0.02243158366980709,
+      "grad_norm": 2.747548142801912,
+      "learning_rate": 7.46268656716418e-06,
+      "loss": 0.3193,
+      "step": 150
+    },
+    {
+      "epoch": 0.022581127560939135,
+      "grad_norm": 5.014531999850111,
+      "learning_rate": 7.512437810945274e-06,
+      "loss": 0.5367,
+      "step": 151
+    },
+    {
+      "epoch": 0.022730671452071183,
+      "grad_norm": 1.7396197448467992,
+      "learning_rate": 7.5621890547263685e-06,
+      "loss": 0.1602,
+      "step": 152
+    },
+    {
+      "epoch": 0.02288021534320323,
+      "grad_norm": 3.9271159318267452,
+      "learning_rate": 7.611940298507463e-06,
+      "loss": 0.2763,
+      "step": 153
+    },
+    {
+      "epoch": 0.023029759234335277,
+      "grad_norm": 2.093726492507833,
+      "learning_rate": 7.661691542288557e-06,
+      "loss": 0.169,
+      "step": 154
+    },
+    {
+      "epoch": 0.023179303125467326,
+      "grad_norm": 1.5357011381308088,
+      "learning_rate": 7.711442786069654e-06,
+      "loss": 0.1619,
+      "step": 155
+    },
+    {
+      "epoch": 0.02332884701659937,
+      "grad_norm": 2.3824458230974863,
+      "learning_rate": 7.761194029850747e-06,
+      "loss": 0.2094,
+      "step": 156
+    },
+    {
+      "epoch": 0.02347839090773142,
+      "grad_norm": 2.8236663879690784,
+      "learning_rate": 7.810945273631842e-06,
+      "loss": 0.3426,
+      "step": 157
+    },
+    {
+      "epoch": 0.023627934798863465,
+      "grad_norm": 3.1375695638809815,
+      "learning_rate": 7.860696517412935e-06,
+      "loss": 0.5518,
+      "step": 158
+    },
+    {
+      "epoch": 0.023777478689995513,
+      "grad_norm": 3.2182906468856105,
+      "learning_rate": 7.91044776119403e-06,
+      "loss": 0.1995,
+      "step": 159
+    },
+    {
+      "epoch": 0.023927022581127562,
+      "grad_norm": 14.749841980168513,
+      "learning_rate": 7.960199004975125e-06,
+      "loss": 0.5578,
+      "step": 160
+    },
+    {
+      "epoch": 0.024076566472259607,
+      "grad_norm": 3.0100123201004045,
+      "learning_rate": 8.00995024875622e-06,
+      "loss": 0.5091,
+      "step": 161
+    },
+    {
+      "epoch": 0.024226110363391656,
+      "grad_norm": 3.5091520525666433,
+      "learning_rate": 8.059701492537314e-06,
+      "loss": 0.5357,
+      "step": 162
+    },
+    {
+      "epoch": 0.024375654254523704,
+      "grad_norm": 2.934851375582722,
+      "learning_rate": 8.109452736318409e-06,
+      "loss": 0.2267,
+      "step": 163
+    },
+    {
+      "epoch": 0.02452519814565575,
+      "grad_norm": 2.5911339240383544,
+      "learning_rate": 8.159203980099502e-06,
+      "loss": 0.1782,
+      "step": 164
+    },
+    {
+      "epoch": 0.024674742036787798,
+      "grad_norm": 2.847206263316536,
+      "learning_rate": 8.208955223880599e-06,
+      "loss": 0.2252,
+      "step": 165
+    },
+    {
+      "epoch": 0.024824285927919843,
+      "grad_norm": 3.5380431553535976,
+      "learning_rate": 8.258706467661692e-06,
+      "loss": 0.4295,
+      "step": 166
+    },
+    {
+      "epoch": 0.024973829819051892,
+      "grad_norm": 3.150492354924513,
+      "learning_rate": 8.308457711442787e-06,
+      "loss": 0.3276,
+      "step": 167
+    },
+    {
+      "epoch": 0.02512337371018394,
+      "grad_norm": 3.114695975436696,
+      "learning_rate": 8.35820895522388e-06,
+      "loss": 0.5181,
+      "step": 168
+    },
+    {
+      "epoch": 0.025272917601315985,
+      "grad_norm": 2.6180846619509355,
+      "learning_rate": 8.407960199004975e-06,
+      "loss": 0.2577,
+      "step": 169
+    },
+    {
+      "epoch": 0.025422461492448034,
+      "grad_norm": 1.859950631659999,
+      "learning_rate": 8.45771144278607e-06,
+      "loss": 0.1838,
+      "step": 170
+    },
+    {
+      "epoch": 0.02557200538358008,
+      "grad_norm": 4.092195798232618,
+      "learning_rate": 8.507462686567165e-06,
+      "loss": 0.2676,
+      "step": 171
+    },
+    {
+      "epoch": 0.025721549274712128,
+      "grad_norm": 2.0820308098425766,
+      "learning_rate": 8.557213930348259e-06,
+      "loss": 0.2528,
+      "step": 172
+    },
+    {
+      "epoch": 0.025871093165844176,
+      "grad_norm": 2.8153771201369087,
+      "learning_rate": 8.606965174129354e-06,
+      "loss": 0.3374,
+      "step": 173
+    },
+    {
+      "epoch": 0.02602063705697622,
+      "grad_norm": 2.6417342231989114,
+      "learning_rate": 8.656716417910447e-06,
+      "loss": 0.4309,
+      "step": 174
+    },
+    {
+      "epoch": 0.02617018094810827,
+      "grad_norm": 3.3553357791865825,
+      "learning_rate": 8.706467661691544e-06,
+      "loss": 0.279,
+      "step": 175
+    },
+    {
+      "epoch": 0.026319724839240315,
+      "grad_norm": 2.5896987414147707,
+      "learning_rate": 8.756218905472637e-06,
+      "loss": 0.2505,
+      "step": 176
+    },
+    {
+      "epoch": 0.026469268730372364,
+      "grad_norm": 15.917959164107543,
+      "learning_rate": 8.805970149253732e-06,
+      "loss": 0.3903,
+      "step": 177
+    },
+    {
+      "epoch": 0.026618812621504413,
+      "grad_norm": 1.897502276352634,
+      "learning_rate": 8.855721393034826e-06,
+      "loss": 0.3051,
+      "step": 178
+    },
+    {
+      "epoch": 0.026768356512636458,
+      "grad_norm": 3.498345426750877,
+      "learning_rate": 8.905472636815922e-06,
+      "loss": 0.8122,
+      "step": 179
+    },
+    {
+      "epoch": 0.026917900403768506,
+      "grad_norm": 3.2270107650642297,
+      "learning_rate": 8.955223880597016e-06,
+      "loss": 0.2312,
+      "step": 180
+    },
+    {
+      "epoch": 0.027067444294900555,
+      "grad_norm": 2.373617987334166,
+      "learning_rate": 9.00497512437811e-06,
+      "loss": 0.3553,
+      "step": 181
+    },
+    {
+      "epoch": 0.0272169881860326,
+      "grad_norm": 2.022495433415561,
+      "learning_rate": 9.054726368159204e-06,
+      "loss": 0.3372,
+      "step": 182
+    },
+    {
+      "epoch": 0.02736653207716465,
+      "grad_norm": 2.471303542690233,
+      "learning_rate": 9.104477611940299e-06,
+      "loss": 0.2764,
+      "step": 183
+    },
+    {
+      "epoch": 0.027516075968296694,
+      "grad_norm": 2.170550660433261,
+      "learning_rate": 9.154228855721394e-06,
+      "loss": 0.2429,
+      "step": 184
+    },
+    {
+      "epoch": 0.027665619859428742,
+      "grad_norm": 1.7750572924031363,
+      "learning_rate": 9.203980099502489e-06,
+      "loss": 0.1749,
+      "step": 185
+    },
+    {
+      "epoch": 0.02781516375056079,
+      "grad_norm": 1.9803173977955488,
+      "learning_rate": 9.253731343283582e-06,
+      "loss": 0.3061,
+      "step": 186
+    },
+    {
+      "epoch": 0.027964707641692836,
+      "grad_norm": 2.686793479118654,
+      "learning_rate": 9.303482587064677e-06,
+      "loss": 0.2704,
+      "step": 187
+    },
+    {
+      "epoch": 0.028114251532824885,
+      "grad_norm": 3.0095995560762088,
+      "learning_rate": 9.353233830845772e-06,
+      "loss": 0.3935,
+      "step": 188
+    },
+    {
+      "epoch": 0.02826379542395693,
+      "grad_norm": 3.296780241377357,
+      "learning_rate": 9.402985074626867e-06,
+      "loss": 0.4349,
+      "step": 189
+    },
+    {
+      "epoch": 0.02841333931508898,
+      "grad_norm": 2.0473844316492262,
+      "learning_rate": 9.45273631840796e-06,
+      "loss": 0.3594,
+      "step": 190
+    },
+    {
+      "epoch": 0.028562883206221027,
+      "grad_norm": 2.6746439974295986,
+      "learning_rate": 9.502487562189056e-06,
+      "loss": 0.2507,
+      "step": 191
+    },
+    {
+      "epoch": 0.028712427097353072,
+      "grad_norm": 2.171372767224107,
+      "learning_rate": 9.552238805970149e-06,
+      "loss": 0.4442,
+      "step": 192
+    },
+    {
+      "epoch": 0.02886197098848512,
+      "grad_norm": 3.412610878033882,
+      "learning_rate": 9.601990049751244e-06,
+      "loss": 0.5065,
+      "step": 193
+    },
+    {
+      "epoch": 0.029011514879617166,
+      "grad_norm": 2.5249672849820843,
+      "learning_rate": 9.651741293532339e-06,
+      "loss": 0.2775,
+      "step": 194
+    },
+    {
+      "epoch": 0.029161058770749215,
+      "grad_norm": 1.9244063665371054,
+      "learning_rate": 9.701492537313434e-06,
+      "loss": 0.2501,
+      "step": 195
+    },
+    {
+      "epoch": 0.029310602661881263,
+      "grad_norm": 2.2928756876943788,
+      "learning_rate": 9.751243781094527e-06,
+      "loss": 0.391,
+      "step": 196
+    },
+    {
+      "epoch": 0.02946014655301331,
+      "grad_norm": 3.2090175671059464,
+      "learning_rate": 9.800995024875622e-06,
+      "loss": 0.355,
+      "step": 197
+    },
+    {
+      "epoch": 0.029609690444145357,
+      "grad_norm": 2.564275054094989,
+      "learning_rate": 9.850746268656717e-06,
+      "loss": 0.3824,
+      "step": 198
+    },
+    {
+      "epoch": 0.029759234335277406,
+      "grad_norm": 2.2612313847384473,
+      "learning_rate": 9.900497512437812e-06,
+      "loss": 0.255,
+      "step": 199
+    },
+    {
+      "epoch": 0.02990877822640945,
+      "grad_norm": 2.867410801811384,
+      "learning_rate": 9.950248756218906e-06,
+      "loss": 0.2321,
+      "step": 200
+    },
+    {
+      "epoch": 0.0300583221175415,
+      "grad_norm": 2.7017080308625316,
+      "learning_rate": 1e-05,
+      "loss": 0.5355,
+      "step": 201
+    },
+    {
+      "epoch": 0.030207866008673544,
+      "grad_norm": 1.7563631058650533,
+      "learning_rate": 9.999999413475907e-06,
+      "loss": 0.2366,
+      "step": 202
+    },
+    {
+      "epoch": 0.030357409899805593,
+      "grad_norm": 2.7923486514729134,
+      "learning_rate": 9.999997653903764e-06,
+      "loss": 0.5735,
+      "step": 203
+    },
+    {
+      "epoch": 0.03050695379093764,
+      "grad_norm": 2.5477270678585935,
+      "learning_rate": 9.999994721283985e-06,
+      "loss": 0.2316,
+      "step": 204
+    },
+    {
+      "epoch": 0.030656497682069687,
+      "grad_norm": 1.6435827637040603,
+      "learning_rate": 9.99999061561726e-06,
+      "loss": 0.1958,
+      "step": 205
+    },
+    {
+      "epoch": 0.030806041573201735,
+      "grad_norm": 4.225438559077688,
+      "learning_rate": 9.999985336904546e-06,
+      "loss": 0.6052,
+      "step": 206
+    },
+    {
+      "epoch": 0.03095558546433378,
+      "grad_norm": 2.384218907777814,
+      "learning_rate": 9.999978885147086e-06,
+      "loss": 0.382,
+      "step": 207
+    },
+    {
+      "epoch": 0.03110512935546583,
+      "grad_norm": 3.082533240684358,
+      "learning_rate": 9.999971260346394e-06,
+      "loss": 0.4615,
+      "step": 208
+    },
+    {
+      "epoch": 0.03125467324659788,
+      "grad_norm": 2.126341746782405,
+      "learning_rate": 9.999962462504259e-06,
+      "loss": 0.3489,
+      "step": 209
+    },
+    {
+      "epoch": 0.03140421713772992,
+      "grad_norm": 2.3157719584793974,
+      "learning_rate": 9.99995249162274e-06,
+      "loss": 0.351,
+      "step": 210
+    },
+    {
+      "epoch": 0.03155376102886197,
+      "grad_norm": 3.2569828989709046,
+      "learning_rate": 9.999941347704183e-06,
+      "loss": 0.5452,
+      "step": 211
+    },
+    {
+      "epoch": 0.03170330491999402,
+      "grad_norm": 2.4010549422177747,
+      "learning_rate": 9.999929030751199e-06,
+      "loss": 0.5511,
+      "step": 212
+    },
+    {
+      "epoch": 0.031852848811126065,
+      "grad_norm": 2.2021354319659956,
+      "learning_rate": 9.999915540766679e-06,
+      "loss": 0.409,
+      "step": 213
+    },
+    {
+      "epoch": 0.03200239270225811,
+      "grad_norm": 2.7467598032746467,
+      "learning_rate": 9.999900877753786e-06,
+      "loss": 0.2769,
+      "step": 214
+    },
+    {
+      "epoch": 0.03215193659339016,
+      "grad_norm": 2.250991470386846,
+      "learning_rate": 9.99988504171596e-06,
+      "loss": 0.4243,
+      "step": 215
+    },
+    {
+      "epoch": 0.03230148048452221,
+      "grad_norm": 7.389570164962262,
+      "learning_rate": 9.999868032656921e-06,
+      "loss": 0.5661,
+      "step": 216
+    },
+    {
+      "epoch": 0.03245102437565425,
+      "grad_norm": 2.3232325152419904,
+      "learning_rate": 9.999849850580653e-06,
+      "loss": 0.3622,
+      "step": 217
+    },
+    {
+      "epoch": 0.032600568266786305,
+      "grad_norm": 2.8448629192721153,
+      "learning_rate": 9.999830495491425e-06,
+      "loss": 0.5013,
+      "step": 218
+    },
+    {
+      "epoch": 0.03275011215791835,
+      "grad_norm": 1.9203985094095042,
+      "learning_rate": 9.99980996739378e-06,
+      "loss": 0.2597,
+      "step": 219
+    },
+    {
+      "epoch": 0.032899656049050395,
+      "grad_norm": 2.1343351176097705,
+      "learning_rate": 9.99978826629253e-06,
+      "loss": 0.333,
+      "step": 220
+    },
+    {
+      "epoch": 0.03304919994018244,
+      "grad_norm": 2.675496675158128,
+      "learning_rate": 9.999765392192766e-06,
+      "loss": 0.4679,
+      "step": 221
+    },
+    {
+      "epoch": 0.03319874383131449,
+      "grad_norm": 2.954897252892918,
+      "learning_rate": 9.99974134509986e-06,
+      "loss": 0.5779,
+      "step": 222
+    },
+    {
+      "epoch": 0.03334828772244654,
+      "grad_norm": 3.164155125145253,
+      "learning_rate": 9.999716125019448e-06,
+      "loss": 0.5192,
+      "step": 223
+    },
+    {
+      "epoch": 0.03349783161357858,
+      "grad_norm": 2.9422429580445377,
+      "learning_rate": 9.99968973195745e-06,
+      "loss": 0.3514,
+      "step": 224
+    },
+    {
+      "epoch": 0.033647375504710635,
+      "grad_norm": 2.016818218277119,
+      "learning_rate": 9.999662165920056e-06,
+      "loss": 0.3657,
+      "step": 225
+    },
+    {
+      "epoch": 0.03379691939584268,
+      "grad_norm": 2.805692301474297,
+      "learning_rate": 9.999633426913733e-06,
+      "loss": 0.1912,
+      "step": 226
+    },
+    {
+      "epoch": 0.033946463286974725,
+      "grad_norm": 2.205403428118743,
+      "learning_rate": 9.999603514945227e-06,
+      "loss": 0.234,
+      "step": 227
+    },
+    {
+      "epoch": 0.03409600717810678,
+      "grad_norm": 2.013271573198516,
+      "learning_rate": 9.999572430021553e-06,
+      "loss": 0.464,
+      "step": 228
+    },
+    {
+      "epoch": 0.03424555106923882,
+      "grad_norm": 3.033803346792209,
+      "learning_rate": 9.999540172150005e-06,
+      "loss": 0.2599,
+      "step": 229
+    },
+    {
+      "epoch": 0.03439509496037087,
+      "grad_norm": 2.854186400231596,
+      "learning_rate": 9.99950674133815e-06,
+      "loss": 0.6431,
+      "step": 230
+    },
+    {
+      "epoch": 0.03454463885150292,
+      "grad_norm": 2.162434347622467,
+      "learning_rate": 9.999472137593829e-06,
+      "loss": 0.4779,
+      "step": 231
+    },
+    {
+      "epoch": 0.034694182742634964,
+      "grad_norm": 1.4691335020169023,
+      "learning_rate": 9.999436360925165e-06,
+      "loss": 0.1827,
+      "step": 232
+    },
+    {
+      "epoch": 0.03484372663376701,
+      "grad_norm": 1.6955188606947214,
+      "learning_rate": 9.99939941134055e-06,
+      "loss": 0.2336,
+      "step": 233
+    },
+    {
+      "epoch": 0.034993270524899055,
+      "grad_norm": 2.0710606069082167,
+      "learning_rate": 9.99936128884865e-06,
+      "loss": 0.3671,
+      "step": 234
+    },
+    {
+      "epoch": 0.03514281441603111,
+      "grad_norm": 2.128464465717484,
+      "learning_rate": 9.999321993458411e-06,
+      "loss": 0.2928,
+      "step": 235
+    },
+    {
+      "epoch": 0.03529235830716315,
+      "grad_norm": 1.9685227247781487,
+      "learning_rate": 9.999281525179054e-06,
+      "loss": 0.185,
+      "step": 236
+    },
+    {
+      "epoch": 0.0354419021982952,
+      "grad_norm": 2.3203573768463115,
+      "learning_rate": 9.99923988402007e-06,
+      "loss": 0.3733,
+      "step": 237
+    },
+    {
+      "epoch": 0.03559144608942725,
+      "grad_norm": 2.2161639851963457,
+      "learning_rate": 9.99919706999123e-06,
+      "loss": 0.4,
+      "step": 238
+    },
+    {
+      "epoch": 0.035740989980559294,
+      "grad_norm": 1.551687214387557,
+      "learning_rate": 9.99915308310258e-06,
+      "loss": 0.1723,
+      "step": 239
+    },
+    {
+      "epoch": 0.03589053387169134,
+      "grad_norm": 1.9544776771870587,
+      "learning_rate": 9.999107923364436e-06,
+      "loss": 0.2587,
+      "step": 240
+    },
+    {
+      "epoch": 0.03604007776282339,
+      "grad_norm": 2.1986380601508375,
+      "learning_rate": 9.999061590787394e-06,
+      "loss": 0.544,
+      "step": 241
+    },
+    {
+      "epoch": 0.03618962165395544,
+      "grad_norm": 2.5816888510040457,
+      "learning_rate": 9.999014085382326e-06,
+      "loss": 0.4619,
+      "step": 242
+    },
+    {
+      "epoch": 0.03633916554508748,
+      "grad_norm": 1.8291845348661409,
+      "learning_rate": 9.998965407160377e-06,
+      "loss": 0.2052,
+      "step": 243
+    },
+    {
+      "epoch": 0.036488709436219534,
+      "grad_norm": 3.167062575704647,
+      "learning_rate": 9.998915556132966e-06,
+      "loss": 0.6123,
+      "step": 244
+    },
+    {
+      "epoch": 0.03663825332735158,
+      "grad_norm": 1.8628898225455814,
+      "learning_rate": 9.99886453231179e-06,
+      "loss": 0.3634,
+      "step": 245
+    },
+    {
+      "epoch": 0.036787797218483624,
+      "grad_norm": 1.7903762911789451,
+      "learning_rate": 9.998812335708818e-06,
+      "loss": 0.2162,
+      "step": 246
+    },
+    {
+      "epoch": 0.03693734110961567,
+      "grad_norm": 1.3282642487848175,
+      "learning_rate": 9.998758966336296e-06,
+      "loss": 0.1875,
+      "step": 247
+    },
+    {
+      "epoch": 0.03708688500074772,
+      "grad_norm": 1.8364953512469955,
+      "learning_rate": 9.998704424206747e-06,
+      "loss": 0.208,
+      "step": 248
+    },
+    {
+      "epoch": 0.037236428891879766,
+      "grad_norm": 1.3941303606582691,
+      "learning_rate": 9.998648709332965e-06,
+      "loss": 0.1737,
+      "step": 249
+    },
+    {
+      "epoch": 0.03738597278301181,
+      "grad_norm": 1.7239196409011197,
+      "learning_rate": 9.998591821728022e-06,
+      "loss": 0.2339,
+      "step": 250
+    },
+    {
+      "epoch": 0.037535516674143864,
+      "grad_norm": 2.623262386600702,
+      "learning_rate": 9.998533761405265e-06,
+      "loss": 0.3988,
+      "step": 251
+    },
+    {
+      "epoch": 0.03768506056527591,
+      "grad_norm": 3.0417113736320354,
+      "learning_rate": 9.998474528378315e-06,
+      "loss": 0.3998,
+      "step": 252
+    },
+    {
+      "epoch": 0.037834604456407954,
+      "grad_norm": 2.3389769972346532,
+      "learning_rate": 9.998414122661066e-06,
+      "loss": 0.2157,
+      "step": 253
+    },
+    {
+      "epoch": 0.037984148347540006,
+      "grad_norm": 2.776666496961099,
+      "learning_rate": 9.998352544267696e-06,
+      "loss": 0.5598,
+      "step": 254
+    },
+    {
+      "epoch": 0.03813369223867205,
+      "grad_norm": 2.1472401976055746,
+      "learning_rate": 9.998289793212645e-06,
+      "loss": 0.2375,
+      "step": 255
+    },
+    {
+      "epoch": 0.038283236129804096,
+      "grad_norm": 2.258529852719024,
+      "learning_rate": 9.99822586951064e-06,
+      "loss": 0.257,
+      "step": 256
+    },
+    {
+      "epoch": 0.03843278002093614,
+      "grad_norm": 2.234662282588329,
+      "learning_rate": 9.998160773176676e-06,
+      "loss": 0.2513,
+      "step": 257
+    },
+    {
+      "epoch": 0.038582323912068194,
+      "grad_norm": 1.557075634748184,
+      "learning_rate": 9.998094504226025e-06,
+      "loss": 0.2154,
+      "step": 258
+    },
+    {
+      "epoch": 0.03873186780320024,
+      "grad_norm": 1.2782097805836874,
+      "learning_rate": 9.998027062674236e-06,
+      "loss": 0.1997,
+      "step": 259
+    },
+    {
+      "epoch": 0.038881411694332284,
+      "grad_norm": 1.5754692941437902,
+      "learning_rate": 9.997958448537129e-06,
+      "loss": 0.2271,
+      "step": 260
+    },
+    {
+      "epoch": 0.039030955585464336,
+      "grad_norm": 2.3273358127526516,
+      "learning_rate": 9.997888661830803e-06,
+      "loss": 0.4129,
+      "step": 261
+    },
+    {
+      "epoch": 0.03918049947659638,
+      "grad_norm": 2.5932478274973705,
+      "learning_rate": 9.997817702571631e-06,
+      "loss": 0.2762,
+      "step": 262
+    },
+    {
+      "epoch": 0.039330043367728426,
+      "grad_norm": 1.7415819067090217,
+      "learning_rate": 9.99774557077626e-06,
+      "loss": 0.2677,
+      "step": 263
+    },
+    {
+      "epoch": 0.03947958725886048,
+      "grad_norm": 2.1983315861883974,
+      "learning_rate": 9.997672266461613e-06,
+      "loss": 0.3412,
+      "step": 264
+    },
+    {
+      "epoch": 0.03962913114999252,
+      "grad_norm": 2.8445138272257666,
+      "learning_rate": 9.997597789644889e-06,
+      "loss": 0.3471,
+      "step": 265
+    },
+    {
+      "epoch": 0.03977867504112457,
+      "grad_norm": 2.6658347323464575,
+      "learning_rate": 9.997522140343558e-06,
+      "loss": 0.3785,
+      "step": 266
+    },
+    {
+      "epoch": 0.03992821893225662,
+      "grad_norm": 1.2913669477506569,
+      "learning_rate": 9.997445318575371e-06,
+      "loss": 0.2089,
+      "step": 267
+    },
+    {
+      "epoch": 0.040077762823388666,
+      "grad_norm": 2.440102551085522,
+      "learning_rate": 9.99736732435835e-06,
+      "loss": 0.5639,
+      "step": 268
+    },
+    {
+      "epoch": 0.04022730671452071,
+      "grad_norm": 2.252623935384866,
+      "learning_rate": 9.997288157710795e-06,
+      "loss": 0.447,
+      "step": 269
+    },
+    {
+      "epoch": 0.040376850605652756,
+      "grad_norm": 1.9038309319538977,
+      "learning_rate": 9.997207818651273e-06,
+      "loss": 0.2784,
+      "step": 270
+    },
+    {
+      "epoch": 0.04052639449678481,
+      "grad_norm": 2.05316637395224,
+      "learning_rate": 9.99712630719864e-06,
+      "loss": 0.3874,
+      "step": 271
+    },
+    {
+      "epoch": 0.04067593838791685,
+      "grad_norm": 4.663034399257074,
+      "learning_rate": 9.997043623372016e-06,
+      "loss": 0.3558,
+      "step": 272
+    },
+    {
+      "epoch": 0.0408254822790489,
+      "grad_norm": 2.0324793909935375,
+      "learning_rate": 9.996959767190799e-06,
+      "loss": 0.3884,
+      "step": 273
+    },
+    {
+      "epoch": 0.04097502617018095,
+      "grad_norm": 2.1897027573531003,
+      "learning_rate": 9.996874738674663e-06,
+      "loss": 0.2372,
+      "step": 274
+    },
+    {
+      "epoch": 0.041124570061312996,
+      "grad_norm": 1.9410471939157525,
+      "learning_rate": 9.996788537843558e-06,
+      "loss": 0.3478,
+      "step": 275
+    },
+    {
+      "epoch": 0.04127411395244504,
+      "grad_norm": 3.650983914269082,
+      "learning_rate": 9.996701164717704e-06,
+      "loss": 0.4213,
+      "step": 276
+    },
+    {
+      "epoch": 0.04142365784357709,
+      "grad_norm": 3.067988013237884,
+      "learning_rate": 9.996612619317602e-06,
+      "loss": 0.7209,
+      "step": 277
+    },
+    {
+      "epoch": 0.04157320173470914,
+      "grad_norm": 2.5863303551652033,
+      "learning_rate": 9.996522901664028e-06,
+      "loss": 0.5418,
+      "step": 278
+    },
+    {
+      "epoch": 0.04172274562584118,
+      "grad_norm": 2.1885641779249476,
+      "learning_rate": 9.996432011778026e-06,
+      "loss": 0.371,
+      "step": 279
+    },
+    {
+      "epoch": 0.041872289516973235,
+      "grad_norm": 2.398824728854803,
+      "learning_rate": 9.99633994968092e-06,
+      "loss": 0.5508,
+      "step": 280
+    },
+    {
+      "epoch": 0.04202183340810528,
+      "grad_norm": 1.5732032420608302,
+      "learning_rate": 9.996246715394314e-06,
+      "loss": 0.2468,
+      "step": 281
+    },
+    {
+      "epoch": 0.042171377299237325,
+      "grad_norm": 2.8532279807617944,
+      "learning_rate": 9.996152308940075e-06,
+      "loss": 0.5503,
+      "step": 282
+    },
+    {
+      "epoch": 0.04232092119036937,
+      "grad_norm": 2.4502727303222733,
+      "learning_rate": 9.996056730340356e-06,
+      "loss": 0.4046,
+      "step": 283
+    },
+    {
+      "epoch": 0.04247046508150142,
+      "grad_norm": 1.9272098426705169,
+      "learning_rate": 9.995959979617578e-06,
+      "loss": 0.3906,
+      "step": 284
+    },
+    {
+      "epoch": 0.04262000897263347,
+      "grad_norm": 2.290690335549339,
+      "learning_rate": 9.995862056794441e-06,
+      "loss": 0.2464,
+      "step": 285
+    },
+    {
+      "epoch": 0.04276955286376551,
+      "grad_norm": 1.656564250859485,
+      "learning_rate": 9.99576296189392e-06,
+      "loss": 0.1996,
+      "step": 286
+    },
+    {
+      "epoch": 0.042919096754897565,
+      "grad_norm": 2.1259148220336965,
+      "learning_rate": 9.995662694939262e-06,
+      "loss": 0.3994,
+      "step": 287
+    },
+    {
+      "epoch": 0.04306864064602961,
+      "grad_norm": 2.286901143642134,
+      "learning_rate": 9.99556125595399e-06,
+      "loss": 0.4047,
+      "step": 288
+    },
+    {
+      "epoch": 0.043218184537161655,
+      "grad_norm": 1.3559455912309712,
+      "learning_rate": 9.995458644961902e-06,
+      "loss": 0.2228,
+      "step": 289
+    },
+    {
+      "epoch": 0.04336772842829371,
+      "grad_norm": 2.285750924681825,
+      "learning_rate": 9.995354861987075e-06,
+      "loss": 0.2367,
+      "step": 290
+    },
+    {
+      "epoch": 0.04351727231942575,
+      "grad_norm": 1.923824453592428,
+      "learning_rate": 9.995249907053854e-06,
+      "loss": 0.3951,
+      "step": 291
+    },
+    {
+      "epoch": 0.0436668162105578,
+      "grad_norm": 1.968047953500074,
+      "learning_rate": 9.995143780186865e-06,
+      "loss": 0.2149,
+      "step": 292
+    },
+    {
+      "epoch": 0.04381636010168984,
+      "grad_norm": 2.3975790519132074,
+      "learning_rate": 9.995036481411005e-06,
+      "loss": 0.5312,
+      "step": 293
+    },
+    {
+      "epoch": 0.043965903992821895,
+      "grad_norm": 1.9664546058841197,
+      "learning_rate": 9.994928010751447e-06,
+      "loss": 0.4832,
+      "step": 294
+    },
+    {
+      "epoch": 0.04411544788395394,
+      "grad_norm": 2.1609011533249785,
+      "learning_rate": 9.994818368233639e-06,
+      "loss": 0.571,
+      "step": 295
+    },
+    {
+      "epoch": 0.044264991775085985,
+      "grad_norm": 1.2099666806993736,
+      "learning_rate": 9.994707553883305e-06,
+      "loss": 0.1801,
+      "step": 296
+    },
+    {
+      "epoch": 0.04441453566621804,
+      "grad_norm": 1.8811137964659612,
+      "learning_rate": 9.994595567726444e-06,
+      "loss": 0.2708,
+      "step": 297
+    },
+    {
+      "epoch": 0.04456407955735008,
+      "grad_norm": 1.6387011737954997,
+      "learning_rate": 9.994482409789329e-06,
+      "loss": 0.245,
+      "step": 298
+    },
+    {
+      "epoch": 0.04471362344848213,
+      "grad_norm": 2.4061797367092486,
+      "learning_rate": 9.994368080098505e-06,
+      "loss": 0.204,
+      "step": 299
+    },
+    {
+      "epoch": 0.04486316733961418,
+      "grad_norm": 2.555264958903577,
+      "learning_rate": 9.994252578680796e-06,
+      "loss": 0.5251,
+      "step": 300
+    },
+    {
+      "epoch": 0.045012711230746225,
+      "grad_norm": 3.1965886018503897,
+      "learning_rate": 9.994135905563302e-06,
+      "loss": 0.4353,
+      "step": 301
+    },
+    {
+      "epoch": 0.04516225512187827,
+      "grad_norm": 2.390530599961774,
+      "learning_rate": 9.994018060773396e-06,
+      "loss": 0.4199,
+      "step": 302
+    },
+    {
+      "epoch": 0.04531179901301032,
+      "grad_norm": 2.694731420269419,
+      "learning_rate": 9.993899044338722e-06,
+      "loss": 0.4029,
+      "step": 303
+    },
+    {
+      "epoch": 0.04546134290414237,
+      "grad_norm": 2.5518583518075437,
+      "learning_rate": 9.993778856287205e-06,
+      "loss": 0.3712,
+      "step": 304
+    },
+    {
+      "epoch": 0.04561088679527441,
+      "grad_norm": 1.958382495979976,
+      "learning_rate": 9.99365749664704e-06,
+      "loss": 0.3617,
+      "step": 305
+    },
+    {
+      "epoch": 0.04576043068640646,
+      "grad_norm": 2.299652220902115,
+      "learning_rate": 9.993534965446701e-06,
+      "loss": 0.4059,
+      "step": 306
+    },
+    {
+      "epoch": 0.04590997457753851,
+      "grad_norm": 4.086258301258261,
+      "learning_rate": 9.993411262714934e-06,
+      "loss": 0.2774,
+      "step": 307
+    },
+    {
+      "epoch": 0.046059518468670554,
+      "grad_norm": 2.0081624141767156,
+      "learning_rate": 9.993286388480763e-06,
+      "loss": 0.2724,
+      "step": 308
+    },
+    {
+      "epoch": 0.0462090623598026,
+      "grad_norm": 2.388037596587926,
+      "learning_rate": 9.993160342773483e-06,
+      "loss": 0.2706,
+      "step": 309
+    },
+    {
+      "epoch": 0.04635860625093465,
+      "grad_norm": 1.5868739255084185,
+      "learning_rate": 9.993033125622665e-06,
+      "loss": 0.256,
+      "step": 310
+    },
+    {
+      "epoch": 0.0465081501420667,
+      "grad_norm": 1.8286822342955051,
+      "learning_rate": 9.992904737058157e-06,
+      "loss": 0.209,
+      "step": 311
+    },
+    {
+      "epoch": 0.04665769403319874,
+      "grad_norm": 2.2060332987484306,
+      "learning_rate": 9.992775177110078e-06,
+      "loss": 0.4253,
+      "step": 312
+    },
+    {
+      "epoch": 0.046807237924330794,
+      "grad_norm": 1.39628419375001,
+      "learning_rate": 9.992644445808826e-06,
+      "loss": 0.1693,
+      "step": 313
+    },
+    {
+      "epoch": 0.04695678181546284,
+      "grad_norm": 1.5668060198088787,
+      "learning_rate": 9.99251254318507e-06,
+      "loss": 0.24,
+      "step": 314
+    },
+    {
+      "epoch": 0.047106325706594884,
+      "grad_norm": 1.998270389587923,
+      "learning_rate": 9.992379469269758e-06,
+      "loss": 0.2519,
+      "step": 315
+    },
+    {
+      "epoch": 0.04725586959772693,
+      "grad_norm": 1.9609810436779118,
+      "learning_rate": 9.99224522409411e-06,
+      "loss": 0.2023,
+      "step": 316
+    },
+    {
+      "epoch": 0.04740541348885898,
+      "grad_norm": 1.4580736241239847,
+      "learning_rate": 9.992109807689619e-06,
+      "loss": 0.2387,
+      "step": 317
+    },
+    {
+      "epoch": 0.04755495737999103,
+      "grad_norm": 2.710681694340303,
+      "learning_rate": 9.991973220088057e-06,
+      "loss": 0.6738,
+      "step": 318
+    },
+    {
+      "epoch": 0.04770450127112307,
+      "grad_norm": 1.2469776099691643,
+      "learning_rate": 9.991835461321466e-06,
+      "loss": 0.2013,
+      "step": 319
+    },
+    {
+      "epoch": 0.047854045162255124,
+      "grad_norm": 2.128896128779159,
+      "learning_rate": 9.99169653142217e-06,
+      "loss": 0.3432,
+      "step": 320
+    },
+    {
+      "epoch": 0.04800358905338717,
+      "grad_norm": 1.6053097848087672,
+      "learning_rate": 9.991556430422759e-06,
+      "loss": 0.2301,
+      "step": 321
+    },
+    {
+      "epoch": 0.048153132944519214,
+      "grad_norm": 1.7774787600035602,
+      "learning_rate": 9.991415158356106e-06,
+      "loss": 0.2535,
+      "step": 322
+    },
+    {
+      "epoch": 0.048302676835651266,
+      "grad_norm": 1.449815289318445,
+      "learning_rate": 9.991272715255351e-06,
+      "loss": 0.1878,
+      "step": 323
+    },
+    {
+      "epoch": 0.04845222072678331,
+      "grad_norm": 1.5118547669168991,
+      "learning_rate": 9.991129101153916e-06,
+      "loss": 0.3186,
+      "step": 324
+    },
+    {
+      "epoch": 0.048601764617915356,
+      "grad_norm": 1.461388444407636,
+      "learning_rate": 9.99098431608549e-06,
+      "loss": 0.1747,
+      "step": 325
+    },
+    {
+      "epoch": 0.04875130850904741,
+      "grad_norm": 2.3912366570769974,
+      "learning_rate": 9.990838360084045e-06,
+      "loss": 0.5325,
+      "step": 326
+    },
+    {
+      "epoch": 0.048900852400179454,
+      "grad_norm": 2.5611474084390937,
+      "learning_rate": 9.990691233183823e-06,
+      "loss": 0.2606,
+      "step": 327
+    },
+    {
+      "epoch": 0.0490503962913115,
+      "grad_norm": 2.21899436894442,
+      "learning_rate": 9.990542935419341e-06,
+      "loss": 0.4253,
+      "step": 328
+    },
+    {
+      "epoch": 0.049199940182443544,
+      "grad_norm": 1.6883179263006298,
+      "learning_rate": 9.99039346682539e-06,
+      "loss": 0.1768,
+      "step": 329
+    },
+    {
+      "epoch": 0.049349484073575596,
+      "grad_norm": 3.2358870266119006,
+      "learning_rate": 9.990242827437036e-06,
+      "loss": 0.7866,
+      "step": 330
+    },
+    {
+      "epoch": 0.04949902796470764,
+      "grad_norm": 2.0627143054944153,
+      "learning_rate": 9.990091017289623e-06,
+      "loss": 0.3286,
+      "step": 331
+    },
+    {
+      "epoch": 0.049648571855839686,
+      "grad_norm": 2.1246533005850523,
+      "learning_rate": 9.989938036418766e-06,
+      "loss": 0.2716,
+      "step": 332
+    },
+    {
+      "epoch": 0.04979811574697174,
+      "grad_norm": 2.6250279686209828,
+      "learning_rate": 9.989783884860355e-06,
+      "loss": 0.5058,
+      "step": 333
+    },
+    {
+      "epoch": 0.049947659638103784,
+      "grad_norm": 2.3409062617647627,
+      "learning_rate": 9.989628562650558e-06,
+      "loss": 0.2589,
+      "step": 334
+    },
+    {
+      "epoch": 0.05009720352923583,
+      "grad_norm": 1.835901073337933,
+      "learning_rate": 9.989472069825811e-06,
+      "loss": 0.3493,
+      "step": 335
+    },
+    {
+      "epoch": 0.05024674742036788,
+      "grad_norm": 2.2454393810241298,
+      "learning_rate": 9.989314406422835e-06,
+      "loss": 0.4113,
+      "step": 336
+    },
+    {
+      "epoch": 0.050396291311499926,
+      "grad_norm": 2.2906853778474674,
+      "learning_rate": 9.989155572478611e-06,
+      "loss": 0.5289,
+      "step": 337
+    },
+    {
+      "epoch": 0.05054583520263197,
+      "grad_norm": 2.3899442476389665,
+      "learning_rate": 9.98899556803041e-06,
+      "loss": 0.2174,
+      "step": 338
+    },
+    {
+      "epoch": 0.05069537909376402,
+      "grad_norm": 1.3681982854338133,
+      "learning_rate": 9.988834393115768e-06,
+      "loss": 0.2021,
+      "step": 339
+    },
+    {
+      "epoch": 0.05084492298489607,
+      "grad_norm": 1.5118760155287632,
+      "learning_rate": 9.988672047772497e-06,
+      "loss": 0.1927,
+      "step": 340
+    },
+    {
+      "epoch": 0.05099446687602811,
+      "grad_norm": 2.1144895431001105,
+      "learning_rate": 9.988508532038685e-06,
+      "loss": 0.3325,
+      "step": 341
+    },
+    {
+      "epoch": 0.05114401076716016,
+      "grad_norm": 1.8616803287346595,
+      "learning_rate": 9.988343845952697e-06,
+      "loss": 0.3018,
+      "step": 342
+    },
+    {
+      "epoch": 0.05129355465829221,
+      "grad_norm": 2.787967616575242,
+      "learning_rate": 9.988177989553167e-06,
+      "loss": 0.4641,
+      "step": 343
+    },
+    {
+      "epoch": 0.051443098549424256,
+      "grad_norm": 2.2905797584406242,
+      "learning_rate": 9.98801096287901e-06,
+      "loss": 0.5336,
+      "step": 344
+    },
+    {
+      "epoch": 0.0515926424405563,
+      "grad_norm": 1.769311364935245,
+      "learning_rate": 9.987842765969408e-06,
+      "loss": 0.2843,
+      "step": 345
+    },
+    {
+      "epoch": 0.05174218633168835,
+      "grad_norm": 1.7122732613639495,
+      "learning_rate": 9.987673398863824e-06,
+      "loss": 0.2272,
+      "step": 346
+    },
+    {
+      "epoch": 0.0518917302228204,
+      "grad_norm": 2.328359950454365,
+      "learning_rate": 9.987502861601991e-06,
+      "loss": 0.2645,
+      "step": 347
+    },
+    {
+      "epoch": 0.05204127411395244,
+      "grad_norm": 2.208277642399548,
+      "learning_rate": 9.987331154223922e-06,
+      "loss": 0.5877,
+      "step": 348
+    },
+    {
+      "epoch": 0.052190818005084495,
+      "grad_norm": 2.154817789687723,
+      "learning_rate": 9.9871582767699e-06,
+      "loss": 0.3414,
+      "step": 349
+    },
+    {
+      "epoch": 0.05234036189621654,
+      "grad_norm": 2.0510314098551814,
+      "learning_rate": 9.986984229280483e-06,
+      "loss": 0.3981,
+      "step": 350
+    },
+    {
+      "epoch": 0.052489905787348586,
+      "grad_norm": 2.346735661125246,
+      "learning_rate": 9.986809011796503e-06,
+      "loss": 0.6596,
+      "step": 351
+    },
+    {
+      "epoch": 0.05263944967848063,
+      "grad_norm": 1.641693244293744,
+      "learning_rate": 9.98663262435907e-06,
+      "loss": 0.3657,
+      "step": 352
+    },
+    {
+      "epoch": 0.05278899356961268,
+      "grad_norm": 2.240226359797858,
+      "learning_rate": 9.986455067009566e-06,
+      "loss": 0.3706,
+      "step": 353
+    },
+    {
+      "epoch": 0.05293853746074473,
+      "grad_norm": 2.3791485993411357,
+      "learning_rate": 9.986276339789648e-06,
+      "loss": 0.5428,
+      "step": 354
+    },
+    {
+      "epoch": 0.05308808135187677,
+      "grad_norm": 1.7806897327965683,
+      "learning_rate": 9.986096442741241e-06,
+      "loss": 0.2336,
+      "step": 355
+    },
+    {
+      "epoch": 0.053237625243008825,
+      "grad_norm": 1.8563417208131827,
+      "learning_rate": 9.98591537590656e-06,
+      "loss": 0.2129,
+      "step": 356
+    },
+    {
+      "epoch": 0.05338716913414087,
+      "grad_norm": 2.2115041121315895,
+      "learning_rate": 9.98573313932808e-06,
+      "loss": 0.5232,
+      "step": 357
+    },
+    {
+      "epoch": 0.053536713025272915,
+      "grad_norm": 1.3693709893910027,
+      "learning_rate": 9.985549733048556e-06,
+      "loss": 0.3524,
+      "step": 358
+    },
+    {
+      "epoch": 0.05368625691640497,
+      "grad_norm": 2.033727598383455,
+      "learning_rate": 9.985365157111017e-06,
+      "loss": 0.3987,
+      "step": 359
+    },
+    {
+      "epoch": 0.05383580080753701,
+      "grad_norm": 2.3258255541409505,
+      "learning_rate": 9.985179411558767e-06,
+      "loss": 0.5489,
+      "step": 360
+    },
+    {
+      "epoch": 0.05398534469866906,
+      "grad_norm": 2.0805855861837057,
+      "learning_rate": 9.984992496435383e-06,
+      "loss": 0.3982,
+      "step": 361
+    },
+    {
+      "epoch": 0.05413488858980111,
+      "grad_norm": 1.4938394292792039,
+      "learning_rate": 9.984804411784717e-06,
+      "loss": 0.2279,
+      "step": 362
+    },
+    {
+      "epoch": 0.054284432480933155,
+      "grad_norm": 1.935765339737269,
+      "learning_rate": 9.984615157650896e-06,
+      "loss": 0.2208,
+      "step": 363
+    },
+    {
+      "epoch": 0.0544339763720652,
+      "grad_norm": 2.294825440673555,
+      "learning_rate": 9.98442473407832e-06,
+      "loss": 0.4006,
+      "step": 364
+    },
+    {
+      "epoch": 0.054583520263197245,
+      "grad_norm": 1.7404498428206792,
+      "learning_rate": 9.984233141111663e-06,
+      "loss": 0.3859,
+      "step": 365
+    },
+    {
+      "epoch": 0.0547330641543293,
+      "grad_norm": 2.382616866788976,
+      "learning_rate": 9.984040378795879e-06,
+      "loss": 0.5393,
+      "step": 366
+    },
+    {
+      "epoch": 0.05488260804546134,
+      "grad_norm": 2.121310368782044,
+      "learning_rate": 9.983846447176186e-06,
+      "loss": 0.3808,
+      "step": 367
+    },
+    {
+      "epoch": 0.05503215193659339,
+      "grad_norm": 1.4327836947551182,
+      "learning_rate": 9.983651346298089e-06,
+      "loss": 0.21,
+      "step": 368
+    },
+    {
+      "epoch": 0.05518169582772544,
+      "grad_norm": 1.8551217286702022,
+      "learning_rate": 9.983455076207353e-06,
+      "loss": 0.3611,
+      "step": 369
+    },
+    {
+      "epoch": 0.055331239718857485,
+      "grad_norm": 1.1962615317465979,
+      "learning_rate": 9.983257636950032e-06,
+      "loss": 0.1632,
+      "step": 370
+    },
+    {
+      "epoch": 0.05548078360998953,
+      "grad_norm": 2.210937603202386,
+      "learning_rate": 9.983059028572443e-06,
+      "loss": 0.2054,
+      "step": 371
+    },
+    {
+      "epoch": 0.05563032750112158,
+      "grad_norm": 1.3676870965949202,
+      "learning_rate": 9.982859251121183e-06,
+      "loss": 0.2257,
+      "step": 372
+    },
+    {
+      "epoch": 0.05577987139225363,
+      "grad_norm": 1.877238753038072,
+      "learning_rate": 9.98265830464312e-06,
+      "loss": 0.3069,
+      "step": 373
+    },
+    {
+      "epoch": 0.05592941528338567,
+      "grad_norm": 2.6215120058588743,
+      "learning_rate": 9.9824561891854e-06,
+      "loss": 0.3812,
+      "step": 374
+    },
+    {
+      "epoch": 0.056078959174517724,
+      "grad_norm": 1.5353869053774183,
+      "learning_rate": 9.982252904795437e-06,
+      "loss": 0.3038,
+      "step": 375
+    },
+    {
+      "epoch": 0.05622850306564977,
+      "grad_norm": 1.5387274188562523,
+      "learning_rate": 9.98204845152093e-06,
+      "loss": 0.1784,
+      "step": 376
+    },
+    {
+      "epoch": 0.056378046956781815,
+      "grad_norm": 2.3221296907492444,
+      "learning_rate": 9.981842829409842e-06,
+      "loss": 0.4253,
+      "step": 377
+    },
+    {
+      "epoch": 0.05652759084791386,
+      "grad_norm": 1.8464138105889263,
+      "learning_rate": 9.981636038510414e-06,
+      "loss": 0.2137,
+      "step": 378
+    },
+    {
+      "epoch": 0.05667713473904591,
+      "grad_norm": 1.9213502252741161,
+      "learning_rate": 9.98142807887116e-06,
+      "loss": 0.2652,
+      "step": 379
+    },
+    {
+      "epoch": 0.05682667863017796,
+      "grad_norm": 1.7697460473662174,
+      "learning_rate": 9.981218950540874e-06,
+      "loss": 0.2525,
+      "step": 380
+    },
+    {
+      "epoch": 0.05697622252131,
+      "grad_norm": 2.001502054151958,
+      "learning_rate": 9.981008653568613e-06,
+      "loss": 0.3749,
+      "step": 381
+    },
+    {
+      "epoch": 0.057125766412442054,
+      "grad_norm": 1.7507480997796745,
+      "learning_rate": 9.98079718800372e-06,
+      "loss": 0.3293,
+      "step": 382
+    },
+    {
+      "epoch": 0.0572753103035741,
+      "grad_norm": 1.8995856376763527,
+      "learning_rate": 9.980584553895805e-06,
+      "loss": 0.2595,
+      "step": 383
+    },
+    {
+      "epoch": 0.057424854194706144,
+      "grad_norm": 1.6960817341003291,
+      "learning_rate": 9.980370751294754e-06,
+      "loss": 0.3214,
+      "step": 384
+    },
+    {
+      "epoch": 0.057574398085838197,
+      "grad_norm": 2.747620756274178,
+      "learning_rate": 9.980155780250728e-06,
+      "loss": 0.4678,
+      "step": 385
+    },
+    {
+      "epoch": 0.05772394197697024,
+      "grad_norm": 1.429295181164985,
+      "learning_rate": 9.979939640814158e-06,
+      "loss": 0.3417,
+      "step": 386
+    },
+    {
+      "epoch": 0.05787348586810229,
+      "grad_norm": 1.546941524577904,
+      "learning_rate": 9.979722333035757e-06,
+      "loss": 0.3017,
+      "step": 387
+    },
+    {
+      "epoch": 0.05802302975923433,
+      "grad_norm": 2.3243262803022753,
+      "learning_rate": 9.979503856966504e-06,
+      "loss": 0.3906,
+      "step": 388
+    },
+    {
+      "epoch": 0.058172573650366384,
+      "grad_norm": 1.5367077444523152,
+      "learning_rate": 9.979284212657658e-06,
+      "loss": 0.2735,
+      "step": 389
+    },
+    {
+      "epoch": 0.05832211754149843,
+      "grad_norm": 1.0259751361449947,
+      "learning_rate": 9.979063400160747e-06,
+      "loss": 0.1788,
+      "step": 390
+    },
+    {
+      "epoch": 0.058471661432630474,
+      "grad_norm": 1.7811616961442123,
+      "learning_rate": 9.97884141952758e-06,
+      "loss": 0.2071,
+      "step": 391
+    },
+    {
+      "epoch": 0.058621205323762526,
+      "grad_norm": 2.347009922116326,
+      "learning_rate": 9.978618270810229e-06,
+      "loss": 0.4248,
+      "step": 392
+    },
+    {
+      "epoch": 0.05877074921489457,
+      "grad_norm": 1.3076474084417338,
+      "learning_rate": 9.978393954061052e-06,
+      "loss": 0.1771,
+      "step": 393
+    },
+    {
+      "epoch": 0.05892029310602662,
+      "grad_norm": 2.4165379692755455,
+      "learning_rate": 9.978168469332677e-06,
+      "loss": 0.4913,
+      "step": 394
+    },
+    {
+      "epoch": 0.05906983699715867,
+      "grad_norm": 1.6584516839965744,
+      "learning_rate": 9.977941816678e-06,
+      "loss": 0.2292,
+      "step": 395
+    },
+    {
+      "epoch": 0.059219380888290714,
+      "grad_norm": 1.3323879687206615,
+      "learning_rate": 9.9777139961502e-06,
+      "loss": 0.2042,
+      "step": 396
+    },
+    {
+      "epoch": 0.05936892477942276,
+      "grad_norm": 1.242996863833067,
+      "learning_rate": 9.977485007802725e-06,
+      "loss": 0.1759,
+      "step": 397
+    },
+    {
+      "epoch": 0.05951846867055481,
+      "grad_norm": 2.0289613301318057,
+      "learning_rate": 9.977254851689297e-06,
+      "loss": 0.3391,
+      "step": 398
+    },
+    {
+      "epoch": 0.059668012561686856,
+      "grad_norm": 1.7111890076718022,
+      "learning_rate": 9.977023527863913e-06,
+      "loss": 0.318,
+      "step": 399
+    },
+    {
+      "epoch": 0.0598175564528189,
+      "grad_norm": 2.360289838407607,
+      "learning_rate": 9.976791036380844e-06,
+      "loss": 0.7436,
+      "step": 400
+    },
+    {
+      "epoch": 0.059967100343950946,
+      "grad_norm": 1.6556682149662436,
+      "learning_rate": 9.976557377294634e-06,
+      "loss": 0.3579,
+      "step": 401
+    },
+    {
+      "epoch": 0.060116644235083,
+      "grad_norm": 1.9472299876725607,
+      "learning_rate": 9.976322550660103e-06,
+      "loss": 0.3939,
+      "step": 402
+    },
+    {
+      "epoch": 0.060266188126215044,
+      "grad_norm": 1.2625006623785717,
+      "learning_rate": 9.976086556532343e-06,
+      "loss": 0.1777,
+      "step": 403
+    },
+    {
+      "epoch": 0.06041573201734709,
+      "grad_norm": 2.142440158571368,
+      "learning_rate": 9.975849394966721e-06,
+      "loss": 0.4728,
+      "step": 404
+    },
+    {
+      "epoch": 0.06056527590847914,
+      "grad_norm": 1.3109446375337697,
+      "learning_rate": 9.975611066018876e-06,
+      "loss": 0.2035,
+      "step": 405
+    },
+    {
+      "epoch": 0.060714819799611186,
+      "grad_norm": 1.473069250695052,
+      "learning_rate": 9.975371569744723e-06,
+      "loss": 0.2502,
+      "step": 406
+    },
+    {
+      "epoch": 0.06086436369074323,
+      "grad_norm": 1.4147256960977963,
+      "learning_rate": 9.975130906200453e-06,
+      "loss": 0.1861,
+      "step": 407
+    },
+    {
+      "epoch": 0.06101390758187528,
+      "grad_norm": 1.5107559691714745,
+      "learning_rate": 9.97488907544252e-06,
+      "loss": 0.2309,
+      "step": 408
+    },
+    {
+      "epoch": 0.06116345147300733,
+      "grad_norm": 1.5467720756101462,
+      "learning_rate": 9.97464607752767e-06,
+      "loss": 0.235,
+      "step": 409
+    },
+    {
+      "epoch": 0.061312995364139374,
+      "grad_norm": 1.2901444374034334,
+      "learning_rate": 9.974401912512905e-06,
+      "loss": 0.1877,
+      "step": 410
+    },
+    {
+      "epoch": 0.061462539255271426,
+      "grad_norm": 1.8751659558285558,
+      "learning_rate": 9.974156580455512e-06,
+      "loss": 0.2941,
+      "step": 411
+    },
+    {
+      "epoch": 0.06161208314640347,
+      "grad_norm": 1.2187366523072891,
+      "learning_rate": 9.973910081413048e-06,
+      "loss": 0.2,
+      "step": 412
+    },
+    {
+      "epoch": 0.061761627037535516,
+      "grad_norm": 2.56665763030278,
+      "learning_rate": 9.973662415443342e-06,
+      "loss": 0.4259,
+      "step": 413
+    },
+    {
+      "epoch": 0.06191117092866756,
+      "grad_norm": 1.5201509236946156,
+      "learning_rate": 9.973413582604502e-06,
+      "loss": 0.2098,
+      "step": 414
+    },
+    {
+      "epoch": 0.06206071481979961,
+      "grad_norm": 2.2299268067487183,
+      "learning_rate": 9.973163582954903e-06,
+      "loss": 0.5054,
+      "step": 415
+    },
+    {
+      "epoch": 0.06221025871093166,
+      "grad_norm": 2.195400724979985,
+      "learning_rate": 9.972912416553202e-06,
+      "loss": 0.3856,
+      "step": 416
+    },
+    {
+      "epoch": 0.0623598026020637,
+      "grad_norm": 2.3196273331545876,
+      "learning_rate": 9.972660083458321e-06,
+      "loss": 0.5608,
+      "step": 417
+    },
+    {
+      "epoch": 0.06250934649319576,
+      "grad_norm": 1.6815269422927719,
+      "learning_rate": 9.97240658372946e-06,
+      "loss": 0.3682,
+      "step": 418
+    },
+    {
+      "epoch": 0.0626588903843278,
+      "grad_norm": 1.7582779956751238,
+      "learning_rate": 9.972151917426095e-06,
+      "loss": 0.2256,
+      "step": 419
+    },
+    {
+      "epoch": 0.06280843427545985,
+      "grad_norm": 1.9523974169697056,
+      "learning_rate": 9.97189608460797e-06,
+      "loss": 0.2303,
+      "step": 420
+    },
+    {
+      "epoch": 0.06295797816659189,
+      "grad_norm": 2.120409254412015,
+      "learning_rate": 9.97163908533511e-06,
+      "loss": 0.2198,
+      "step": 421
+    },
+    {
+      "epoch": 0.06310752205772394,
+      "grad_norm": 1.7213130956608376,
+      "learning_rate": 9.971380919667806e-06,
+      "loss": 0.3355,
+      "step": 422
+    },
+    {
+      "epoch": 0.063257065948856,
+      "grad_norm": 1.6609701125154137,
+      "learning_rate": 9.971121587666627e-06,
+      "loss": 0.2354,
+      "step": 423
+    },
+    {
+      "epoch": 0.06340660983998804,
+      "grad_norm": 1.2809919353271448,
+      "learning_rate": 9.970861089392415e-06,
+      "loss": 0.2043,
+      "step": 424
+    },
+    {
+      "epoch": 0.06355615373112009,
+      "grad_norm": 1.137987748410028,
+      "learning_rate": 9.970599424906285e-06,
+      "loss": 0.1714,
+      "step": 425
+    },
+    {
+      "epoch": 0.06370569762225213,
+      "grad_norm": 2.241505455994119,
+      "learning_rate": 9.970336594269627e-06,
+      "loss": 0.559,
+      "step": 426
+    },
+    {
+      "epoch": 0.06385524151338418,
+      "grad_norm": 1.8145782296174282,
+      "learning_rate": 9.970072597544102e-06,
+      "loss": 0.4695,
+      "step": 427
+    },
+    {
+      "epoch": 0.06400478540451622,
+      "grad_norm": 2.6609160560733924,
+      "learning_rate": 9.96980743479165e-06,
+      "loss": 0.3927,
+      "step": 428
+    },
+    {
+      "epoch": 0.06415432929564828,
+      "grad_norm": 1.5902127205656447,
+      "learning_rate": 9.969541106074477e-06,
+      "loss": 0.3221,
+      "step": 429
+    },
+    {
+      "epoch": 0.06430387318678032,
+      "grad_norm": 1.354440824254012,
+      "learning_rate": 9.969273611455066e-06,
+      "loss": 0.1982,
+      "step": 430
+    },
+    {
+      "epoch": 0.06445341707791237,
+      "grad_norm": 2.1796464676908682,
+      "learning_rate": 9.969004950996175e-06,
+      "loss": 0.5947,
+      "step": 431
+    },
+    {
+      "epoch": 0.06460296096904442,
+      "grad_norm": 1.6772295444343943,
+      "learning_rate": 9.968735124760834e-06,
+      "loss": 0.3567,
+      "step": 432
+    },
+    {
+      "epoch": 0.06475250486017646,
+      "grad_norm": 2.326608368656497,
+      "learning_rate": 9.968464132812348e-06,
+      "loss": 0.3934,
+      "step": 433
+    },
+    {
+      "epoch": 0.0649020487513085,
+      "grad_norm": 1.9737750855760885,
+      "learning_rate": 9.968191975214293e-06,
+      "loss": 0.3936,
+      "step": 434
+    },
+    {
+      "epoch": 0.06505159264244055,
+      "grad_norm": 2.09687169461338,
+      "learning_rate": 9.967918652030522e-06,
+      "loss": 0.3644,
+      "step": 435
+    },
+    {
+      "epoch": 0.06520113653357261,
+      "grad_norm": 2.1122151786614967,
+      "learning_rate": 9.967644163325157e-06,
+      "loss": 0.2169,
+      "step": 436
+    },
+    {
+      "epoch": 0.06535068042470465,
+      "grad_norm": 1.8368706867911107,
+      "learning_rate": 9.967368509162595e-06,
+      "loss": 0.3956,
+      "step": 437
+    },
+    {
+      "epoch": 0.0655002243158367,
+      "grad_norm": 1.7823169737575542,
+      "learning_rate": 9.96709168960751e-06,
+      "loss": 0.232,
+      "step": 438
+    },
+    {
+      "epoch": 0.06564976820696874,
+      "grad_norm": 2.1565508943507194,
+      "learning_rate": 9.966813704724844e-06,
+      "loss": 0.2228,
+      "step": 439
+    },
+    {
+      "epoch": 0.06579931209810079,
+      "grad_norm": 2.2075342060994414,
+      "learning_rate": 9.966534554579816e-06,
+      "loss": 0.204,
+      "step": 440
+    },
+    {
+      "epoch": 0.06594885598923284,
+      "grad_norm": 2.0929887441012602,
+      "learning_rate": 9.966254239237917e-06,
+      "loss": 0.3946,
+      "step": 441
+    },
+    {
+      "epoch": 0.06609839988036488,
+      "grad_norm": 2.0382287962872834,
+      "learning_rate": 9.965972758764912e-06,
+      "loss": 0.4633,
+      "step": 442
+    },
+    {
+      "epoch": 0.06624794377149694,
+      "grad_norm": 1.2772439274586147,
+      "learning_rate": 9.96569011322684e-06,
+      "loss": 0.1784,
+      "step": 443
+    },
+    {
+      "epoch": 0.06639748766262898,
+      "grad_norm": 1.1024457344648066,
+      "learning_rate": 9.965406302690011e-06,
+      "loss": 0.1625,
+      "step": 444
+    },
+    {
+      "epoch": 0.06654703155376103,
+      "grad_norm": 1.2184559623271476,
+      "learning_rate": 9.965121327221007e-06,
+      "loss": 0.1959,
+      "step": 445
+    },
+    {
+      "epoch": 0.06669657544489307,
+      "grad_norm": 1.9215235980087064,
+      "learning_rate": 9.964835186886692e-06,
+      "loss": 0.2493,
+      "step": 446
+    },
+    {
+      "epoch": 0.06684611933602512,
+      "grad_norm": 2.1443052954533974,
+      "learning_rate": 9.964547881754194e-06,
+      "loss": 0.3611,
+      "step": 447
+    },
+    {
+      "epoch": 0.06699566322715717,
+      "grad_norm": 2.6967138020110712,
+      "learning_rate": 9.964259411890918e-06,
+      "loss": 0.5427,
+      "step": 448
+    },
+    {
+      "epoch": 0.06714520711828922,
+      "grad_norm": 1.688779610685555,
+      "learning_rate": 9.96396977736454e-06,
+      "loss": 0.2569,
+      "step": 449
+    },
+    {
+      "epoch": 0.06729475100942127,
+      "grad_norm": 2.1241026975378694,
+      "learning_rate": 9.963678978243014e-06,
+      "loss": 0.3863,
+      "step": 450
+    },
+    {
+      "epoch": 0.06744429490055331,
+      "grad_norm": 1.9388647656441462,
+      "learning_rate": 9.96338701459456e-06,
+      "loss": 0.2726,
+      "step": 451
+    },
+    {
+      "epoch": 0.06759383879168536,
+      "grad_norm": 1.4657993620125664,
+      "learning_rate": 9.963093886487683e-06,
+      "loss": 0.2338,
+      "step": 452
+    },
+    {
+      "epoch": 0.0677433826828174,
+      "grad_norm": 2.307173509923502,
+      "learning_rate": 9.962799593991146e-06,
+      "loss": 0.8039,
+      "step": 453
+    },
+    {
+      "epoch": 0.06789292657394945,
+      "grad_norm": 1.2669540134016812,
+      "learning_rate": 9.962504137173997e-06,
+      "loss": 0.169,
+      "step": 454
+    },
+    {
+      "epoch": 0.0680424704650815,
+      "grad_norm": 1.5981790001004936,
+      "learning_rate": 9.962207516105552e-06,
+      "loss": 0.2019,
+      "step": 455
+    },
+    {
+      "epoch": 0.06819201435621355,
+      "grad_norm": 1.740837427237262,
+      "learning_rate": 9.9619097308554e-06,
+      "loss": 0.2116,
+      "step": 456
+    },
+    {
+      "epoch": 0.0683415582473456,
+      "grad_norm": 1.9511590671787182,
+      "learning_rate": 9.961610781493407e-06,
+      "loss": 0.2611,
+      "step": 457
+    },
+    {
+      "epoch": 0.06849110213847764,
+      "grad_norm": 1.9814713665794252,
+      "learning_rate": 9.961310668089708e-06,
+      "loss": 0.3714,
+      "step": 458
+    },
+    {
+      "epoch": 0.06864064602960969,
+      "grad_norm": 2.755804773731971,
+      "learning_rate": 9.96100939071471e-06,
+      "loss": 0.5178,
+      "step": 459
+    },
+    {
+      "epoch": 0.06879018992074173,
+      "grad_norm": 2.5378159735000225,
+      "learning_rate": 9.960706949439101e-06,
+      "loss": 0.7334,
+      "step": 460
+    },
+    {
+      "epoch": 0.06893973381187378,
+      "grad_norm": 2.3557582569765003,
+      "learning_rate": 9.960403344333832e-06,
+      "loss": 0.5763,
+      "step": 461
+    },
+    {
+      "epoch": 0.06908927770300584,
+      "grad_norm": 1.6501148783544786,
+      "learning_rate": 9.960098575470131e-06,
+      "loss": 0.3681,
+      "step": 462
+    },
+    {
+      "epoch": 0.06923882159413788,
+      "grad_norm": 1.3521314881367383,
+      "learning_rate": 9.959792642919505e-06,
+      "loss": 0.216,
+      "step": 463
+    },
+    {
+      "epoch": 0.06938836548526993,
+      "grad_norm": 1.9967115308447656,
+      "learning_rate": 9.959485546753724e-06,
+      "loss": 0.4411,
+      "step": 464
+    },
+    {
+      "epoch": 0.06953790937640197,
+      "grad_norm": 1.6934835527025132,
+      "learning_rate": 9.959177287044839e-06,
+      "loss": 0.3013,
+      "step": 465
+    },
+    {
+      "epoch": 0.06968745326753402,
+      "grad_norm": 2.1881268216288703,
+      "learning_rate": 9.958867863865168e-06,
+      "loss": 0.386,
+      "step": 466
+    },
+    {
+      "epoch": 0.06983699715866606,
+      "grad_norm": 1.746249573857031,
+      "learning_rate": 9.958557277287307e-06,
+      "loss": 0.3486,
+      "step": 467
+    },
+    {
+      "epoch": 0.06998654104979811,
+      "grad_norm": 1.3309239290400467,
+      "learning_rate": 9.958245527384118e-06,
+      "loss": 0.2512,
+      "step": 468
+    },
+    {
+      "epoch": 0.07013608494093017,
+      "grad_norm": 1.780095751208227,
+      "learning_rate": 9.957932614228746e-06,
+      "loss": 0.3579,
+      "step": 469
+    },
+    {
+      "epoch": 0.07028562883206221,
+      "grad_norm": 2.058627302052003,
+      "learning_rate": 9.957618537894602e-06,
+      "loss": 0.2234,
+      "step": 470
+    },
+    {
+      "epoch": 0.07043517272319426,
+      "grad_norm": 2.1643867800571286,
+      "learning_rate": 9.95730329845537e-06,
+      "loss": 0.2658,
+      "step": 471
+    },
+    {
+      "epoch": 0.0705847166143263,
+      "grad_norm": 1.9162877246393155,
+      "learning_rate": 9.956986895985009e-06,
+      "loss": 0.3514,
+      "step": 472
+    },
+    {
+      "epoch": 0.07073426050545835,
+      "grad_norm": 2.0198300655217474,
+      "learning_rate": 9.95666933055775e-06,
+      "loss": 0.4191,
+      "step": 473
+    },
+    {
+      "epoch": 0.0708838043965904,
+      "grad_norm": 1.8174642496449622,
+      "learning_rate": 9.956350602248095e-06,
+      "loss": 0.1802,
+      "step": 474
+    },
+    {
+      "epoch": 0.07103334828772245,
+      "grad_norm": 1.7641599345266465,
+      "learning_rate": 9.956030711130824e-06,
+      "loss": 0.2181,
+      "step": 475
+    },
+    {
+      "epoch": 0.0711828921788545,
+      "grad_norm": 1.5149058769435404,
+      "learning_rate": 9.955709657280985e-06,
+      "loss": 0.2068,
+      "step": 476
+    },
+    {
+      "epoch": 0.07133243606998654,
+      "grad_norm": 2.14267612952952,
+      "learning_rate": 9.955387440773902e-06,
+      "loss": 0.2799,
+      "step": 477
+    },
+    {
+      "epoch": 0.07148197996111859,
+      "grad_norm": 1.8794948861297893,
+      "learning_rate": 9.955064061685166e-06,
+      "loss": 0.3437,
+      "step": 478
+    },
+    {
+      "epoch": 0.07163152385225063,
+      "grad_norm": 1.595856928796192,
+      "learning_rate": 9.954739520090649e-06,
+      "loss": 0.1741,
+      "step": 479
+    },
+    {
+      "epoch": 0.07178106774338268,
+      "grad_norm": 1.4775459266699813,
+      "learning_rate": 9.95441381606649e-06,
+      "loss": 0.2009,
+      "step": 480
+    },
+    {
+      "epoch": 0.07193061163451472,
+      "grad_norm": 1.4624583034603231,
+      "learning_rate": 9.954086949689102e-06,
+      "loss": 0.2413,
+      "step": 481
+    },
+    {
+      "epoch": 0.07208015552564678,
+      "grad_norm": 1.5685428117813849,
+      "learning_rate": 9.953758921035171e-06,
+      "loss": 0.2381,
+      "step": 482
+    },
+    {
+      "epoch": 0.07222969941677883,
+      "grad_norm": 2.0490413587537524,
+      "learning_rate": 9.953429730181653e-06,
+      "loss": 0.4092,
+      "step": 483
+    },
+    {
+      "epoch": 0.07237924330791087,
+      "grad_norm": 2.605633491672469,
+      "learning_rate": 9.953099377205786e-06,
+      "loss": 0.56,
+      "step": 484
+    },
+    {
+      "epoch": 0.07252878719904292,
+      "grad_norm": 1.6836189923086853,
+      "learning_rate": 9.952767862185071e-06,
+      "loss": 0.3514,
+      "step": 485
+    },
+    {
+      "epoch": 0.07267833109017496,
+      "grad_norm": 2.165692386982445,
+      "learning_rate": 9.952435185197281e-06,
+      "loss": 0.4363,
+      "step": 486
+    },
+    {
+      "epoch": 0.07282787498130701,
+      "grad_norm": 2.328987566639375,
+      "learning_rate": 9.952101346320471e-06,
+      "loss": 0.5953,
+      "step": 487
+    },
+    {
+      "epoch": 0.07297741887243907,
+      "grad_norm": 1.857109300243422,
+      "learning_rate": 9.951766345632957e-06,
+      "loss": 0.4125,
+      "step": 488
+    },
+    {
+      "epoch": 0.07312696276357111,
+      "grad_norm": 1.780608988332075,
+      "learning_rate": 9.951430183213338e-06,
+      "loss": 0.2793,
+      "step": 489
+    },
+    {
+      "epoch": 0.07327650665470316,
+      "grad_norm": 1.2718866410706833,
+      "learning_rate": 9.951092859140479e-06,
+      "loss": 0.1878,
+      "step": 490
+    },
+    {
+      "epoch": 0.0734260505458352,
+      "grad_norm": 1.389385388824981,
+      "learning_rate": 9.95075437349352e-06,
+      "loss": 0.1922,
+      "step": 491
+    },
+    {
+      "epoch": 0.07357559443696725,
+      "grad_norm": 1.2364018773804621,
+      "learning_rate": 9.950414726351873e-06,
+      "loss": 0.1972,
+      "step": 492
+    },
+    {
+      "epoch": 0.0737251383280993,
+      "grad_norm": 1.6438922682719497,
+      "learning_rate": 9.95007391779522e-06,
+      "loss": 0.3835,
+      "step": 493
+    },
+    {
+      "epoch": 0.07387468221923134,
+      "grad_norm": 1.9223258334837023,
+      "learning_rate": 9.949731947903523e-06,
+      "loss": 0.5421,
+      "step": 494
+    },
+    {
+      "epoch": 0.0740242261103634,
+      "grad_norm": 2.1294087718057955,
+      "learning_rate": 9.949388816757009e-06,
+      "loss": 0.6584,
+      "step": 495
+    },
+    {
+      "epoch": 0.07417377000149544,
+      "grad_norm": 1.9620720670123732,
+      "learning_rate": 9.949044524436178e-06,
+      "loss": 0.3427,
+      "step": 496
+    },
+    {
+      "epoch": 0.07432331389262749,
+      "grad_norm": 1.8767982308843718,
+      "learning_rate": 9.948699071021806e-06,
+      "loss": 0.2221,
+      "step": 497
+    },
+    {
+      "epoch": 0.07447285778375953,
+      "grad_norm": 1.5717369659821445,
+      "learning_rate": 9.948352456594938e-06,
+      "loss": 0.3915,
+      "step": 498
+    },
+    {
+      "epoch": 0.07462240167489158,
+      "grad_norm": 1.9105988284269253,
+      "learning_rate": 9.948004681236896e-06,
+      "loss": 0.4049,
+      "step": 499
+    },
+    {
+      "epoch": 0.07477194556602362,
+      "grad_norm": 2.051255434710168,
+      "learning_rate": 9.94765574502927e-06,
+      "loss": 0.263,
+      "step": 500
+    },
+    {
+      "epoch": 0.07492148945715567,
+      "grad_norm": 1.1727115808022262,
+      "learning_rate": 9.947305648053924e-06,
+      "loss": 0.2061,
+      "step": 501
+    },
+    {
+      "epoch": 0.07507103334828773,
+      "grad_norm": 2.3851218898633566,
+      "learning_rate": 9.946954390392995e-06,
+      "loss": 0.3587,
+      "step": 502
+    },
+    {
+      "epoch": 0.07522057723941977,
+      "grad_norm": 2.668333899893354,
+      "learning_rate": 9.94660197212889e-06,
+      "loss": 0.279,
+      "step": 503
+    },
+    {
+      "epoch": 0.07537012113055182,
+      "grad_norm": 2.324044177768054,
+      "learning_rate": 9.946248393344289e-06,
+      "loss": 0.5219,
+      "step": 504
+    },
+    {
+      "epoch": 0.07551966502168386,
+      "grad_norm": 2.252535927387564,
+      "learning_rate": 9.945893654122147e-06,
+      "loss": 0.4462,
+      "step": 505
+    },
+    {
+      "epoch": 0.07566920891281591,
+      "grad_norm": 1.2553962948323492,
+      "learning_rate": 9.945537754545689e-06,
+      "loss": 0.1829,
+      "step": 506
+    },
+    {
+      "epoch": 0.07581875280394795,
+      "grad_norm": 2.009514792075129,
+      "learning_rate": 9.94518069469841e-06,
+      "loss": 0.334,
+      "step": 507
+    },
+    {
+      "epoch": 0.07596829669508001,
+      "grad_norm": 1.7045023449590413,
+      "learning_rate": 9.944822474664082e-06,
+      "loss": 0.3202,
+      "step": 508
+    },
+    {
+      "epoch": 0.07611784058621206,
+      "grad_norm": 1.0508191419172128,
+      "learning_rate": 9.944463094526747e-06,
+      "loss": 0.205,
+      "step": 509
+    },
+    {
+      "epoch": 0.0762673844773441,
+      "grad_norm": 1.6097293192900886,
+      "learning_rate": 9.944102554370718e-06,
+      "loss": 0.2324,
+      "step": 510
+    },
+    {
+      "epoch": 0.07641692836847615,
+      "grad_norm": 1.9399148366487866,
+      "learning_rate": 9.943740854280582e-06,
+      "loss": 0.4526,
+      "step": 511
+    },
+    {
+      "epoch": 0.07656647225960819,
+      "grad_norm": 2.0362256511499335,
+      "learning_rate": 9.943377994341197e-06,
+      "loss": 0.3979,
+      "step": 512
+    },
+    {
+      "epoch": 0.07671601615074024,
+      "grad_norm": 1.5296316888698338,
+      "learning_rate": 9.943013974637693e-06,
+      "loss": 0.3789,
+      "step": 513
+    },
+    {
+      "epoch": 0.07686556004187228,
+      "grad_norm": 1.496691000675503,
+      "learning_rate": 9.942648795255473e-06,
+      "loss": 0.2497,
+      "step": 514
+    },
+    {
+      "epoch": 0.07701510393300434,
+      "grad_norm": 1.4146486247851384,
+      "learning_rate": 9.942282456280212e-06,
+      "loss": 0.3088,
+      "step": 515
+    },
+    {
+      "epoch": 0.07716464782413639,
+      "grad_norm": 1.3671722765483707,
+      "learning_rate": 9.941914957797855e-06,
+      "loss": 0.2076,
+      "step": 516
+    },
+    {
+      "epoch": 0.07731419171526843,
+      "grad_norm": 1.8485057563465108,
+      "learning_rate": 9.941546299894623e-06,
+      "loss": 0.3676,
+      "step": 517
+    },
+    {
+      "epoch": 0.07746373560640048,
+      "grad_norm": 2.0438588429845255,
+      "learning_rate": 9.941176482657005e-06,
+      "loss": 0.4905,
+      "step": 518
+    },
+    {
+      "epoch": 0.07761327949753252,
+      "grad_norm": 1.3215533906334498,
+      "learning_rate": 9.940805506171765e-06,
+      "loss": 0.2028,
+      "step": 519
+    },
+    {
+      "epoch": 0.07776282338866457,
+      "grad_norm": 2.499241081917891,
+      "learning_rate": 9.940433370525937e-06,
+      "loss": 0.4323,
+      "step": 520
+    },
+    {
+      "epoch": 0.07791236727979663,
+      "grad_norm": 1.4654220634749195,
+      "learning_rate": 9.940060075806827e-06,
+      "loss": 0.1928,
+      "step": 521
+    },
+    {
+      "epoch": 0.07806191117092867,
+      "grad_norm": 2.32501667334618,
+      "learning_rate": 9.939685622102013e-06,
+      "loss": 0.6039,
+      "step": 522
+    },
+    {
+      "epoch": 0.07821145506206072,
+      "grad_norm": 2.0353313744113644,
+      "learning_rate": 9.939310009499348e-06,
+      "loss": 0.434,
+      "step": 523
+    },
+    {
+      "epoch": 0.07836099895319276,
+      "grad_norm": 1.5916248439200642,
+      "learning_rate": 9.938933238086952e-06,
+      "loss": 0.2484,
+      "step": 524
+    },
+    {
+      "epoch": 0.07851054284432481,
+      "grad_norm": 1.510761606083,
+      "learning_rate": 9.938555307953221e-06,
+      "loss": 0.2761,
+      "step": 525
+    },
+    {
+      "epoch": 0.07866008673545685,
+      "grad_norm": 1.6041562012438388,
+      "learning_rate": 9.93817621918682e-06,
+      "loss": 0.3032,
+      "step": 526
+    },
+    {
+      "epoch": 0.0788096306265889,
+      "grad_norm": 1.5831322947558841,
+      "learning_rate": 9.937795971876686e-06,
+      "loss": 0.3486,
+      "step": 527
+    },
+    {
+      "epoch": 0.07895917451772096,
+      "grad_norm": 2.2247878916503856,
+      "learning_rate": 9.93741456611203e-06,
+      "loss": 0.4087,
+      "step": 528
+    },
+    {
+      "epoch": 0.079108718408853,
+      "grad_norm": 2.152252638423622,
+      "learning_rate": 9.937032001982334e-06,
+      "loss": 0.5629,
+      "step": 529
+    },
+    {
+      "epoch": 0.07925826229998505,
+      "grad_norm": 2.0483514105705525,
+      "learning_rate": 9.93664827957735e-06,
+      "loss": 0.5279,
+      "step": 530
+    },
+    {
+      "epoch": 0.07940780619111709,
+      "grad_norm": 1.2448870158155207,
+      "learning_rate": 9.936263398987103e-06,
+      "loss": 0.3744,
+      "step": 531
+    },
+    {
+      "epoch": 0.07955735008224914,
+      "grad_norm": 0.9489762178863248,
+      "learning_rate": 9.93587736030189e-06,
+      "loss": 0.1631,
+      "step": 532
+    },
+    {
+      "epoch": 0.07970689397338118,
+      "grad_norm": 1.3545590640653586,
+      "learning_rate": 9.935490163612279e-06,
+      "loss": 0.1975,
+      "step": 533
+    },
+    {
+      "epoch": 0.07985643786451324,
+      "grad_norm": 1.3663228011672384,
+      "learning_rate": 9.93510180900911e-06,
+      "loss": 0.184,
+      "step": 534
+    },
+    {
+      "epoch": 0.08000598175564529,
+      "grad_norm": 1.5768436668872405,
+      "learning_rate": 9.934712296583497e-06,
+      "loss": 0.3183,
+      "step": 535
+    },
+    {
+      "epoch": 0.08015552564677733,
+      "grad_norm": 1.926347057489139,
+      "learning_rate": 9.93432162642682e-06,
+      "loss": 0.3305,
+      "step": 536
+    },
+    {
+      "epoch": 0.08030506953790938,
+      "grad_norm": 2.0791782850566474,
+      "learning_rate": 9.933929798630738e-06,
+      "loss": 0.5009,
+      "step": 537
+    },
+    {
+      "epoch": 0.08045461342904142,
+      "grad_norm": 2.1023331544425523,
+      "learning_rate": 9.933536813287172e-06,
+      "loss": 0.4292,
+      "step": 538
+    },
+    {
+      "epoch": 0.08060415732017347,
+      "grad_norm": 2.8605361415271493,
+      "learning_rate": 9.933142670488324e-06,
+      "loss": 0.2666,
+      "step": 539
+    },
+    {
+      "epoch": 0.08075370121130551,
+      "grad_norm": 2.7087693572573968,
+      "learning_rate": 9.932747370326664e-06,
+      "loss": 0.2544,
+      "step": 540
+    },
+    {
+      "epoch": 0.08090324510243757,
+      "grad_norm": 1.5804074183588281,
+      "learning_rate": 9.932350912894932e-06,
+      "loss": 0.2089,
+      "step": 541
+    },
+    {
+      "epoch": 0.08105278899356962,
+      "grad_norm": 1.6448934387271092,
+      "learning_rate": 9.931953298286141e-06,
+      "loss": 0.181,
+      "step": 542
+    },
+    {
+      "epoch": 0.08120233288470166,
+      "grad_norm": 1.373017928034036,
+      "learning_rate": 9.931554526593576e-06,
+      "loss": 0.3218,
+      "step": 543
+    },
+    {
+      "epoch": 0.0813518767758337,
+      "grad_norm": 1.4895748889012388,
+      "learning_rate": 9.931154597910791e-06,
+      "loss": 0.2472,
+      "step": 544
+    },
+    {
+      "epoch": 0.08150142066696575,
+      "grad_norm": 2.064608760225509,
+      "learning_rate": 9.930753512331615e-06,
+      "loss": 0.3765,
+      "step": 545
+    },
+    {
+      "epoch": 0.0816509645580978,
+      "grad_norm": 1.6526846905937504,
+      "learning_rate": 9.930351269950144e-06,
+      "loss": 0.3177,
+      "step": 546
+    },
+    {
+      "epoch": 0.08180050844922986,
+      "grad_norm": 2.047798829134187,
+      "learning_rate": 9.92994787086075e-06,
+      "loss": 0.3192,
+      "step": 547
+    },
+    {
+      "epoch": 0.0819500523403619,
+      "grad_norm": 2.122394373762569,
+      "learning_rate": 9.929543315158073e-06,
+      "loss": 0.5554,
+      "step": 548
+    },
+    {
+      "epoch": 0.08209959623149395,
+      "grad_norm": 2.311960518258969,
+      "learning_rate": 9.929137602937028e-06,
+      "loss": 0.3797,
+      "step": 549
+    },
+    {
+      "epoch": 0.08224914012262599,
+      "grad_norm": 1.8449832380251867,
+      "learning_rate": 9.928730734292797e-06,
+      "loss": 0.3894,
+      "step": 550
+    },
+    {
+      "epoch": 0.08239868401375804,
+      "grad_norm": 1.995255157883457,
+      "learning_rate": 9.928322709320834e-06,
+      "loss": 0.3925,
+      "step": 551
+    },
+    {
+      "epoch": 0.08254822790489008,
+      "grad_norm": 2.755405061449222,
+      "learning_rate": 9.92791352811687e-06,
+      "loss": 0.6899,
+      "step": 552
+    },
+    {
+      "epoch": 0.08269777179602213,
+      "grad_norm": 1.2254981142470793,
+      "learning_rate": 9.9275031907769e-06,
+      "loss": 0.2225,
+      "step": 553
+    },
+    {
+      "epoch": 0.08284731568715419,
+      "grad_norm": 1.9323036995913243,
+      "learning_rate": 9.927091697397192e-06,
+      "loss": 0.3865,
+      "step": 554
+    },
+    {
+      "epoch": 0.08299685957828623,
+      "grad_norm": 2.0962863974348593,
+      "learning_rate": 9.926679048074289e-06,
+      "loss": 0.4,
+      "step": 555
+    },
+    {
+      "epoch": 0.08314640346941828,
+      "grad_norm": 1.5847691098448267,
+      "learning_rate": 9.926265242904998e-06,
+      "loss": 0.247,
+      "step": 556
+    },
+    {
+      "epoch": 0.08329594736055032,
+      "grad_norm": 2.5967594290859903,
+      "learning_rate": 9.925850281986408e-06,
+      "loss": 0.2083,
+      "step": 557
+    },
+    {
+      "epoch": 0.08344549125168237,
+      "grad_norm": 2.0426826933231226,
+      "learning_rate": 9.925434165415868e-06,
+      "loss": 0.449,
+      "step": 558
+    },
+    {
+      "epoch": 0.08359503514281441,
+      "grad_norm": 1.7693278888452375,
+      "learning_rate": 9.925016893291007e-06,
+      "loss": 0.2789,
+      "step": 559
+    },
+    {
+      "epoch": 0.08374457903394647,
+      "grad_norm": 1.6227416269049326,
+      "learning_rate": 9.924598465709717e-06,
+      "loss": 0.2209,
+      "step": 560
+    },
+    {
+      "epoch": 0.08389412292507852,
+      "grad_norm": 1.7055307729140163,
+      "learning_rate": 9.924178882770166e-06,
+      "loss": 0.3554,
+      "step": 561
+    },
+    {
+      "epoch": 0.08404366681621056,
+      "grad_norm": 1.9245436136675982,
+      "learning_rate": 9.923758144570792e-06,
+      "loss": 0.5343,
+      "step": 562
+    },
+    {
+      "epoch": 0.0841932107073426,
+      "grad_norm": 1.3916186974123048,
+      "learning_rate": 9.923336251210306e-06,
+      "loss": 0.2328,
+      "step": 563
+    },
+    {
+      "epoch": 0.08434275459847465,
+      "grad_norm": 1.8724253939088875,
+      "learning_rate": 9.92291320278769e-06,
+      "loss": 0.2691,
+      "step": 564
+    },
+    {
+      "epoch": 0.0844922984896067,
+      "grad_norm": 1.545927153493535,
+      "learning_rate": 9.922488999402191e-06,
+      "loss": 0.2049,
+      "step": 565
+    },
+    {
+      "epoch": 0.08464184238073874,
+      "grad_norm": 2.216312298348258,
+      "learning_rate": 9.922063641153332e-06,
+      "loss": 0.5844,
+      "step": 566
+    },
+    {
+      "epoch": 0.0847913862718708,
+      "grad_norm": 1.2444734652143745,
+      "learning_rate": 9.921637128140909e-06,
+      "loss": 0.2872,
+      "step": 567
+    },
+    {
+      "epoch": 0.08494093016300285,
+      "grad_norm": 2.133851301389792,
+      "learning_rate": 9.921209460464983e-06,
+      "loss": 0.2418,
+      "step": 568
+    },
+    {
+      "epoch": 0.08509047405413489,
+      "grad_norm": 1.5462263702909163,
+      "learning_rate": 9.92078063822589e-06,
+      "loss": 0.3438,
+      "step": 569
+    },
+    {
+      "epoch": 0.08524001794526694,
+      "grad_norm": 2.341879963295622,
+      "learning_rate": 9.920350661524237e-06,
+      "loss": 0.5783,
+      "step": 570
+    },
+    {
+      "epoch": 0.08538956183639898,
+      "grad_norm": 1.7633187330163729,
+      "learning_rate": 9.919919530460899e-06,
+      "loss": 0.3503,
+      "step": 571
+    },
+    {
+      "epoch": 0.08553910572753103,
+      "grad_norm": 2.1676160714531107,
+      "learning_rate": 9.919487245137024e-06,
+      "loss": 0.2098,
+      "step": 572
+    },
+    {
+      "epoch": 0.08568864961866307,
+      "grad_norm": 2.198855334486466,
+      "learning_rate": 9.919053805654029e-06,
+      "loss": 0.3876,
+      "step": 573
+    },
+    {
+      "epoch": 0.08583819350979513,
+      "grad_norm": 1.821472616891953,
+      "learning_rate": 9.918619212113607e-06,
+      "loss": 0.391,
+      "step": 574
+    },
+    {
+      "epoch": 0.08598773740092717,
+      "grad_norm": 1.4553776733520012,
+      "learning_rate": 9.918183464617714e-06,
+      "loss": 0.2032,
+      "step": 575
+    },
+    {
+      "epoch": 0.08613728129205922,
+      "grad_norm": 1.5817735791823646,
+      "learning_rate": 9.917746563268581e-06,
+      "loss": 0.2658,
+      "step": 576
+    },
+    {
+      "epoch": 0.08628682518319127,
+      "grad_norm": 2.255323258805483,
+      "learning_rate": 9.917308508168712e-06,
+      "loss": 0.39,
+      "step": 577
+    },
+    {
+      "epoch": 0.08643636907432331,
+      "grad_norm": 1.699175902078527,
+      "learning_rate": 9.916869299420875e-06,
+      "loss": 0.1906,
+      "step": 578
+    },
+    {
+      "epoch": 0.08658591296545536,
+      "grad_norm": 1.5572993513277051,
+      "learning_rate": 9.916428937128117e-06,
+      "loss": 0.3438,
+      "step": 579
+    },
+    {
+      "epoch": 0.08673545685658741,
+      "grad_norm": 1.5095119263162684,
+      "learning_rate": 9.915987421393747e-06,
+      "loss": 0.272,
+      "step": 580
+    },
+    {
+      "epoch": 0.08688500074771946,
+      "grad_norm": 2.8137128440101735,
+      "learning_rate": 9.91554475232135e-06,
+      "loss": 0.3833,
+      "step": 581
+    },
+    {
+      "epoch": 0.0870345446388515,
+      "grad_norm": 1.845156278788705,
+      "learning_rate": 9.915100930014786e-06,
+      "loss": 0.4658,
+      "step": 582
+    },
+    {
+      "epoch": 0.08718408852998355,
+      "grad_norm": 1.7624433765379017,
+      "learning_rate": 9.914655954578171e-06,
+      "loss": 0.3968,
+      "step": 583
+    },
+    {
+      "epoch": 0.0873336324211156,
+      "grad_norm": 1.7915618837196812,
+      "learning_rate": 9.914209826115906e-06,
+      "loss": 0.4901,
+      "step": 584
+    },
+    {
+      "epoch": 0.08748317631224764,
+      "grad_norm": 1.8335500777788887,
+      "learning_rate": 9.913762544732654e-06,
+      "loss": 0.249,
+      "step": 585
+    },
+    {
+      "epoch": 0.08763272020337969,
+      "grad_norm": 1.5116580783389033,
+      "learning_rate": 9.913314110533355e-06,
+      "loss": 0.3999,
+      "step": 586
+    },
+    {
+      "epoch": 0.08778226409451174,
+      "grad_norm": 1.9828537343745032,
+      "learning_rate": 9.912864523623214e-06,
+      "loss": 0.4153,
+      "step": 587
+    },
+    {
+      "epoch": 0.08793180798564379,
+      "grad_norm": 1.6056147158647165,
+      "learning_rate": 9.912413784107709e-06,
+      "loss": 0.357,
+      "step": 588
+    },
+    {
+      "epoch": 0.08808135187677583,
+      "grad_norm": 1.7642170812152784,
+      "learning_rate": 9.911961892092587e-06,
+      "loss": 0.3425,
+      "step": 589
+    },
+    {
+      "epoch": 0.08823089576790788,
+      "grad_norm": 1.925307511563271,
+      "learning_rate": 9.911508847683867e-06,
+      "loss": 0.4476,
+      "step": 590
+    },
+    {
+      "epoch": 0.08838043965903992,
+      "grad_norm": 1.9824372539957273,
+      "learning_rate": 9.911054650987837e-06,
+      "loss": 0.4597,
+      "step": 591
+    },
+    {
+      "epoch": 0.08852998355017197,
+      "grad_norm": 1.5805088418089035,
+      "learning_rate": 9.910599302111057e-06,
+      "loss": 0.1935,
+      "step": 592
+    },
+    {
+      "epoch": 0.08867952744130403,
+      "grad_norm": 2.157404890931188,
+      "learning_rate": 9.910142801160355e-06,
+      "loss": 0.3443,
+      "step": 593
+    },
+    {
+      "epoch": 0.08882907133243607,
+      "grad_norm": 2.094900000445731,
+      "learning_rate": 9.909685148242831e-06,
+      "loss": 0.404,
+      "step": 594
+    },
+    {
+      "epoch": 0.08897861522356812,
+      "grad_norm": 2.336415519412793,
+      "learning_rate": 9.909226343465856e-06,
+      "loss": 0.6382,
+      "step": 595
+    },
+    {
+      "epoch": 0.08912815911470016,
+      "grad_norm": 2.0552137049182497,
+      "learning_rate": 9.908766386937067e-06,
+      "loss": 0.3908,
+      "step": 596
+    },
+    {
+      "epoch": 0.08927770300583221,
+      "grad_norm": 1.1564393734179468,
+      "learning_rate": 9.908305278764376e-06,
+      "loss": 0.2457,
+      "step": 597
+    },
+    {
+      "epoch": 0.08942724689696425,
+      "grad_norm": 1.8704284289450437,
+      "learning_rate": 9.907843019055966e-06,
+      "loss": 0.3604,
+      "step": 598
+    },
+    {
+      "epoch": 0.0895767907880963,
+      "grad_norm": 1.295042190600909,
+      "learning_rate": 9.907379607920281e-06,
+      "loss": 0.2075,
+      "step": 599
+    },
+    {
+      "epoch": 0.08972633467922836,
+      "grad_norm": 1.8305770820800886,
+      "learning_rate": 9.90691504546605e-06,
+      "loss": 0.2698,
+      "step": 600
+    },
+    {
+      "epoch": 0.0898758785703604,
+      "grad_norm": 1.7240290275544472,
+      "learning_rate": 9.906449331802256e-06,
+      "loss": 0.2504,
+      "step": 601
+    },
+    {
+      "epoch": 0.09002542246149245,
+      "grad_norm": 1.0036789417827203,
+      "learning_rate": 9.905982467038167e-06,
+      "loss": 0.195,
+      "step": 602
+    },
+    {
+      "epoch": 0.0901749663526245,
+      "grad_norm": 1.6777253578130231,
+      "learning_rate": 9.905514451283308e-06,
+      "loss": 0.2436,
+      "step": 603
+    },
+    {
+      "epoch": 0.09032451024375654,
+      "grad_norm": 1.9190873052270145,
+      "learning_rate": 9.905045284647483e-06,
+      "loss": 0.4006,
+      "step": 604
+    },
+    {
+      "epoch": 0.09047405413488858,
+      "grad_norm": 1.77001911452716,
+      "learning_rate": 9.904574967240764e-06,
+      "loss": 0.3703,
+      "step": 605
+    },
+    {
+      "epoch": 0.09062359802602064,
+      "grad_norm": 1.3114492277508998,
+      "learning_rate": 9.904103499173487e-06,
+      "loss": 0.2323,
+      "step": 606
+    },
+    {
+      "epoch": 0.09077314191715269,
+      "grad_norm": 1.6694643051834908,
+      "learning_rate": 9.90363088055627e-06,
+      "loss": 0.2881,
+      "step": 607
+    },
+    {
+      "epoch": 0.09092268580828473,
+      "grad_norm": 1.4448454411512122,
+      "learning_rate": 9.903157111499988e-06,
+      "loss": 0.2341,
+      "step": 608
+    },
+    {
+      "epoch": 0.09107222969941678,
+      "grad_norm": 1.8302982894061834,
+      "learning_rate": 9.902682192115795e-06,
+      "loss": 0.3497,
+      "step": 609
+    },
+    {
+      "epoch": 0.09122177359054882,
+      "grad_norm": 1.4089802820999182,
+      "learning_rate": 9.902206122515113e-06,
+      "loss": 0.1565,
+      "step": 610
+    },
+    {
+      "epoch": 0.09137131748168087,
+      "grad_norm": 2.275670976517465,
+      "learning_rate": 9.901728902809627e-06,
+      "loss": 0.482,
+      "step": 611
+    },
+    {
+      "epoch": 0.09152086137281291,
+      "grad_norm": 2.3916744409549997,
+      "learning_rate": 9.901250533111301e-06,
+      "loss": 0.539,
+      "step": 612
+    },
+    {
+      "epoch": 0.09167040526394497,
+      "grad_norm": 1.110965438282227,
+      "learning_rate": 9.900771013532367e-06,
+      "loss": 0.2257,
+      "step": 613
+    },
+    {
+      "epoch": 0.09181994915507702,
+      "grad_norm": 1.6169969209154105,
+      "learning_rate": 9.900290344185321e-06,
+      "loss": 0.2316,
+      "step": 614
+    },
+    {
+      "epoch": 0.09196949304620906,
+      "grad_norm": 1.390950490331229,
+      "learning_rate": 9.899808525182935e-06,
+      "loss": 0.1735,
+      "step": 615
+    },
+    {
+      "epoch": 0.09211903693734111,
+      "grad_norm": 1.26641152514348,
+      "learning_rate": 9.899325556638247e-06,
+      "loss": 0.2269,
+      "step": 616
+    },
+    {
+      "epoch": 0.09226858082847315,
+      "grad_norm": 1.107259968960053,
+      "learning_rate": 9.898841438664568e-06,
+      "loss": 0.2082,
+      "step": 617
+    },
+    {
+      "epoch": 0.0924181247196052,
+      "grad_norm": 1.6779136428714192,
+      "learning_rate": 9.898356171375473e-06,
+      "loss": 0.3744,
+      "step": 618
+    },
+    {
+      "epoch": 0.09256766861073726,
+      "grad_norm": 1.8012739115801626,
+      "learning_rate": 9.897869754884816e-06,
+      "loss": 0.2438,
+      "step": 619
+    },
+    {
+      "epoch": 0.0927172125018693,
+      "grad_norm": 1.6400812519548655,
+      "learning_rate": 9.89738218930671e-06,
+      "loss": 0.3692,
+      "step": 620
+    },
+    {
+      "epoch": 0.09286675639300135,
+      "grad_norm": 2.7659374426954972,
+      "learning_rate": 9.896893474755547e-06,
+      "loss": 0.5873,
+      "step": 621
+    },
+    {
+      "epoch": 0.0930163002841334,
+      "grad_norm": 3.020452608035097,
+      "learning_rate": 9.89640361134598e-06,
+      "loss": 0.4177,
+      "step": 622
+    },
+    {
+      "epoch": 0.09316584417526544,
+      "grad_norm": 1.4907614824403637,
+      "learning_rate": 9.895912599192937e-06,
+      "loss": 0.2516,
+      "step": 623
+    },
+    {
+      "epoch": 0.09331538806639748,
+      "grad_norm": 1.6636615032724535,
+      "learning_rate": 9.895420438411616e-06,
+      "loss": 0.1935,
+      "step": 624
+    },
+    {
+      "epoch": 0.09346493195752953,
+      "grad_norm": 1.9719905447621995,
+      "learning_rate": 9.89492712911748e-06,
+      "loss": 0.2135,
+      "step": 625
+    },
+    {
+      "epoch": 0.09361447584866159,
+      "grad_norm": 1.3681787330772102,
+      "learning_rate": 9.894432671426264e-06,
+      "loss": 0.208,
+      "step": 626
+    },
+    {
+      "epoch": 0.09376401973979363,
+      "grad_norm": 2.0793649946453043,
+      "learning_rate": 9.893937065453976e-06,
+      "loss": 0.3719,
+      "step": 627
+    },
+    {
+      "epoch": 0.09391356363092568,
+      "grad_norm": 1.685584025343787,
+      "learning_rate": 9.893440311316887e-06,
+      "loss": 0.2164,
+      "step": 628
+    },
+    {
+      "epoch": 0.09406310752205772,
+      "grad_norm": 1.2145425693019332,
+      "learning_rate": 9.892942409131541e-06,
+      "loss": 0.1725,
+      "step": 629
+    },
+    {
+      "epoch": 0.09421265141318977,
+      "grad_norm": 1.1438517718036314,
+      "learning_rate": 9.892443359014752e-06,
+      "loss": 0.2367,
+      "step": 630
+    },
+    {
+      "epoch": 0.09436219530432181,
+      "grad_norm": 1.4416913213257094,
+      "learning_rate": 9.8919431610836e-06,
+      "loss": 0.2254,
+      "step": 631
+    },
+    {
+      "epoch": 0.09451173919545386,
+      "grad_norm": 1.2656296241346114,
+      "learning_rate": 9.891441815455436e-06,
+      "loss": 0.2485,
+      "step": 632
+    },
+    {
+      "epoch": 0.09466128308658592,
+      "grad_norm": 1.4276056880724206,
+      "learning_rate": 9.890939322247881e-06,
+      "loss": 0.1908,
+      "step": 633
+    },
+    {
+      "epoch": 0.09481082697771796,
+      "grad_norm": 1.8185771152087218,
+      "learning_rate": 9.890435681578827e-06,
+      "loss": 0.2096,
+      "step": 634
+    },
+    {
+      "epoch": 0.09496037086885001,
+      "grad_norm": 1.2794518689910337,
+      "learning_rate": 9.88993089356643e-06,
+      "loss": 0.2394,
+      "step": 635
+    },
+    {
+      "epoch": 0.09510991475998205,
+      "grad_norm": 2.0227594086297738,
+      "learning_rate": 9.88942495832912e-06,
+      "loss": 0.59,
+      "step": 636
+    },
+    {
+      "epoch": 0.0952594586511141,
+      "grad_norm": 1.3323082817593526,
+      "learning_rate": 9.888917875985593e-06,
+      "loss": 0.2073,
+      "step": 637
+    },
+    {
+      "epoch": 0.09540900254224614,
+      "grad_norm": 1.7884206661676574,
+      "learning_rate": 9.888409646654818e-06,
+      "loss": 0.3897,
+      "step": 638
+    },
+    {
+      "epoch": 0.0955585464333782,
+      "grad_norm": 2.124144136353745,
+      "learning_rate": 9.887900270456025e-06,
+      "loss": 0.5683,
+      "step": 639
+    },
+    {
+      "epoch": 0.09570809032451025,
+      "grad_norm": 1.4793433841619534,
+      "learning_rate": 9.887389747508725e-06,
+      "loss": 0.3727,
+      "step": 640
+    },
+    {
+      "epoch": 0.09585763421564229,
+      "grad_norm": 1.0661747667222115,
+      "learning_rate": 9.88687807793269e-06,
+      "loss": 0.1983,
+      "step": 641
+    },
+    {
+      "epoch": 0.09600717810677434,
+      "grad_norm": 1.615153009655538,
+      "learning_rate": 9.886365261847957e-06,
+      "loss": 0.3675,
+      "step": 642
+    },
+    {
+      "epoch": 0.09615672199790638,
+      "grad_norm": 1.4963878387365324,
+      "learning_rate": 9.885851299374844e-06,
+      "loss": 0.1805,
+      "step": 643
+    },
+    {
+      "epoch": 0.09630626588903843,
+      "grad_norm": 1.8529323065992462,
+      "learning_rate": 9.88533619063393e-06,
+      "loss": 0.391,
+      "step": 644
+    },
+    {
+      "epoch": 0.09645580978017047,
+      "grad_norm": 2.4764246014732145,
+      "learning_rate": 9.884819935746063e-06,
+      "loss": 0.2605,
+      "step": 645
+    },
+    {
+      "epoch": 0.09660535367130253,
+      "grad_norm": 1.904672440883197,
+      "learning_rate": 9.884302534832361e-06,
+      "loss": 0.3935,
+      "step": 646
+    },
+    {
+      "epoch": 0.09675489756243458,
+      "grad_norm": 1.9431435460380113,
+      "learning_rate": 9.883783988014216e-06,
+      "loss": 0.2092,
+      "step": 647
+    },
+    {
+      "epoch": 0.09690444145356662,
+      "grad_norm": 2.0946695671241553,
+      "learning_rate": 9.883264295413278e-06,
+      "loss": 0.3957,
+      "step": 648
+    },
+    {
+      "epoch": 0.09705398534469867,
+      "grad_norm": 1.0944344711946927,
+      "learning_rate": 9.882743457151476e-06,
+      "loss": 0.202,
+      "step": 649
+    },
+    {
+      "epoch": 0.09720352923583071,
+      "grad_norm": 1.5147259026498003,
+      "learning_rate": 9.882221473351e-06,
+      "loss": 0.3029,
+      "step": 650
+    },
+    {
+      "epoch": 0.09735307312696276,
+      "grad_norm": 1.3452835965457643,
+      "learning_rate": 9.881698344134316e-06,
+      "loss": 0.2159,
+      "step": 651
+    },
+    {
+      "epoch": 0.09750261701809482,
+      "grad_norm": 1.7952640402406481,
+      "learning_rate": 9.881174069624155e-06,
+      "loss": 0.4006,
+      "step": 652
+    },
+    {
+      "epoch": 0.09765216090922686,
+      "grad_norm": 2.468540255171398,
+      "learning_rate": 9.880648649943515e-06,
+      "loss": 0.4393,
+      "step": 653
+    },
+    {
+      "epoch": 0.09780170480035891,
+      "grad_norm": 1.5332585075726441,
+      "learning_rate": 9.880122085215664e-06,
+      "loss": 0.2401,
+      "step": 654
+    },
+    {
+      "epoch": 0.09795124869149095,
+      "grad_norm": 1.5882881108110953,
+      "learning_rate": 9.87959437556414e-06,
+      "loss": 0.2078,
+      "step": 655
+    },
+    {
+      "epoch": 0.098100792582623,
+      "grad_norm": 1.7962702189497488,
+      "learning_rate": 9.87906552111275e-06,
+      "loss": 0.4793,
+      "step": 656
+    },
+    {
+      "epoch": 0.09825033647375504,
+      "grad_norm": 1.860004859316795,
+      "learning_rate": 9.878535521985568e-06,
+      "loss": 0.2388,
+      "step": 657
+    },
+    {
+      "epoch": 0.09839988036488709,
+      "grad_norm": 1.9861019609665855,
+      "learning_rate": 9.878004378306934e-06,
+      "loss": 0.3721,
+      "step": 658
+    },
+    {
+      "epoch": 0.09854942425601915,
+      "grad_norm": 1.5404208138898199,
+      "learning_rate": 9.877472090201463e-06,
+      "loss": 0.3534,
+      "step": 659
+    },
+    {
+      "epoch": 0.09869896814715119,
+      "grad_norm": 3.0119825067072306,
+      "learning_rate": 9.876938657794036e-06,
+      "loss": 0.6732,
+      "step": 660
+    },
+    {
+      "epoch": 0.09884851203828324,
+      "grad_norm": 1.5069735817087104,
+      "learning_rate": 9.876404081209796e-06,
+      "loss": 0.4004,
+      "step": 661
+    },
+    {
+      "epoch": 0.09899805592941528,
+      "grad_norm": 1.6856753387650372,
+      "learning_rate": 9.875868360574164e-06,
+      "loss": 0.2942,
+      "step": 662
+    },
+    {
+      "epoch": 0.09914759982054733,
+      "grad_norm": 1.6896901311725145,
+      "learning_rate": 9.875331496012822e-06,
+      "loss": 0.239,
+      "step": 663
+    },
+    {
+      "epoch": 0.09929714371167937,
+      "grad_norm": 2.2770505228904225,
+      "learning_rate": 9.87479348765173e-06,
+      "loss": 0.4755,
+      "step": 664
+    },
+    {
+      "epoch": 0.09944668760281143,
+      "grad_norm": 1.9016485099179228,
+      "learning_rate": 9.874254335617102e-06,
+      "loss": 0.4645,
+      "step": 665
+    },
+    {
+      "epoch": 0.09959623149394348,
+      "grad_norm": 1.6638896812103354,
+      "learning_rate": 9.873714040035434e-06,
+      "loss": 0.2512,
+      "step": 666
+    },
+    {
+      "epoch": 0.09974577538507552,
+      "grad_norm": 1.7233554952000107,
+      "learning_rate": 9.873172601033482e-06,
+      "loss": 0.3958,
+      "step": 667
+    },
+    {
+      "epoch": 0.09989531927620757,
+      "grad_norm": 1.7250170911584946,
+      "learning_rate": 9.872630018738271e-06,
+      "loss": 0.3115,
+      "step": 668
+    },
+    {
+      "epoch": 0.10004486316733961,
+      "grad_norm": 1.8843746906489027,
+      "learning_rate": 9.872086293277101e-06,
+      "loss": 0.3789,
+      "step": 669
+    },
+    {
+      "epoch": 0.10019440705847166,
+      "grad_norm": 1.943275185299739,
+      "learning_rate": 9.871541424777534e-06,
+      "loss": 0.4192,
+      "step": 670
+    },
+    {
+      "epoch": 0.1003439509496037,
+      "grad_norm": 1.4918005726247283,
+      "learning_rate": 9.870995413367397e-06,
+      "loss": 0.2538,
+      "step": 671
+    },
+    {
+      "epoch": 0.10049349484073576,
+      "grad_norm": 1.6441123648652987,
+      "learning_rate": 9.870448259174791e-06,
+      "loss": 0.2295,
+      "step": 672
+    },
+    {
+      "epoch": 0.1006430387318678,
+      "grad_norm": 1.933429186975597,
+      "learning_rate": 9.86989996232809e-06,
+      "loss": 0.4015,
+      "step": 673
+    },
+    {
+      "epoch": 0.10079258262299985,
+      "grad_norm": 1.8125640882474123,
+      "learning_rate": 9.869350522955921e-06,
+      "loss": 0.3807,
+      "step": 674
+    },
+    {
+      "epoch": 0.1009421265141319,
+      "grad_norm": 1.9369733002230116,
+      "learning_rate": 9.868799941187193e-06,
+      "loss": 0.5201,
+      "step": 675
+    },
+    {
+      "epoch": 0.10109167040526394,
+      "grad_norm": 1.5216959755972845,
+      "learning_rate": 9.868248217151075e-06,
+      "loss": 0.3624,
+      "step": 676
+    },
+    {
+      "epoch": 0.10124121429639599,
+      "grad_norm": 1.532054269025379,
+      "learning_rate": 9.867695350977009e-06,
+      "loss": 0.2738,
+      "step": 677
+    },
+    {
+      "epoch": 0.10139075818752805,
+      "grad_norm": 1.9725714032650388,
+      "learning_rate": 9.867141342794703e-06,
+      "loss": 0.5802,
+      "step": 678
+    },
+    {
+      "epoch": 0.10154030207866009,
+      "grad_norm": 1.9107978616944274,
+      "learning_rate": 9.86658619273413e-06,
+      "loss": 0.482,
+      "step": 679
+    },
+    {
+      "epoch": 0.10168984596979214,
+      "grad_norm": 1.919639496784501,
+      "learning_rate": 9.866029900925535e-06,
+      "loss": 0.3558,
+      "step": 680
+    },
+    {
+      "epoch": 0.10183938986092418,
+      "grad_norm": 2.174841069849439,
+      "learning_rate": 9.865472467499431e-06,
+      "loss": 0.6996,
+      "step": 681
+    },
+    {
+      "epoch": 0.10198893375205623,
+      "grad_norm": 2.2558702972279807,
+      "learning_rate": 9.864913892586596e-06,
+      "loss": 0.2397,
+      "step": 682
+    },
+    {
+      "epoch": 0.10213847764318827,
+      "grad_norm": 2.1196800969183105,
+      "learning_rate": 9.864354176318076e-06,
+      "loss": 0.3793,
+      "step": 683
+    },
+    {
+      "epoch": 0.10228802153432032,
+      "grad_norm": 2.165719475550091,
+      "learning_rate": 9.863793318825186e-06,
+      "loss": 0.2154,
+      "step": 684
+    },
+    {
+      "epoch": 0.10243756542545238,
+      "grad_norm": 1.7513134063770632,
+      "learning_rate": 9.86323132023951e-06,
+      "loss": 0.3816,
+      "step": 685
+    },
+    {
+      "epoch": 0.10258710931658442,
+      "grad_norm": 1.7103742255808732,
+      "learning_rate": 9.862668180692897e-06,
+      "loss": 0.2469,
+      "step": 686
+    },
+    {
+      "epoch": 0.10273665320771647,
+      "grad_norm": 1.9784764768939407,
+      "learning_rate": 9.862103900317467e-06,
+      "loss": 0.2279,
+      "step": 687
+    },
+    {
+      "epoch": 0.10288619709884851,
+      "grad_norm": 2.5430996153598877,
+      "learning_rate": 9.861538479245603e-06,
+      "loss": 0.4512,
+      "step": 688
+    },
+    {
+      "epoch": 0.10303574098998056,
+      "grad_norm": 1.3584315188319882,
+      "learning_rate": 9.86097191760996e-06,
+      "loss": 0.2521,
+      "step": 689
+    },
+    {
+      "epoch": 0.1031852848811126,
+      "grad_norm": 1.8041511333081743,
+      "learning_rate": 9.860404215543458e-06,
+      "loss": 0.3794,
+      "step": 690
+    },
+    {
+      "epoch": 0.10333482877224466,
+      "grad_norm": 2.261581805469511,
+      "learning_rate": 9.859835373179285e-06,
+      "loss": 0.5264,
+      "step": 691
+    },
+    {
+      "epoch": 0.1034843726633767,
+      "grad_norm": 1.4531049528328563,
+      "learning_rate": 9.859265390650897e-06,
+      "loss": 0.2069,
+      "step": 692
+    },
+    {
+      "epoch": 0.10363391655450875,
+      "grad_norm": 1.6530791454319427,
+      "learning_rate": 9.85869426809202e-06,
+      "loss": 0.2304,
+      "step": 693
+    },
+    {
+      "epoch": 0.1037834604456408,
+      "grad_norm": 1.5868398701857311,
+      "learning_rate": 9.85812200563664e-06,
+      "loss": 0.3894,
+      "step": 694
+    },
+    {
+      "epoch": 0.10393300433677284,
+      "grad_norm": 1.4690408418702507,
+      "learning_rate": 9.857548603419019e-06,
+      "loss": 0.3383,
+      "step": 695
+    },
+    {
+      "epoch": 0.10408254822790489,
+      "grad_norm": 1.868386725453433,
+      "learning_rate": 9.856974061573682e-06,
+      "loss": 0.4666,
+      "step": 696
+    },
+    {
+      "epoch": 0.10423209211903693,
+      "grad_norm": 1.3771017197315938,
+      "learning_rate": 9.856398380235422e-06,
+      "loss": 0.2285,
+      "step": 697
+    },
+    {
+      "epoch": 0.10438163601016899,
+      "grad_norm": 2.452990479638216,
+      "learning_rate": 9.855821559539298e-06,
+      "loss": 0.7219,
+      "step": 698
+    },
+    {
+      "epoch": 0.10453117990130104,
+      "grad_norm": 1.8742322224001207,
+      "learning_rate": 9.85524359962064e-06,
+      "loss": 0.4803,
+      "step": 699
+    },
+    {
+      "epoch": 0.10468072379243308,
+      "grad_norm": 1.858692042760981,
+      "learning_rate": 9.854664500615041e-06,
+      "loss": 0.2273,
+      "step": 700
+    },
+    {
+      "epoch": 0.10483026768356513,
+      "grad_norm": 1.1355721780236596,
+      "learning_rate": 9.854084262658365e-06,
+      "loss": 0.1947,
+      "step": 701
+    },
+    {
+      "epoch": 0.10497981157469717,
+      "grad_norm": 1.3464195395769243,
+      "learning_rate": 9.853502885886738e-06,
+      "loss": 0.1988,
+      "step": 702
+    },
+    {
+      "epoch": 0.10512935546582922,
+      "grad_norm": 1.204875080370136,
+      "learning_rate": 9.852920370436561e-06,
+      "loss": 0.3027,
+      "step": 703
+    },
+    {
+      "epoch": 0.10527889935696126,
+      "grad_norm": 1.3557124537174092,
+      "learning_rate": 9.852336716444496e-06,
+      "loss": 0.2158,
+      "step": 704
+    },
+    {
+      "epoch": 0.10542844324809332,
+      "grad_norm": 1.5752529363149261,
+      "learning_rate": 9.851751924047472e-06,
+      "loss": 0.3324,
+      "step": 705
+    },
+    {
+      "epoch": 0.10557798713922537,
+      "grad_norm": 1.7915590890665287,
+      "learning_rate": 9.85116599338269e-06,
+      "loss": 0.4936,
+      "step": 706
+    },
+    {
+      "epoch": 0.10572753103035741,
+      "grad_norm": 1.6842493918087815,
+      "learning_rate": 9.850578924587614e-06,
+      "loss": 0.4249,
+      "step": 707
+    },
+    {
+      "epoch": 0.10587707492148946,
+      "grad_norm": 1.781419189683173,
+      "learning_rate": 9.849990717799975e-06,
+      "loss": 0.2615,
+      "step": 708
+    },
+    {
+      "epoch": 0.1060266188126215,
+      "grad_norm": 1.4202393409091985,
+      "learning_rate": 9.849401373157772e-06,
+      "loss": 0.3256,
+      "step": 709
+    },
+    {
+      "epoch": 0.10617616270375355,
+      "grad_norm": 1.3714522045342281,
+      "learning_rate": 9.84881089079927e-06,
+      "loss": 0.219,
+      "step": 710
+    },
+    {
+      "epoch": 0.1063257065948856,
+      "grad_norm": 1.7391677942386203,
+      "learning_rate": 9.848219270863005e-06,
+      "loss": 0.2249,
+      "step": 711
+    },
+    {
+      "epoch": 0.10647525048601765,
+      "grad_norm": 1.3023890791191592,
+      "learning_rate": 9.847626513487774e-06,
+      "loss": 0.3693,
+      "step": 712
+    },
+    {
+      "epoch": 0.1066247943771497,
+      "grad_norm": 1.7969068078667318,
+      "learning_rate": 9.847032618812647e-06,
+      "loss": 0.2298,
+      "step": 713
+    },
+    {
+      "epoch": 0.10677433826828174,
+      "grad_norm": 2.102291030534645,
+      "learning_rate": 9.846437586976952e-06,
+      "loss": 0.4688,
+      "step": 714
+    },
+    {
+      "epoch": 0.10692388215941379,
+      "grad_norm": 1.072288463866959,
+      "learning_rate": 9.845841418120295e-06,
+      "loss": 0.2023,
+      "step": 715
+    },
+    {
+      "epoch": 0.10707342605054583,
+      "grad_norm": 1.3278088866624802,
+      "learning_rate": 9.845244112382536e-06,
+      "loss": 0.3492,
+      "step": 716
+    },
+    {
+      "epoch": 0.10722296994167788,
+      "grad_norm": 1.3771047197586663,
+      "learning_rate": 9.844645669903816e-06,
+      "loss": 0.2152,
+      "step": 717
+    },
+    {
+      "epoch": 0.10737251383280993,
+      "grad_norm": 1.243148446265919,
+      "learning_rate": 9.844046090824533e-06,
+      "loss": 0.2419,
+      "step": 718
+    },
+    {
+      "epoch": 0.10752205772394198,
+      "grad_norm": 1.3994827626329662,
+      "learning_rate": 9.843445375285351e-06,
+      "loss": 0.3578,
+      "step": 719
+    },
+    {
+      "epoch": 0.10767160161507403,
+      "grad_norm": 2.180600395588636,
+      "learning_rate": 9.842843523427207e-06,
+      "loss": 0.4159,
+      "step": 720
+    },
+    {
+      "epoch": 0.10782114550620607,
+      "grad_norm": 1.853639106134475,
+      "learning_rate": 9.842240535391301e-06,
+      "loss": 0.3929,
+      "step": 721
+    },
+    {
+      "epoch": 0.10797068939733812,
+      "grad_norm": 2.1662710209518306,
+      "learning_rate": 9.841636411319098e-06,
+      "loss": 0.298,
+      "step": 722
+    },
+    {
+      "epoch": 0.10812023328847016,
+      "grad_norm": 1.604340910092426,
+      "learning_rate": 9.841031151352332e-06,
+      "loss": 0.2175,
+      "step": 723
+    },
+    {
+      "epoch": 0.10826977717960222,
+      "grad_norm": 2.486345181702559,
+      "learning_rate": 9.840424755633002e-06,
+      "loss": 0.5179,
+      "step": 724
+    },
+    {
+      "epoch": 0.10841932107073426,
+      "grad_norm": 1.415864057650498,
+      "learning_rate": 9.83981722430338e-06,
+      "loss": 0.3539,
+      "step": 725
+    },
+    {
+      "epoch": 0.10856886496186631,
+      "grad_norm": 1.4949172725362427,
+      "learning_rate": 9.839208557505989e-06,
+      "loss": 0.382,
+      "step": 726
+    },
+    {
+      "epoch": 0.10871840885299835,
+      "grad_norm": 2.2920695398684576,
+      "learning_rate": 9.838598755383636e-06,
+      "loss": 0.5086,
+      "step": 727
+    },
+    {
+      "epoch": 0.1088679527441304,
+      "grad_norm": 1.6667328126020315,
+      "learning_rate": 9.837987818079382e-06,
+      "loss": 0.3736,
+      "step": 728
+    },
+    {
+      "epoch": 0.10901749663526245,
+      "grad_norm": 1.3398213843433537,
+      "learning_rate": 9.837375745736562e-06,
+      "loss": 0.1827,
+      "step": 729
+    },
+    {
+      "epoch": 0.10916704052639449,
+      "grad_norm": 1.9606925081810038,
+      "learning_rate": 9.83676253849877e-06,
+      "loss": 0.3992,
+      "step": 730
+    },
+    {
+      "epoch": 0.10931658441752655,
+      "grad_norm": 1.7774772468148792,
+      "learning_rate": 9.836148196509875e-06,
+      "loss": 0.4769,
+      "step": 731
+    },
+    {
+      "epoch": 0.1094661283086586,
+      "grad_norm": 1.9643639344581283,
+      "learning_rate": 9.835532719914005e-06,
+      "loss": 0.6049,
+      "step": 732
+    },
+    {
+      "epoch": 0.10961567219979064,
+      "grad_norm": 2.205274189695602,
+      "learning_rate": 9.834916108855557e-06,
+      "loss": 0.3679,
+      "step": 733
+    },
+    {
+      "epoch": 0.10976521609092268,
+      "grad_norm": 1.5293962633909763,
+      "learning_rate": 9.834298363479193e-06,
+      "loss": 0.3355,
+      "step": 734
+    },
+    {
+      "epoch": 0.10991475998205473,
+      "grad_norm": 1.985725165029928,
+      "learning_rate": 9.833679483929846e-06,
+      "loss": 0.357,
+      "step": 735
+    },
+    {
+      "epoch": 0.11006430387318678,
+      "grad_norm": 1.938156924090921,
+      "learning_rate": 9.833059470352705e-06,
+      "loss": 0.2667,
+      "step": 736
+    },
+    {
+      "epoch": 0.11021384776431883,
+      "grad_norm": 1.6208704990029696,
+      "learning_rate": 9.832438322893235e-06,
+      "loss": 0.2751,
+      "step": 737
+    },
+    {
+      "epoch": 0.11036339165545088,
+      "grad_norm": 1.6610841289834064,
+      "learning_rate": 9.831816041697164e-06,
+      "loss": 0.2319,
+      "step": 738
+    },
+    {
+      "epoch": 0.11051293554658292,
+      "grad_norm": 2.003195385581558,
+      "learning_rate": 9.831192626910482e-06,
+      "loss": 0.3799,
+      "step": 739
+    },
+    {
+      "epoch": 0.11066247943771497,
+      "grad_norm": 1.8861050802330894,
+      "learning_rate": 9.83056807867945e-06,
+      "loss": 0.4804,
+      "step": 740
+    },
+    {
+      "epoch": 0.11081202332884701,
+      "grad_norm": 1.6483143403386502,
+      "learning_rate": 9.829942397150593e-06,
+      "loss": 0.3658,
+      "step": 741
+    },
+    {
+      "epoch": 0.11096156721997906,
+      "grad_norm": 1.5438600790491723,
+      "learning_rate": 9.829315582470702e-06,
+      "loss": 0.2297,
+      "step": 742
+    },
+    {
+      "epoch": 0.1111111111111111,
+      "grad_norm": 1.4215916414139778,
+      "learning_rate": 9.828687634786834e-06,
+      "loss": 0.2365,
+      "step": 743
+    },
+    {
+      "epoch": 0.11126065500224316,
+      "grad_norm": 1.7761192463313074,
+      "learning_rate": 9.828058554246309e-06,
+      "loss": 0.3052,
+      "step": 744
+    },
+    {
+      "epoch": 0.11141019889337521,
+      "grad_norm": 1.6148872971257344,
+      "learning_rate": 9.82742834099672e-06,
+      "loss": 0.2199,
+      "step": 745
+    },
+    {
+      "epoch": 0.11155974278450725,
+      "grad_norm": 2.0162005563823646,
+      "learning_rate": 9.826796995185916e-06,
+      "loss": 0.3839,
+      "step": 746
+    },
+    {
+      "epoch": 0.1117092866756393,
+      "grad_norm": 1.4737452330787222,
+      "learning_rate": 9.826164516962022e-06,
+      "loss": 0.1869,
+      "step": 747
+    },
+    {
+      "epoch": 0.11185883056677134,
+      "grad_norm": 2.102437337141145,
+      "learning_rate": 9.82553090647342e-06,
+      "loss": 0.2615,
+      "step": 748
+    },
+    {
+      "epoch": 0.11200837445790339,
+      "grad_norm": 1.7050095093194846,
+      "learning_rate": 9.82489616386876e-06,
+      "loss": 0.2518,
+      "step": 749
+    },
+    {
+      "epoch": 0.11215791834903545,
+      "grad_norm": 1.5205595562607015,
+      "learning_rate": 9.824260289296963e-06,
+      "loss": 0.1792,
+      "step": 750
+    },
+    {
+      "epoch": 0.1123074622401675,
+      "grad_norm": 1.7480859507023934,
+      "learning_rate": 9.823623282907207e-06,
+      "loss": 0.4179,
+      "step": 751
+    },
+    {
+      "epoch": 0.11245700613129954,
+      "grad_norm": 1.5498394808983003,
+      "learning_rate": 9.822985144848944e-06,
+      "loss": 0.3358,
+      "step": 752
+    },
+    {
+      "epoch": 0.11260655002243158,
+      "grad_norm": 1.6393482758244988,
+      "learning_rate": 9.822345875271884e-06,
+      "loss": 0.2149,
+      "step": 753
+    },
+    {
+      "epoch": 0.11275609391356363,
+      "grad_norm": 1.9754923234031054,
+      "learning_rate": 9.821705474326006e-06,
+      "loss": 0.434,
+      "step": 754
+    },
+    {
+      "epoch": 0.11290563780469567,
+      "grad_norm": 2.2364190645343154,
+      "learning_rate": 9.821063942161558e-06,
+      "loss": 0.2228,
+      "step": 755
+    },
+    {
+      "epoch": 0.11305518169582772,
+      "grad_norm": 1.4307479990009164,
+      "learning_rate": 9.820421278929045e-06,
+      "loss": 0.3547,
+      "step": 756
+    },
+    {
+      "epoch": 0.11320472558695978,
+      "grad_norm": 1.2078809494224174,
+      "learning_rate": 9.819777484779242e-06,
+      "loss": 0.2245,
+      "step": 757
+    },
+    {
+      "epoch": 0.11335426947809182,
+      "grad_norm": 2.6305960032740354,
+      "learning_rate": 9.819132559863194e-06,
+      "loss": 0.6771,
+      "step": 758
+    },
+    {
+      "epoch": 0.11350381336922387,
+      "grad_norm": 1.4792675137281683,
+      "learning_rate": 9.818486504332203e-06,
+      "loss": 0.318,
+      "step": 759
+    },
+    {
+      "epoch": 0.11365335726035591,
+      "grad_norm": 2.127350110714495,
+      "learning_rate": 9.817839318337839e-06,
+      "loss": 0.4925,
+      "step": 760
+    },
+    {
+      "epoch": 0.11380290115148796,
+      "grad_norm": 1.8233415508114148,
+      "learning_rate": 9.81719100203194e-06,
+      "loss": 0.2747,
+      "step": 761
+    },
+    {
+      "epoch": 0.11395244504262,
+      "grad_norm": 2.1187219443475156,
+      "learning_rate": 9.81654155556661e-06,
+      "loss": 0.4595,
+      "step": 762
+    },
+    {
+      "epoch": 0.11410198893375205,
+      "grad_norm": 1.5759792753813915,
+      "learning_rate": 9.81589097909421e-06,
+      "loss": 0.3553,
+      "step": 763
+    },
+    {
+      "epoch": 0.11425153282488411,
+      "grad_norm": 1.708776908270571,
+      "learning_rate": 9.815239272767373e-06,
+      "loss": 0.3091,
+      "step": 764
+    },
+    {
+      "epoch": 0.11440107671601615,
+      "grad_norm": 1.703775430420963,
+      "learning_rate": 9.814586436738998e-06,
+      "loss": 0.3728,
+      "step": 765
+    },
+    {
+      "epoch": 0.1145506206071482,
+      "grad_norm": 1.6198262441466886,
+      "learning_rate": 9.813932471162245e-06,
+      "loss": 0.2498,
+      "step": 766
+    },
+    {
+      "epoch": 0.11470016449828024,
+      "grad_norm": 1.4858642435718663,
+      "learning_rate": 9.813277376190539e-06,
+      "loss": 0.2299,
+      "step": 767
+    },
+    {
+      "epoch": 0.11484970838941229,
+      "grad_norm": 1.8052387881768808,
+      "learning_rate": 9.812621151977574e-06,
+      "loss": 0.3834,
+      "step": 768
+    },
+    {
+      "epoch": 0.11499925228054433,
+      "grad_norm": 1.7902664470941898,
+      "learning_rate": 9.811963798677306e-06,
+      "loss": 0.2282,
+      "step": 769
+    },
+    {
+      "epoch": 0.11514879617167639,
+      "grad_norm": 1.6316784429425562,
+      "learning_rate": 9.811305316443956e-06,
+      "loss": 0.2396,
+      "step": 770
+    },
+    {
+      "epoch": 0.11529834006280844,
+      "grad_norm": 1.414088700773603,
+      "learning_rate": 9.81064570543201e-06,
+      "loss": 0.2353,
+      "step": 771
+    },
+    {
+      "epoch": 0.11544788395394048,
+      "grad_norm": 1.9219176628835946,
+      "learning_rate": 9.80998496579622e-06,
+      "loss": 0.3379,
+      "step": 772
+    },
+    {
+      "epoch": 0.11559742784507253,
+      "grad_norm": 1.113085528787142,
+      "learning_rate": 9.809323097691602e-06,
+      "loss": 0.213,
+      "step": 773
+    },
+    {
+      "epoch": 0.11574697173620457,
+      "grad_norm": 1.6091723360768109,
+      "learning_rate": 9.808660101273435e-06,
+      "loss": 0.3457,
+      "step": 774
+    },
+    {
+      "epoch": 0.11589651562733662,
+      "grad_norm": 1.4626018681006754,
+      "learning_rate": 9.807995976697267e-06,
+      "loss": 0.1777,
+      "step": 775
+    },
+    {
+      "epoch": 0.11604605951846866,
+      "grad_norm": 2.3953869334660522,
+      "learning_rate": 9.807330724118906e-06,
+      "loss": 0.449,
+      "step": 776
+    },
+    {
+      "epoch": 0.11619560340960072,
+      "grad_norm": 1.595059614098865,
+      "learning_rate": 9.806664343694425e-06,
+      "loss": 0.3367,
+      "step": 777
+    },
+    {
+      "epoch": 0.11634514730073277,
+      "grad_norm": 1.1067814732956414,
+      "learning_rate": 9.805996835580169e-06,
+      "loss": 0.1828,
+      "step": 778
+    },
+    {
+      "epoch": 0.11649469119186481,
+      "grad_norm": 1.62643731894747,
+      "learning_rate": 9.805328199932736e-06,
+      "loss": 0.2005,
+      "step": 779
+    },
+    {
+      "epoch": 0.11664423508299686,
+      "grad_norm": 1.5535907543030336,
+      "learning_rate": 9.804658436908996e-06,
+      "loss": 0.2635,
+      "step": 780
+    },
+    {
+      "epoch": 0.1167937789741289,
+      "grad_norm": 1.0891099881486959,
+      "learning_rate": 9.803987546666083e-06,
+      "loss": 0.2012,
+      "step": 781
+    },
+    {
+      "epoch": 0.11694332286526095,
+      "grad_norm": 1.9927493555965012,
+      "learning_rate": 9.803315529361395e-06,
+      "loss": 0.5297,
+      "step": 782
+    },
+    {
+      "epoch": 0.11709286675639301,
+      "grad_norm": 1.6333695435696685,
+      "learning_rate": 9.802642385152593e-06,
+      "loss": 0.2959,
+      "step": 783
+    },
+    {
+      "epoch": 0.11724241064752505,
+      "grad_norm": 2.0905230122366896,
+      "learning_rate": 9.8019681141976e-06,
+      "loss": 0.3662,
+      "step": 784
+    },
+    {
+      "epoch": 0.1173919545386571,
+      "grad_norm": 1.3626106000772258,
+      "learning_rate": 9.80129271665461e-06,
+      "loss": 0.2065,
+      "step": 785
+    },
+    {
+      "epoch": 0.11754149842978914,
+      "grad_norm": 1.7357589724302078,
+      "learning_rate": 9.800616192682077e-06,
+      "loss": 0.3269,
+      "step": 786
+    },
+    {
+      "epoch": 0.11769104232092119,
+      "grad_norm": 1.3095611640218061,
+      "learning_rate": 9.79993854243872e-06,
+      "loss": 0.1993,
+      "step": 787
+    },
+    {
+      "epoch": 0.11784058621205323,
+      "grad_norm": 2.008966146034465,
+      "learning_rate": 9.799259766083522e-06,
+      "loss": 0.2346,
+      "step": 788
+    },
+    {
+      "epoch": 0.11799013010318528,
+      "grad_norm": 1.4166616715548845,
+      "learning_rate": 9.798579863775733e-06,
+      "loss": 0.2053,
+      "step": 789
+    },
+    {
+      "epoch": 0.11813967399431734,
+      "grad_norm": 2.2231308106975742,
+      "learning_rate": 9.79789883567486e-06,
+      "loss": 0.3138,
+      "step": 790
+    },
+    {
+      "epoch": 0.11828921788544938,
+      "grad_norm": 2.104337845030462,
+      "learning_rate": 9.79721668194068e-06,
+      "loss": 0.5896,
+      "step": 791
+    },
+    {
+      "epoch": 0.11843876177658143,
+      "grad_norm": 1.3961164790172769,
+      "learning_rate": 9.796533402733235e-06,
+      "loss": 0.2023,
+      "step": 792
+    },
+    {
+      "epoch": 0.11858830566771347,
+      "grad_norm": 1.9644410315987328,
+      "learning_rate": 9.79584899821283e-06,
+      "loss": 0.2389,
+      "step": 793
+    },
+    {
+      "epoch": 0.11873784955884552,
+      "grad_norm": 1.911243493993425,
+      "learning_rate": 9.795163468540028e-06,
+      "loss": 0.2319,
+      "step": 794
+    },
+    {
+      "epoch": 0.11888739344997756,
+      "grad_norm": 1.5442762653540485,
+      "learning_rate": 9.794476813875665e-06,
+      "loss": 0.256,
+      "step": 795
+    },
+    {
+      "epoch": 0.11903693734110962,
+      "grad_norm": 1.6347185262551664,
+      "learning_rate": 9.793789034380833e-06,
+      "loss": 0.3659,
+      "step": 796
+    },
+    {
+      "epoch": 0.11918648123224167,
+      "grad_norm": 6.93121052791821,
+      "learning_rate": 9.793100130216895e-06,
+      "loss": 0.3348,
+      "step": 797
+    },
+    {
+      "epoch": 0.11933602512337371,
+      "grad_norm": 1.1914924746720745,
+      "learning_rate": 9.792410101545475e-06,
+      "loss": 0.2475,
+      "step": 798
+    },
+    {
+      "epoch": 0.11948556901450576,
+      "grad_norm": 1.4728413245474197,
+      "learning_rate": 9.791718948528457e-06,
+      "loss": 0.3569,
+      "step": 799
+    },
+    {
+      "epoch": 0.1196351129056378,
+      "grad_norm": 2.0173892018585113,
+      "learning_rate": 9.791026671327996e-06,
+      "loss": 0.3154,
+      "step": 800
+    },
+    {
+      "epoch": 0.11978465679676985,
+      "grad_norm": 1.863844432530015,
+      "learning_rate": 9.790333270106505e-06,
+      "loss": 0.349,
+      "step": 801
+    },
+    {
+      "epoch": 0.11993420068790189,
+      "grad_norm": 1.348620907056274,
+      "learning_rate": 9.789638745026661e-06,
+      "loss": 0.2553,
+      "step": 802
+    },
+    {
+      "epoch": 0.12008374457903395,
+      "grad_norm": 1.7207895705367349,
+      "learning_rate": 9.78894309625141e-06,
+      "loss": 0.3931,
+      "step": 803
+    },
+    {
+      "epoch": 0.120233288470166,
+      "grad_norm": 1.3131471894535065,
+      "learning_rate": 9.788246323943954e-06,
+      "loss": 0.1473,
+      "step": 804
+    },
+    {
+      "epoch": 0.12038283236129804,
+      "grad_norm": 1.054327043113717,
+      "learning_rate": 9.787548428267766e-06,
+      "loss": 0.1945,
+      "step": 805
+    },
+    {
+      "epoch": 0.12053237625243009,
+      "grad_norm": 1.9526157668969721,
+      "learning_rate": 9.786849409386577e-06,
+      "loss": 0.3906,
+      "step": 806
+    },
+    {
+      "epoch": 0.12068192014356213,
+      "grad_norm": 1.8267497285704608,
+      "learning_rate": 9.786149267464382e-06,
+      "loss": 0.4193,
+      "step": 807
+    },
+    {
+      "epoch": 0.12083146403469418,
+      "grad_norm": 0.8038770137897923,
+      "learning_rate": 9.785448002665446e-06,
+      "loss": 0.2392,
+      "step": 808
+    },
+    {
+      "epoch": 0.12098100792582624,
+      "grad_norm": 1.5253624272396114,
+      "learning_rate": 9.784745615154286e-06,
+      "loss": 0.3366,
+      "step": 809
+    },
+    {
+      "epoch": 0.12113055181695828,
+      "grad_norm": 2.093861559130543,
+      "learning_rate": 9.784042105095694e-06,
+      "loss": 0.4947,
+      "step": 810
+    },
+    {
+      "epoch": 0.12128009570809033,
+      "grad_norm": 1.4395999315986885,
+      "learning_rate": 9.78333747265472e-06,
+      "loss": 0.3721,
+      "step": 811
+    },
+    {
+      "epoch": 0.12142963959922237,
+      "grad_norm": 1.852326670020495,
+      "learning_rate": 9.782631717996675e-06,
+      "loss": 0.4779,
+      "step": 812
+    },
+    {
+      "epoch": 0.12157918349035442,
+      "grad_norm": 1.6061911802246367,
+      "learning_rate": 9.781924841287136e-06,
+      "loss": 0.3634,
+      "step": 813
+    },
+    {
+      "epoch": 0.12172872738148646,
+      "grad_norm": 1.6480313202927959,
+      "learning_rate": 9.781216842691945e-06,
+      "loss": 0.3486,
+      "step": 814
+    },
+    {
+      "epoch": 0.12187827127261851,
+      "grad_norm": 1.7352908193060639,
+      "learning_rate": 9.780507722377205e-06,
+      "loss": 0.2405,
+      "step": 815
+    },
+    {
+      "epoch": 0.12202781516375057,
+      "grad_norm": 1.819605505850209,
+      "learning_rate": 9.779797480509281e-06,
+      "loss": 0.2702,
+      "step": 816
+    },
+    {
+      "epoch": 0.12217735905488261,
+      "grad_norm": 5.030925574416197,
+      "learning_rate": 9.779086117254804e-06,
+      "loss": 0.4802,
+      "step": 817
+    },
+    {
+      "epoch": 0.12232690294601466,
+      "grad_norm": 1.4714728946101239,
+      "learning_rate": 9.778373632780665e-06,
+      "loss": 0.4002,
+      "step": 818
+    },
+    {
+      "epoch": 0.1224764468371467,
+      "grad_norm": 1.8961195589940971,
+      "learning_rate": 9.777660027254022e-06,
+      "loss": 0.5022,
+      "step": 819
+    },
+    {
+      "epoch": 0.12262599072827875,
+      "grad_norm": 2.1718036722230343,
+      "learning_rate": 9.776945300842292e-06,
+      "loss": 0.3274,
+      "step": 820
+    },
+    {
+      "epoch": 0.12277553461941079,
+      "grad_norm": 1.620505499680087,
+      "learning_rate": 9.776229453713158e-06,
+      "loss": 0.2316,
+      "step": 821
+    },
+    {
+      "epoch": 0.12292507851054285,
+      "grad_norm": 1.6978035349883904,
+      "learning_rate": 9.775512486034564e-06,
+      "loss": 0.3388,
+      "step": 822
+    },
+    {
+      "epoch": 0.1230746224016749,
+      "grad_norm": 1.8097210824887537,
+      "learning_rate": 9.774794397974715e-06,
+      "loss": 0.2658,
+      "step": 823
+    },
+    {
+      "epoch": 0.12322416629280694,
+      "grad_norm": 1.7832381045534218,
+      "learning_rate": 9.774075189702085e-06,
+      "loss": 0.236,
+      "step": 824
+    },
+    {
+      "epoch": 0.12337371018393899,
+      "grad_norm": 1.5481034516154306,
+      "learning_rate": 9.773354861385408e-06,
+      "loss": 0.2209,
+      "step": 825
+    },
+    {
+      "epoch": 0.12352325407507103,
+      "grad_norm": 1.808929914702085,
+      "learning_rate": 9.772633413193677e-06,
+      "loss": 0.3936,
+      "step": 826
+    },
+    {
+      "epoch": 0.12367279796620308,
+      "grad_norm": 1.4632324741175244,
+      "learning_rate": 9.771910845296151e-06,
+      "loss": 0.1809,
+      "step": 827
+    },
+    {
+      "epoch": 0.12382234185733512,
+      "grad_norm": 1.602480536861921,
+      "learning_rate": 9.771187157862352e-06,
+      "loss": 0.3631,
+      "step": 828
+    },
+    {
+      "epoch": 0.12397188574846718,
+      "grad_norm": 1.695314807275002,
+      "learning_rate": 9.770462351062065e-06,
+      "loss": 0.3419,
+      "step": 829
+    },
+    {
+      "epoch": 0.12412142963959923,
+      "grad_norm": 2.2542289621861262,
+      "learning_rate": 9.769736425065333e-06,
+      "loss": 0.4292,
+      "step": 830
+    },
+    {
+      "epoch": 0.12427097353073127,
+      "grad_norm": 1.7697982755032058,
+      "learning_rate": 9.76900938004247e-06,
+      "loss": 0.3735,
+      "step": 831
+    },
+    {
+      "epoch": 0.12442051742186332,
+      "grad_norm": 1.8120887881814032,
+      "learning_rate": 9.768281216164045e-06,
+      "loss": 0.3568,
+      "step": 832
+    },
+    {
+      "epoch": 0.12457006131299536,
+      "grad_norm": 1.581710048140236,
+      "learning_rate": 9.767551933600896e-06,
+      "loss": 0.1999,
+      "step": 833
+    },
+    {
+      "epoch": 0.1247196052041274,
+      "grad_norm": 1.8611636134135094,
+      "learning_rate": 9.766821532524113e-06,
+      "loss": 0.4111,
+      "step": 834
+    },
+    {
+      "epoch": 0.12486914909525945,
+      "grad_norm": 1.3596930734620556,
+      "learning_rate": 9.76609001310506e-06,
+      "loss": 0.1893,
+      "step": 835
+    },
+    {
+      "epoch": 0.1250186929863915,
+      "grad_norm": 1.7562268713789106,
+      "learning_rate": 9.76535737551536e-06,
+      "loss": 0.3948,
+      "step": 836
+    },
+    {
+      "epoch": 0.12516823687752354,
+      "grad_norm": 2.0138359003020136,
+      "learning_rate": 9.764623619926891e-06,
+      "loss": 0.2182,
+      "step": 837
+    },
+    {
+      "epoch": 0.1253177807686556,
+      "grad_norm": 1.447071144741178,
+      "learning_rate": 9.763888746511804e-06,
+      "loss": 0.2027,
+      "step": 838
+    },
+    {
+      "epoch": 0.12546732465978766,
+      "grad_norm": 2.1652568729944734,
+      "learning_rate": 9.763152755442504e-06,
+      "loss": 0.6314,
+      "step": 839
+    },
+    {
+      "epoch": 0.1256168685509197,
+      "grad_norm": 1.8038679466625127,
+      "learning_rate": 9.762415646891665e-06,
+      "loss": 0.3578,
+      "step": 840
+    },
+    {
+      "epoch": 0.12576641244205175,
+      "grad_norm": 1.6058605347195138,
+      "learning_rate": 9.761677421032218e-06,
+      "loss": 0.4411,
+      "step": 841
+    },
+    {
+      "epoch": 0.12591595633318378,
+      "grad_norm": 1.590419871100753,
+      "learning_rate": 9.760938078037358e-06,
+      "loss": 0.3562,
+      "step": 842
+    },
+    {
+      "epoch": 0.12606550022431584,
+      "grad_norm": 2.1851801492136267,
+      "learning_rate": 9.76019761808054e-06,
+      "loss": 0.5822,
+      "step": 843
+    },
+    {
+      "epoch": 0.12621504411544787,
+      "grad_norm": 1.5855983503039581,
+      "learning_rate": 9.759456041335487e-06,
+      "loss": 0.2229,
+      "step": 844
+    },
+    {
+      "epoch": 0.12636458800657993,
+      "grad_norm": 1.370525319712137,
+      "learning_rate": 9.758713347976179e-06,
+      "loss": 0.2126,
+      "step": 845
+    },
+    {
+      "epoch": 0.126514131897712,
+      "grad_norm": 6.394283226949693,
+      "learning_rate": 9.757969538176856e-06,
+      "loss": 0.5925,
+      "step": 846
+    },
+    {
+      "epoch": 0.12666367578884402,
+      "grad_norm": 1.6599084242802136,
+      "learning_rate": 9.757224612112026e-06,
+      "loss": 0.2939,
+      "step": 847
+    },
+    {
+      "epoch": 0.12681321967997608,
+      "grad_norm": 1.7907787465869436,
+      "learning_rate": 9.756478569956455e-06,
+      "loss": 0.222,
+      "step": 848
+    },
+    {
+      "epoch": 0.1269627635711081,
+      "grad_norm": 2.2366122735755707,
+      "learning_rate": 9.755731411885172e-06,
+      "loss": 0.6684,
+      "step": 849
+    },
+    {
+      "epoch": 0.12711230746224017,
+      "grad_norm": 1.2342377688986181,
+      "learning_rate": 9.754983138073466e-06,
+      "loss": 0.1731,
+      "step": 850
+    },
+    {
+      "epoch": 0.1272618513533722,
+      "grad_norm": 1.890953555602396,
+      "learning_rate": 9.75423374869689e-06,
+      "loss": 0.3518,
+      "step": 851
+    },
+    {
+      "epoch": 0.12741139524450426,
+      "grad_norm": 1.6475639435427634,
+      "learning_rate": 9.75348324393126e-06,
+      "loss": 0.2398,
+      "step": 852
+    },
+    {
+      "epoch": 0.12756093913563632,
+      "grad_norm": 1.6270554683226957,
+      "learning_rate": 9.752731623952647e-06,
+      "loss": 0.4891,
+      "step": 853
+    },
+    {
+      "epoch": 0.12771048302676835,
+      "grad_norm": 1.26855312080081,
+      "learning_rate": 9.751978888937394e-06,
+      "loss": 0.256,
+      "step": 854
+    },
+    {
+      "epoch": 0.1278600269179004,
+      "grad_norm": 1.8185769247015318,
+      "learning_rate": 9.751225039062096e-06,
+      "loss": 0.4165,
+      "step": 855
+    },
+    {
+      "epoch": 0.12800957080903244,
+      "grad_norm": 1.751422967721224,
+      "learning_rate": 9.750470074503616e-06,
+      "loss": 0.4006,
+      "step": 856
+    },
+    {
+      "epoch": 0.1281591147001645,
+      "grad_norm": 1.520424463551602,
+      "learning_rate": 9.749713995439072e-06,
+      "loss": 0.221,
+      "step": 857
+    },
+    {
+      "epoch": 0.12830865859129656,
+      "grad_norm": 1.3721869164602227,
+      "learning_rate": 9.74895680204585e-06,
+      "loss": 0.2902,
+      "step": 858
+    },
+    {
+      "epoch": 0.1284582024824286,
+      "grad_norm": 0.9768480150555632,
+      "learning_rate": 9.748198494501598e-06,
+      "loss": 0.2115,
+      "step": 859
+    },
+    {
+      "epoch": 0.12860774637356065,
+      "grad_norm": 1.1411458146693625,
+      "learning_rate": 9.747439072984217e-06,
+      "loss": 0.2657,
+      "step": 860
+    },
+    {
+      "epoch": 0.12875729026469268,
+      "grad_norm": 1.351877109756975,
+      "learning_rate": 9.746678537671876e-06,
+      "loss": 0.1998,
+      "step": 861
+    },
+    {
+      "epoch": 0.12890683415582474,
+      "grad_norm": 1.735519954859839,
+      "learning_rate": 9.745916888743006e-06,
+      "loss": 0.3916,
+      "step": 862
+    },
+    {
+      "epoch": 0.12905637804695677,
+      "grad_norm": 1.5843585668319187,
+      "learning_rate": 9.745154126376295e-06,
+      "loss": 0.2412,
+      "step": 863
+    },
+    {
+      "epoch": 0.12920592193808883,
+      "grad_norm": 1.2682977552751018,
+      "learning_rate": 9.744390250750694e-06,
+      "loss": 0.2082,
+      "step": 864
+    },
+    {
+      "epoch": 0.1293554658292209,
+      "grad_norm": 1.972644277101951,
+      "learning_rate": 9.74362526204542e-06,
+      "loss": 0.5327,
+      "step": 865
+    },
+    {
+      "epoch": 0.12950500972035292,
+      "grad_norm": 1.9379037313358354,
+      "learning_rate": 9.74285916043994e-06,
+      "loss": 0.5184,
+      "step": 866
+    },
+    {
+      "epoch": 0.12965455361148498,
+      "grad_norm": 2.118494372996469,
+      "learning_rate": 9.742091946113994e-06,
+      "loss": 0.4367,
+      "step": 867
+    },
+    {
+      "epoch": 0.129804097502617,
+      "grad_norm": 3.0042960877566904,
+      "learning_rate": 9.741323619247575e-06,
+      "loss": 0.2971,
+      "step": 868
+    },
+    {
+      "epoch": 0.12995364139374907,
+      "grad_norm": 1.4822743195387478,
+      "learning_rate": 9.740554180020944e-06,
+      "loss": 0.2324,
+      "step": 869
+    },
+    {
+      "epoch": 0.1301031852848811,
+      "grad_norm": 1.629265135056018,
+      "learning_rate": 9.739783628614614e-06,
+      "loss": 0.3717,
+      "step": 870
+    },
+    {
+      "epoch": 0.13025272917601316,
+      "grad_norm": 2.663192450334603,
+      "learning_rate": 9.739011965209366e-06,
+      "loss": 0.4405,
+      "step": 871
+    },
+    {
+      "epoch": 0.13040227306714522,
+      "grad_norm": 1.942071044818614,
+      "learning_rate": 9.738239189986239e-06,
+      "loss": 0.2141,
+      "step": 872
+    },
+    {
+      "epoch": 0.13055181695827725,
+      "grad_norm": 3.5555352510097684,
+      "learning_rate": 9.737465303126533e-06,
+      "loss": 0.5084,
+      "step": 873
+    },
+    {
+      "epoch": 0.1307013608494093,
+      "grad_norm": 131.61785973638575,
+      "learning_rate": 9.736690304811811e-06,
+      "loss": 0.2431,
+      "step": 874
+    },
+    {
+      "epoch": 0.13085090474054134,
+      "grad_norm": 1.190310223040302,
+      "learning_rate": 9.735914195223894e-06,
+      "loss": 0.1586,
+      "step": 875
+    },
+    {
+      "epoch": 0.1310004486316734,
+      "grad_norm": 1.8346983968963104,
+      "learning_rate": 9.735136974544866e-06,
+      "loss": 0.5247,
+      "step": 876
+    },
+    {
+      "epoch": 0.13114999252280543,
+      "grad_norm": 1.7905067752668935,
+      "learning_rate": 9.734358642957068e-06,
+      "loss": 0.2645,
+      "step": 877
+    },
+    {
+      "epoch": 0.1312995364139375,
+      "grad_norm": 1.9545038391601572,
+      "learning_rate": 9.733579200643108e-06,
+      "loss": 0.3769,
+      "step": 878
+    },
+    {
+      "epoch": 0.13144908030506955,
+      "grad_norm": 2.1943279379956477,
+      "learning_rate": 9.732798647785847e-06,
+      "loss": 0.5142,
+      "step": 879
+    },
+    {
+      "epoch": 0.13159862419620158,
+      "grad_norm": 1.8055649624971999,
+      "learning_rate": 9.73201698456841e-06,
+      "loss": 0.1857,
+      "step": 880
+    },
+    {
+      "epoch": 0.13174816808733364,
+      "grad_norm": 1.8149442634221358,
+      "learning_rate": 9.731234211174188e-06,
+      "loss": 0.2233,
+      "step": 881
+    },
+    {
+      "epoch": 0.13189771197846567,
+      "grad_norm": 2.080170101944024,
+      "learning_rate": 9.73045032778682e-06,
+      "loss": 0.3904,
+      "step": 882
+    },
+    {
+      "epoch": 0.13204725586959773,
+      "grad_norm": 1.9510038015167501,
+      "learning_rate": 9.729665334590217e-06,
+      "loss": 0.3821,
+      "step": 883
+    },
+    {
+      "epoch": 0.13219679976072976,
+      "grad_norm": 2.1650257376887545,
+      "learning_rate": 9.728879231768547e-06,
+      "loss": 0.2357,
+      "step": 884
+    },
+    {
+      "epoch": 0.13234634365186182,
+      "grad_norm": 1.1711773735000268,
+      "learning_rate": 9.728092019506233e-06,
+      "loss": 0.181,
+      "step": 885
+    },
+    {
+      "epoch": 0.13249588754299388,
+      "grad_norm": 1.3626230838557185,
+      "learning_rate": 9.727303697987965e-06,
+      "loss": 0.2283,
+      "step": 886
+    },
+    {
+      "epoch": 0.1326454314341259,
+      "grad_norm": 1.9510781854659551,
+      "learning_rate": 9.72651426739869e-06,
+      "loss": 0.5154,
+      "step": 887
+    },
+    {
+      "epoch": 0.13279497532525797,
+      "grad_norm": 0.9715289377560907,
+      "learning_rate": 9.72572372792362e-06,
+      "loss": 0.248,
+      "step": 888
+    },
+    {
+      "epoch": 0.13294451921639,
+      "grad_norm": 2.146774938769699,
+      "learning_rate": 9.724932079748218e-06,
+      "loss": 0.5735,
+      "step": 889
+    },
+    {
+      "epoch": 0.13309406310752206,
+      "grad_norm": 1.9362171525305378,
+      "learning_rate": 9.724139323058213e-06,
+      "loss": 0.4643,
+      "step": 890
+    },
+    {
+      "epoch": 0.13324360699865412,
+      "grad_norm": 1.2101289634017103,
+      "learning_rate": 9.723345458039595e-06,
+      "loss": 0.2266,
+      "step": 891
+    },
+    {
+      "epoch": 0.13339315088978615,
+      "grad_norm": 1.5975435725539564,
+      "learning_rate": 9.722550484878612e-06,
+      "loss": 0.2212,
+      "step": 892
+    },
+    {
+      "epoch": 0.1335426947809182,
+      "grad_norm": 2.1131453338181783,
+      "learning_rate": 9.721754403761773e-06,
+      "loss": 0.5017,
+      "step": 893
+    },
+    {
+      "epoch": 0.13369223867205024,
+      "grad_norm": 1.8229015728915987,
+      "learning_rate": 9.720957214875846e-06,
+      "loss": 0.3833,
+      "step": 894
+    },
+    {
+      "epoch": 0.1338417825631823,
+      "grad_norm": 2.2277630697934416,
+      "learning_rate": 9.720158918407859e-06,
+      "loss": 0.2482,
+      "step": 895
+    },
+    {
+      "epoch": 0.13399132645431433,
+      "grad_norm": 2.7493650830526413,
+      "learning_rate": 9.719359514545097e-06,
+      "loss": 0.3227,
+      "step": 896
+    },
+    {
+      "epoch": 0.1341408703454464,
+      "grad_norm": 1.326522393176365,
+      "learning_rate": 9.718559003475114e-06,
+      "loss": 0.3694,
+      "step": 897
+    },
+    {
+      "epoch": 0.13429041423657845,
+      "grad_norm": 2.031851235687534,
+      "learning_rate": 9.717757385385713e-06,
+      "loss": 0.4018,
+      "step": 898
+    },
+    {
+      "epoch": 0.13443995812771048,
+      "grad_norm": 1.8621806446966467,
+      "learning_rate": 9.716954660464962e-06,
+      "loss": 0.4906,
+      "step": 899
+    },
+    {
+      "epoch": 0.13458950201884254,
+      "grad_norm": 1.2770259017691152,
+      "learning_rate": 9.716150828901189e-06,
+      "loss": 0.1858,
+      "step": 900
+    },
+    {
+      "epoch": 0.13473904590997457,
+      "grad_norm": 1.5913545450304718,
+      "learning_rate": 9.71534589088298e-06,
+      "loss": 0.197,
+      "step": 901
+    },
+    {
+      "epoch": 0.13488858980110663,
+      "grad_norm": 1.2063585747949084,
+      "learning_rate": 9.714539846599183e-06,
+      "loss": 0.2366,
+      "step": 902
+    },
+    {
+      "epoch": 0.13503813369223866,
+      "grad_norm": 2.5200659605885036,
+      "learning_rate": 9.713732696238901e-06,
+      "loss": 0.5999,
+      "step": 903
+    },
+    {
+      "epoch": 0.13518767758337072,
+      "grad_norm": 1.9024340551353864,
+      "learning_rate": 9.7129244399915e-06,
+      "loss": 0.3618,
+      "step": 904
+    },
+    {
+      "epoch": 0.13533722147450278,
+      "grad_norm": 2.069806998096416,
+      "learning_rate": 9.712115078046606e-06,
+      "loss": 0.2392,
+      "step": 905
+    },
+    {
+      "epoch": 0.1354867653656348,
+      "grad_norm": 1.9190350188819136,
+      "learning_rate": 9.711304610594104e-06,
+      "loss": 0.4096,
+      "step": 906
+    },
+    {
+      "epoch": 0.13563630925676687,
+      "grad_norm": 1.0348199460008871,
+      "learning_rate": 9.710493037824133e-06,
+      "loss": 0.2071,
+      "step": 907
+    },
+    {
+      "epoch": 0.1357858531478989,
+      "grad_norm": 2.014537240536291,
+      "learning_rate": 9.709680359927101e-06,
+      "loss": 0.4374,
+      "step": 908
+    },
+    {
+      "epoch": 0.13593539703903096,
+      "grad_norm": 1.7241079783150106,
+      "learning_rate": 9.708866577093665e-06,
+      "loss": 0.4161,
+      "step": 909
+    },
+    {
+      "epoch": 0.136084940930163,
+      "grad_norm": 1.3280448342419884,
+      "learning_rate": 9.70805168951475e-06,
+      "loss": 0.1967,
+      "step": 910
+    },
+    {
+      "epoch": 0.13623448482129505,
+      "grad_norm": 1.6364992809413539,
+      "learning_rate": 9.707235697381536e-06,
+      "loss": 0.3394,
+      "step": 911
+    },
+    {
+      "epoch": 0.1363840287124271,
+      "grad_norm": 1.1211253304635729,
+      "learning_rate": 9.706418600885462e-06,
+      "loss": 0.3542,
+      "step": 912
+    },
+    {
+      "epoch": 0.13653357260355914,
+      "grad_norm": 4.417441150249539,
+      "learning_rate": 9.705600400218227e-06,
+      "loss": 0.2605,
+      "step": 913
+    },
+    {
+      "epoch": 0.1366831164946912,
+      "grad_norm": 1.6849430545358892,
+      "learning_rate": 9.704781095571788e-06,
+      "loss": 0.3434,
+      "step": 914
+    },
+    {
+      "epoch": 0.13683266038582323,
+      "grad_norm": 1.0419590452262997,
+      "learning_rate": 9.703960687138363e-06,
+      "loss": 0.1759,
+      "step": 915
+    },
+    {
+      "epoch": 0.1369822042769553,
+      "grad_norm": 2.3605687929632286,
+      "learning_rate": 9.703139175110425e-06,
+      "loss": 0.6175,
+      "step": 916
+    },
+    {
+      "epoch": 0.13713174816808735,
+      "grad_norm": 1.6017722806543409,
+      "learning_rate": 9.702316559680714e-06,
+      "loss": 0.2687,
+      "step": 917
+    },
+    {
+      "epoch": 0.13728129205921938,
+      "grad_norm": 1.3442020598442603,
+      "learning_rate": 9.701492841042217e-06,
+      "loss": 0.3801,
+      "step": 918
+    },
+    {
+      "epoch": 0.13743083595035144,
+      "grad_norm": 1.6758219573938795,
+      "learning_rate": 9.70066801938819e-06,
+      "loss": 0.3869,
+      "step": 919
+    },
+    {
+      "epoch": 0.13758037984148347,
+      "grad_norm": 1.787015495544375,
+      "learning_rate": 9.699842094912146e-06,
+      "loss": 0.319,
+      "step": 920
+    },
+    {
+      "epoch": 0.13772992373261553,
+      "grad_norm": 1.8015526879631494,
+      "learning_rate": 9.699015067807851e-06,
+      "loss": 0.493,
+      "step": 921
+    },
+    {
+      "epoch": 0.13787946762374756,
+      "grad_norm": 0.9887387021286004,
+      "learning_rate": 9.698186938269334e-06,
+      "loss": 0.1724,
+      "step": 922
+    },
+    {
+      "epoch": 0.13802901151487962,
+      "grad_norm": 1.7253102296559673,
+      "learning_rate": 9.697357706490885e-06,
+      "loss": 0.5363,
+      "step": 923
+    },
+    {
+      "epoch": 0.13817855540601168,
+      "grad_norm": 1.5558864293295054,
+      "learning_rate": 9.696527372667046e-06,
+      "loss": 0.2863,
+      "step": 924
+    },
+    {
+      "epoch": 0.1383280992971437,
+      "grad_norm": 1.8210322672031793,
+      "learning_rate": 9.695695936992624e-06,
+      "loss": 0.4107,
+      "step": 925
+    },
+    {
+      "epoch": 0.13847764318827577,
+      "grad_norm": 1.6117992353983686,
+      "learning_rate": 9.69486339966268e-06,
+      "loss": 0.2162,
+      "step": 926
+    },
+    {
+      "epoch": 0.1386271870794078,
+      "grad_norm": 1.7018476473220923,
+      "learning_rate": 9.694029760872539e-06,
+      "loss": 0.3609,
+      "step": 927
+    },
+    {
+      "epoch": 0.13877673097053986,
+      "grad_norm": 1.5673565467226127,
+      "learning_rate": 9.693195020817776e-06,
+      "loss": 0.3164,
+      "step": 928
+    },
+    {
+      "epoch": 0.1389262748616719,
+      "grad_norm": 1.6536061805316273,
+      "learning_rate": 9.69235917969423e-06,
+      "loss": 0.5039,
+      "step": 929
+    },
+    {
+      "epoch": 0.13907581875280395,
+      "grad_norm": 1.4953772716061529,
+      "learning_rate": 9.691522237698001e-06,
+      "loss": 0.2073,
+      "step": 930
+    },
+    {
+      "epoch": 0.139225362643936,
+      "grad_norm": 1.0372555974478648,
+      "learning_rate": 9.69068419502544e-06,
+      "loss": 0.1904,
+      "step": 931
+    },
+    {
+      "epoch": 0.13937490653506804,
+      "grad_norm": 1.2803091164977878,
+      "learning_rate": 9.689845051873161e-06,
+      "loss": 0.2085,
+      "step": 932
+    },
+    {
+      "epoch": 0.1395244504262001,
+      "grad_norm": 1.4758036204854348,
+      "learning_rate": 9.689004808438036e-06,
+      "loss": 0.2012,
+      "step": 933
+    },
+    {
+      "epoch": 0.13967399431733213,
+      "grad_norm": 1.6660973952553224,
+      "learning_rate": 9.688163464917191e-06,
+      "loss": 0.3286,
+      "step": 934
+    },
+    {
+      "epoch": 0.1398235382084642,
+      "grad_norm": 1.1549059427655604,
+      "learning_rate": 9.687321021508018e-06,
+      "loss": 0.2267,
+      "step": 935
+    },
+    {
+      "epoch": 0.13997308209959622,
+      "grad_norm": 1.272574916603474,
+      "learning_rate": 9.686477478408159e-06,
+      "loss": 0.1829,
+      "step": 936
+    },
+    {
+      "epoch": 0.14012262599072828,
+      "grad_norm": 2.754918857840336,
+      "learning_rate": 9.685632835815519e-06,
+      "loss": 0.4481,
+      "step": 937
+    },
+    {
+      "epoch": 0.14027216988186034,
+      "grad_norm": 1.1790985103907,
+      "learning_rate": 9.684787093928256e-06,
+      "loss": 0.1814,
+      "step": 938
+    },
+    {
+      "epoch": 0.14042171377299237,
+      "grad_norm": 1.011660485817637,
+      "learning_rate": 9.683940252944794e-06,
+      "loss": 0.1863,
+      "step": 939
+    },
+    {
+      "epoch": 0.14057125766412443,
+      "grad_norm": 1.3525074345715755,
+      "learning_rate": 9.68309231306381e-06,
+      "loss": 0.2084,
+      "step": 940
+    },
+    {
+      "epoch": 0.14072080155525646,
+      "grad_norm": 1.6719478297190948,
+      "learning_rate": 9.682243274484231e-06,
+      "loss": 0.3459,
+      "step": 941
+    },
+    {
+      "epoch": 0.14087034544638852,
+      "grad_norm": 1.5225980842484328,
+      "learning_rate": 9.681393137405259e-06,
+      "loss": 0.3082,
+      "step": 942
+    },
+    {
+      "epoch": 0.14101988933752055,
+      "grad_norm": 1.4403779528104341,
+      "learning_rate": 9.680541902026342e-06,
+      "loss": 0.1952,
+      "step": 943
+    },
+    {
+      "epoch": 0.1411694332286526,
+      "grad_norm": 1.7704358094140293,
+      "learning_rate": 9.679689568547184e-06,
+      "loss": 0.2925,
+      "step": 944
+    },
+    {
+      "epoch": 0.14131897711978467,
+      "grad_norm": 1.8325825494125016,
+      "learning_rate": 9.678836137167753e-06,
+      "loss": 0.2354,
+      "step": 945
+    },
+    {
+      "epoch": 0.1414685210109167,
+      "grad_norm": 8.228207444568621,
+      "learning_rate": 9.677981608088274e-06,
+      "loss": 0.1945,
+      "step": 946
+    },
+    {
+      "epoch": 0.14161806490204876,
+      "grad_norm": 1.9420821742118657,
+      "learning_rate": 9.677125981509227e-06,
+      "loss": 0.3745,
+      "step": 947
+    },
+    {
+      "epoch": 0.1417676087931808,
+      "grad_norm": 1.4287526091354055,
+      "learning_rate": 9.676269257631348e-06,
+      "loss": 0.159,
+      "step": 948
+    },
+    {
+      "epoch": 0.14191715268431285,
+      "grad_norm": 2.2979804382628726,
+      "learning_rate": 9.675411436655636e-06,
+      "loss": 0.6715,
+      "step": 949
+    },
+    {
+      "epoch": 0.1420666965754449,
+      "grad_norm": 1.2906292785961546,
+      "learning_rate": 9.67455251878334e-06,
+      "loss": 0.1863,
+      "step": 950
+    },
+    {
+      "epoch": 0.14221624046657694,
+      "grad_norm": 1.542985394545003,
+      "learning_rate": 9.673692504215974e-06,
+      "loss": 0.276,
+      "step": 951
+    },
+    {
+      "epoch": 0.142365784357709,
+      "grad_norm": 2.076155049712511,
+      "learning_rate": 9.672831393155304e-06,
+      "loss": 0.2878,
+      "step": 952
+    },
+    {
+      "epoch": 0.14251532824884103,
+      "grad_norm": 3.2389836303118265,
+      "learning_rate": 9.671969185803357e-06,
+      "loss": 0.4539,
+      "step": 953
+    },
+    {
+      "epoch": 0.1426648721399731,
+      "grad_norm": 0.8145950923200616,
+      "learning_rate": 9.671105882362412e-06,
+      "loss": 0.1916,
+      "step": 954
+    },
+    {
+      "epoch": 0.14281441603110512,
+      "grad_norm": 1.7529614161433102,
+      "learning_rate": 9.67024148303501e-06,
+      "loss": 0.3852,
+      "step": 955
+    },
+    {
+      "epoch": 0.14296395992223718,
+      "grad_norm": 1.6140653815156045,
+      "learning_rate": 9.669375988023947e-06,
+      "loss": 0.3317,
+      "step": 956
+    },
+    {
+      "epoch": 0.14311350381336924,
+      "grad_norm": 1.5917328640289674,
+      "learning_rate": 9.668509397532278e-06,
+      "loss": 0.205,
+      "step": 957
+    },
+    {
+      "epoch": 0.14326304770450127,
+      "grad_norm": 1.9053910041720175,
+      "learning_rate": 9.667641711763311e-06,
+      "loss": 0.2016,
+      "step": 958
+    },
+    {
+      "epoch": 0.14341259159563333,
+      "grad_norm": 1.2223818916012819,
+      "learning_rate": 9.666772930920614e-06,
+      "loss": 0.1818,
+      "step": 959
+    },
+    {
+      "epoch": 0.14356213548676536,
+      "grad_norm": 1.4130639929342779,
+      "learning_rate": 9.665903055208013e-06,
+      "loss": 0.1776,
+      "step": 960
+    },
+    {
+      "epoch": 0.14371167937789742,
+      "grad_norm": 2.515402250912616,
+      "learning_rate": 9.665032084829588e-06,
+      "loss": 0.7429,
+      "step": 961
+    },
+    {
+      "epoch": 0.14386122326902945,
+      "grad_norm": 2.118211041321287,
+      "learning_rate": 9.66416001998968e-06,
+      "loss": 0.5489,
+      "step": 962
+    },
+    {
+      "epoch": 0.1440107671601615,
+      "grad_norm": 1.6323921378905693,
+      "learning_rate": 9.663286860892877e-06,
+      "loss": 0.3446,
+      "step": 963
+    },
+    {
+      "epoch": 0.14416031105129357,
+      "grad_norm": 2.4057165427715335,
+      "learning_rate": 9.662412607744036e-06,
+      "loss": 0.2152,
+      "step": 964
+    },
+    {
+      "epoch": 0.1443098549424256,
+      "grad_norm": 1.5942384074329368,
+      "learning_rate": 9.661537260748264e-06,
+      "loss": 0.3746,
+      "step": 965
+    },
+    {
+      "epoch": 0.14445939883355766,
+      "grad_norm": 1.3815463338875527,
+      "learning_rate": 9.660660820110926e-06,
+      "loss": 0.2255,
+      "step": 966
+    },
+    {
+      "epoch": 0.1446089427246897,
+      "grad_norm": 2.071453535788066,
+      "learning_rate": 9.659783286037643e-06,
+      "loss": 0.5075,
+      "step": 967
+    },
+    {
+      "epoch": 0.14475848661582175,
+      "grad_norm": 1.3265840791490535,
+      "learning_rate": 9.658904658734293e-06,
+      "loss": 0.2295,
+      "step": 968
+    },
+    {
+      "epoch": 0.14490803050695378,
+      "grad_norm": 1.5352887928749521,
+      "learning_rate": 9.658024938407011e-06,
+      "loss": 0.3484,
+      "step": 969
+    },
+    {
+      "epoch": 0.14505757439808584,
+      "grad_norm": 1.5831033158953907,
+      "learning_rate": 9.657144125262186e-06,
+      "loss": 0.4039,
+      "step": 970
+    },
+    {
+      "epoch": 0.1452071182892179,
+      "grad_norm": 1.6289190913913172,
+      "learning_rate": 9.65626221950647e-06,
+      "loss": 0.2471,
+      "step": 971
+    },
+    {
+      "epoch": 0.14535666218034993,
+      "grad_norm": 1.6042830876012686,
+      "learning_rate": 9.655379221346758e-06,
+      "loss": 0.4886,
+      "step": 972
+    },
+    {
+      "epoch": 0.14550620607148199,
+      "grad_norm": 1.2790362471165744,
+      "learning_rate": 9.654495130990218e-06,
+      "loss": 0.2065,
+      "step": 973
+    },
+    {
+      "epoch": 0.14565574996261402,
+      "grad_norm": 1.72777953773445,
+      "learning_rate": 9.653609948644263e-06,
+      "loss": 0.4006,
+      "step": 974
+    },
+    {
+      "epoch": 0.14580529385374608,
+      "grad_norm": 1.7629618868465593,
+      "learning_rate": 9.652723674516566e-06,
+      "loss": 0.2142,
+      "step": 975
+    },
+    {
+      "epoch": 0.14595483774487814,
+      "grad_norm": 2.0277477662995174,
+      "learning_rate": 9.651836308815055e-06,
+      "loss": 0.5248,
+      "step": 976
+    },
+    {
+      "epoch": 0.14610438163601017,
+      "grad_norm": 1.972708159415252,
+      "learning_rate": 9.650947851747913e-06,
+      "loss": 0.6236,
+      "step": 977
+    },
+    {
+      "epoch": 0.14625392552714223,
+      "grad_norm": 1.760683770673202,
+      "learning_rate": 9.650058303523583e-06,
+      "loss": 0.3418,
+      "step": 978
+    },
+    {
+      "epoch": 0.14640346941827426,
+      "grad_norm": 1.8429433949445744,
+      "learning_rate": 9.649167664350762e-06,
+      "loss": 0.475,
+      "step": 979
+    },
+    {
+      "epoch": 0.14655301330940632,
+      "grad_norm": 1.9248727992532182,
+      "learning_rate": 9.6482759344384e-06,
+      "loss": 0.3368,
+      "step": 980
+    },
+    {
+      "epoch": 0.14670255720053835,
+      "grad_norm": 1.9573379093599792,
+      "learning_rate": 9.647383113995707e-06,
+      "loss": 0.4203,
+      "step": 981
+    },
+    {
+      "epoch": 0.1468521010916704,
+      "grad_norm": 2.7789135237612803,
+      "learning_rate": 9.646489203232145e-06,
+      "loss": 0.4581,
+      "step": 982
+    },
+    {
+      "epoch": 0.14700164498280247,
+      "grad_norm": 2.5119171064254724,
+      "learning_rate": 9.645594202357438e-06,
+      "loss": 0.8202,
+      "step": 983
+    },
+    {
+      "epoch": 0.1471511888739345,
+      "grad_norm": 1.4269567118604072,
+      "learning_rate": 9.644698111581562e-06,
+      "loss": 0.1954,
+      "step": 984
+    },
+    {
+      "epoch": 0.14730073276506656,
+      "grad_norm": 1.5211432396508893,
+      "learning_rate": 9.643800931114742e-06,
+      "loss": 0.207,
+      "step": 985
+    },
+    {
+      "epoch": 0.1474502766561986,
+      "grad_norm": 1.607901993958729,
+      "learning_rate": 9.642902661167472e-06,
+      "loss": 0.3046,
+      "step": 986
+    },
+    {
+      "epoch": 0.14759982054733065,
+      "grad_norm": 2.0792720647718776,
+      "learning_rate": 9.642003301950491e-06,
+      "loss": 0.4314,
+      "step": 987
+    },
+    {
+      "epoch": 0.14774936443846268,
+      "grad_norm": 1.6967562403841654,
+      "learning_rate": 9.641102853674799e-06,
+      "loss": 0.2142,
+      "step": 988
+    },
+    {
+      "epoch": 0.14789890832959474,
+      "grad_norm": 1.8666018241841429,
+      "learning_rate": 9.640201316551651e-06,
+      "loss": 0.4817,
+      "step": 989
+    },
+    {
+      "epoch": 0.1480484522207268,
+      "grad_norm": 1.6231253521213436,
+      "learning_rate": 9.639298690792554e-06,
+      "loss": 0.304,
+      "step": 990
+    },
+    {
+      "epoch": 0.14819799611185883,
+      "grad_norm": 0.9588780548142521,
+      "learning_rate": 9.638394976609274e-06,
+      "loss": 0.1709,
+      "step": 991
+    },
+    {
+      "epoch": 0.14834754000299089,
+      "grad_norm": 1.7153802262812925,
+      "learning_rate": 9.637490174213828e-06,
+      "loss": 0.1959,
+      "step": 992
+    },
+    {
+      "epoch": 0.14849708389412292,
+      "grad_norm": 1.8524843327507126,
+      "learning_rate": 9.636584283818496e-06,
+      "loss": 0.3957,
+      "step": 993
+    },
+    {
+      "epoch": 0.14864662778525498,
+      "grad_norm": 1.8045411979525945,
+      "learning_rate": 9.635677305635807e-06,
+      "loss": 0.2565,
+      "step": 994
+    },
+    {
+      "epoch": 0.148796171676387,
+      "grad_norm": 2.1506980932575175,
+      "learning_rate": 9.634769239878545e-06,
+      "loss": 0.3777,
+      "step": 995
+    },
+    {
+      "epoch": 0.14894571556751907,
+      "grad_norm": 2.1465696781275563,
+      "learning_rate": 9.633860086759753e-06,
+      "loss": 0.6056,
+      "step": 996
+    },
+    {
+      "epoch": 0.14909525945865112,
+      "grad_norm": 1.3341555017873934,
+      "learning_rate": 9.632949846492728e-06,
+      "loss": 0.2219,
+      "step": 997
+    },
+    {
+      "epoch": 0.14924480334978316,
+      "grad_norm": 1.2276798431588063,
+      "learning_rate": 9.632038519291017e-06,
+      "loss": 0.2074,
+      "step": 998
+    },
+    {
+      "epoch": 0.14939434724091522,
+      "grad_norm": 1.53279693471178,
+      "learning_rate": 9.63112610536843e-06,
+      "loss": 0.4373,
+      "step": 999
+    },
+    {
+      "epoch": 0.14954389113204725,
+      "grad_norm": 1.770529951492653,
+      "learning_rate": 9.630212604939026e-06,
+      "loss": 0.2495,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1496934350231793,
+      "grad_norm": 1.6948398115693124,
+      "learning_rate": 9.629298018217123e-06,
+      "loss": 0.467,
+      "step": 1001
+    },
+    {
+      "epoch": 0.14984297891431134,
+      "grad_norm": 2.0292952301635534,
+      "learning_rate": 9.628382345417291e-06,
+      "loss": 0.5509,
+      "step": 1002
+    },
+    {
+      "epoch": 0.1499925228054434,
+      "grad_norm": 2.4704877435341186,
+      "learning_rate": 9.627465586754354e-06,
+      "loss": 0.2214,
+      "step": 1003
+    },
+    {
+      "epoch": 0.15014206669657545,
+      "grad_norm": 1.6257900509098846,
+      "learning_rate": 9.626547742443394e-06,
+      "loss": 0.3365,
+      "step": 1004
+    },
+    {
+      "epoch": 0.15029161058770749,
+      "grad_norm": 1.8331565532381362,
+      "learning_rate": 9.625628812699747e-06,
+      "loss": 0.4435,
+      "step": 1005
+    },
+    {
+      "epoch": 0.15044115447883954,
+      "grad_norm": 2.977053464112802,
+      "learning_rate": 9.624708797739002e-06,
+      "loss": 0.3177,
+      "step": 1006
+    },
+    {
+      "epoch": 0.15059069836997158,
+      "grad_norm": 1.756417225108672,
+      "learning_rate": 9.623787697777001e-06,
+      "loss": 0.3878,
+      "step": 1007
+    },
+    {
+      "epoch": 0.15074024226110364,
+      "grad_norm": 1.6276234348428167,
+      "learning_rate": 9.622865513029846e-06,
+      "loss": 0.1901,
+      "step": 1008
+    },
+    {
+      "epoch": 0.1508897861522357,
+      "grad_norm": 3.3786401544653253,
+      "learning_rate": 9.62194224371389e-06,
+      "loss": 0.2454,
+      "step": 1009
+    },
+    {
+      "epoch": 0.15103933004336773,
+      "grad_norm": 2.0516682586662456,
+      "learning_rate": 9.621017890045739e-06,
+      "loss": 0.7163,
+      "step": 1010
+    },
+    {
+      "epoch": 0.15118887393449978,
+      "grad_norm": 1.0885572394332461,
+      "learning_rate": 9.620092452242257e-06,
+      "loss": 0.2232,
+      "step": 1011
+    },
+    {
+      "epoch": 0.15133841782563182,
+      "grad_norm": 1.287954715641653,
+      "learning_rate": 9.61916593052056e-06,
+      "loss": 0.2124,
+      "step": 1012
+    },
+    {
+      "epoch": 0.15148796171676387,
+      "grad_norm": 2.089649291216167,
+      "learning_rate": 9.618238325098021e-06,
+      "loss": 0.5129,
+      "step": 1013
+    },
+    {
+      "epoch": 0.1516375056078959,
+      "grad_norm": 1.512994828933595,
+      "learning_rate": 9.617309636192262e-06,
+      "loss": 0.2986,
+      "step": 1014
+    },
+    {
+      "epoch": 0.15178704949902797,
+      "grad_norm": 1.6901350993268085,
+      "learning_rate": 9.616379864021163e-06,
+      "loss": 0.1893,
+      "step": 1015
+    },
+    {
+      "epoch": 0.15193659339016002,
+      "grad_norm": 1.6240769043159409,
+      "learning_rate": 9.615449008802858e-06,
+      "loss": 0.1984,
+      "step": 1016
+    },
+    {
+      "epoch": 0.15208613728129206,
+      "grad_norm": 2.2110380817085815,
+      "learning_rate": 9.614517070755736e-06,
+      "loss": 0.4573,
+      "step": 1017
+    },
+    {
+      "epoch": 0.15223568117242411,
+      "grad_norm": 2.3590384489615452,
+      "learning_rate": 9.613584050098436e-06,
+      "loss": 0.536,
+      "step": 1018
+    },
+    {
+      "epoch": 0.15238522506355615,
+      "grad_norm": 1.1021064926756596,
+      "learning_rate": 9.612649947049856e-06,
+      "loss": 0.224,
+      "step": 1019
+    },
+    {
+      "epoch": 0.1525347689546882,
+      "grad_norm": 1.73759279762283,
+      "learning_rate": 9.611714761829146e-06,
+      "loss": 0.2581,
+      "step": 1020
+    },
+    {
+      "epoch": 0.15268431284582024,
+      "grad_norm": 1.9739121848543864,
+      "learning_rate": 9.610778494655706e-06,
+      "loss": 0.2116,
+      "step": 1021
+    },
+    {
+      "epoch": 0.1528338567369523,
+      "grad_norm": 1.7608676639305771,
+      "learning_rate": 9.609841145749196e-06,
+      "loss": 0.2343,
+      "step": 1022
+    },
+    {
+      "epoch": 0.15298340062808435,
+      "grad_norm": 1.8973842729049986,
+      "learning_rate": 9.608902715329527e-06,
+      "loss": 0.416,
+      "step": 1023
+    },
+    {
+      "epoch": 0.15313294451921639,
+      "grad_norm": 1.232118282013805,
+      "learning_rate": 9.607963203616862e-06,
+      "loss": 0.2428,
+      "step": 1024
+    },
+    {
+      "epoch": 0.15328248841034844,
+      "grad_norm": 1.65077206870404,
+      "learning_rate": 9.607022610831623e-06,
+      "loss": 0.395,
+      "step": 1025
+    },
+    {
+      "epoch": 0.15343203230148048,
+      "grad_norm": 1.8519962462671438,
+      "learning_rate": 9.606080937194478e-06,
+      "loss": 0.3225,
+      "step": 1026
+    },
+    {
+      "epoch": 0.15358157619261253,
+      "grad_norm": 1.5406605876000279,
+      "learning_rate": 9.605138182926355e-06,
+      "loss": 0.1962,
+      "step": 1027
+    },
+    {
+      "epoch": 0.15373112008374457,
+      "grad_norm": 1.68724090741763,
+      "learning_rate": 9.604194348248432e-06,
+      "loss": 0.3412,
+      "step": 1028
+    },
+    {
+      "epoch": 0.15388066397487662,
+      "grad_norm": 1.3196225832605089,
+      "learning_rate": 9.603249433382145e-06,
+      "loss": 0.203,
+      "step": 1029
+    },
+    {
+      "epoch": 0.15403020786600868,
+      "grad_norm": 1.4552647720547889,
+      "learning_rate": 9.602303438549177e-06,
+      "loss": 0.2683,
+      "step": 1030
+    },
+    {
+      "epoch": 0.15417975175714072,
+      "grad_norm": 1.9356496958747662,
+      "learning_rate": 9.601356363971467e-06,
+      "loss": 0.4085,
+      "step": 1031
+    },
+    {
+      "epoch": 0.15432929564827277,
+      "grad_norm": 1.5727583829762162,
+      "learning_rate": 9.60040820987121e-06,
+      "loss": 0.2702,
+      "step": 1032
+    },
+    {
+      "epoch": 0.1544788395394048,
+      "grad_norm": 1.9130911912208222,
+      "learning_rate": 9.59945897647085e-06,
+      "loss": 0.3831,
+      "step": 1033
+    },
+    {
+      "epoch": 0.15462838343053686,
+      "grad_norm": 1.8240341076741502,
+      "learning_rate": 9.59850866399309e-06,
+      "loss": 0.4938,
+      "step": 1034
+    },
+    {
+      "epoch": 0.15477792732166892,
+      "grad_norm": 3.1510439843833953,
+      "learning_rate": 9.597557272660878e-06,
+      "loss": 0.4082,
+      "step": 1035
+    },
+    {
+      "epoch": 0.15492747121280095,
+      "grad_norm": 2.7952488908518998,
+      "learning_rate": 9.596604802697422e-06,
+      "loss": 1.0219,
+      "step": 1036
+    },
+    {
+      "epoch": 0.155077015103933,
+      "grad_norm": 1.5968867223126475,
+      "learning_rate": 9.595651254326179e-06,
+      "loss": 0.3575,
+      "step": 1037
+    },
+    {
+      "epoch": 0.15522655899506504,
+      "grad_norm": 1.7270374421609462,
+      "learning_rate": 9.594696627770863e-06,
+      "loss": 0.4184,
+      "step": 1038
+    },
+    {
+      "epoch": 0.1553761028861971,
+      "grad_norm": 1.623026584906972,
+      "learning_rate": 9.593740923255437e-06,
+      "loss": 0.3364,
+      "step": 1039
+    },
+    {
+      "epoch": 0.15552564677732914,
+      "grad_norm": 1.724702934068898,
+      "learning_rate": 9.592784141004118e-06,
+      "loss": 0.5197,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1556751906684612,
+      "grad_norm": 1.2087417414530044,
+      "learning_rate": 9.591826281241379e-06,
+      "loss": 0.3232,
+      "step": 1041
+    },
+    {
+      "epoch": 0.15582473455959325,
+      "grad_norm": 2.0110371073704716,
+      "learning_rate": 9.590867344191941e-06,
+      "loss": 0.4617,
+      "step": 1042
+    },
+    {
+      "epoch": 0.15597427845072528,
+      "grad_norm": 2.300355596627081,
+      "learning_rate": 9.58990733008078e-06,
+      "loss": 0.3483,
+      "step": 1043
+    },
+    {
+      "epoch": 0.15612382234185734,
+      "grad_norm": 1.9228199791045963,
+      "learning_rate": 9.588946239133123e-06,
+      "loss": 0.4723,
+      "step": 1044
+    },
+    {
+      "epoch": 0.15627336623298937,
+      "grad_norm": 1.3725700472934328,
+      "learning_rate": 9.587984071574455e-06,
+      "loss": 0.212,
+      "step": 1045
+    },
+    {
+      "epoch": 0.15642291012412143,
+      "grad_norm": 1.9304969649682522,
+      "learning_rate": 9.587020827630507e-06,
+      "loss": 0.2317,
+      "step": 1046
+    },
+    {
+      "epoch": 0.15657245401525346,
+      "grad_norm": 1.6204729661048964,
+      "learning_rate": 9.586056507527266e-06,
+      "loss": 0.2135,
+      "step": 1047
+    },
+    {
+      "epoch": 0.15672199790638552,
+      "grad_norm": 1.2120441425361188,
+      "learning_rate": 9.58509111149097e-06,
+      "loss": 0.2785,
+      "step": 1048
+    },
+    {
+      "epoch": 0.15687154179751758,
+      "grad_norm": 2.1856010368938135,
+      "learning_rate": 9.584124639748114e-06,
+      "loss": 0.4117,
+      "step": 1049
+    },
+    {
+      "epoch": 0.15702108568864961,
+      "grad_norm": 1.4004860796245104,
+      "learning_rate": 9.583157092525435e-06,
+      "loss": 0.375,
+      "step": 1050
+    },
+    {
+      "epoch": 0.15717062957978167,
+      "grad_norm": 1.1728395243844014,
+      "learning_rate": 9.582188470049935e-06,
+      "loss": 0.2286,
+      "step": 1051
+    },
+    {
+      "epoch": 0.1573201734709137,
+      "grad_norm": 1.3417244382182179,
+      "learning_rate": 9.58121877254886e-06,
+      "loss": 0.2105,
+      "step": 1052
+    },
+    {
+      "epoch": 0.15746971736204576,
+      "grad_norm": 1.3092141915312672,
+      "learning_rate": 9.580248000249709e-06,
+      "loss": 0.253,
+      "step": 1053
+    },
+    {
+      "epoch": 0.1576192612531778,
+      "grad_norm": 0.8768020427524035,
+      "learning_rate": 9.579276153380234e-06,
+      "loss": 0.199,
+      "step": 1054
+    },
+    {
+      "epoch": 0.15776880514430985,
+      "grad_norm": 1.8273743204900599,
+      "learning_rate": 9.578303232168442e-06,
+      "loss": 0.5377,
+      "step": 1055
+    },
+    {
+      "epoch": 0.1579183490354419,
+      "grad_norm": 1.4617178204858172,
+      "learning_rate": 9.57732923684259e-06,
+      "loss": 0.2318,
+      "step": 1056
+    },
+    {
+      "epoch": 0.15806789292657394,
+      "grad_norm": 1.7897112328181541,
+      "learning_rate": 9.576354167631186e-06,
+      "loss": 0.2312,
+      "step": 1057
+    },
+    {
+      "epoch": 0.158217436817706,
+      "grad_norm": 2.055748147218842,
+      "learning_rate": 9.575378024762991e-06,
+      "loss": 0.208,
+      "step": 1058
+    },
+    {
+      "epoch": 0.15836698070883803,
+      "grad_norm": 1.9825123132080376,
+      "learning_rate": 9.574400808467015e-06,
+      "loss": 0.4415,
+      "step": 1059
+    },
+    {
+      "epoch": 0.1585165245999701,
+      "grad_norm": 1.203644734681587,
+      "learning_rate": 9.573422518972524e-06,
+      "loss": 0.1651,
+      "step": 1060
+    },
+    {
+      "epoch": 0.15866606849110212,
+      "grad_norm": 2.143121544386558,
+      "learning_rate": 9.572443156509035e-06,
+      "loss": 0.2813,
+      "step": 1061
+    },
+    {
+      "epoch": 0.15881561238223418,
+      "grad_norm": 1.7645898839865752,
+      "learning_rate": 9.571462721306315e-06,
+      "loss": 0.314,
+      "step": 1062
+    },
+    {
+      "epoch": 0.15896515627336624,
+      "grad_norm": 1.6678516650591295,
+      "learning_rate": 9.570481213594385e-06,
+      "loss": 0.289,
+      "step": 1063
+    },
+    {
+      "epoch": 0.15911470016449827,
+      "grad_norm": 1.2850170647115653,
+      "learning_rate": 9.569498633603513e-06,
+      "loss": 0.2024,
+      "step": 1064
+    },
+    {
+      "epoch": 0.15926424405563033,
+      "grad_norm": 1.814229267711803,
+      "learning_rate": 9.568514981564226e-06,
+      "loss": 0.4606,
+      "step": 1065
+    },
+    {
+      "epoch": 0.15941378794676236,
+      "grad_norm": 1.243462382180521,
+      "learning_rate": 9.567530257707294e-06,
+      "loss": 0.204,
+      "step": 1066
+    },
+    {
+      "epoch": 0.15956333183789442,
+      "grad_norm": 1.2909371933651599,
+      "learning_rate": 9.566544462263744e-06,
+      "loss": 0.2209,
+      "step": 1067
+    },
+    {
+      "epoch": 0.15971287572902648,
+      "grad_norm": 1.6817484277248902,
+      "learning_rate": 9.565557595464854e-06,
+      "loss": 0.2266,
+      "step": 1068
+    },
+    {
+      "epoch": 0.1598624196201585,
+      "grad_norm": 1.6783737367668008,
+      "learning_rate": 9.564569657542153e-06,
+      "loss": 0.3126,
+      "step": 1069
+    },
+    {
+      "epoch": 0.16001196351129057,
+      "grad_norm": 1.7121174458116146,
+      "learning_rate": 9.56358064872742e-06,
+      "loss": 0.4938,
+      "step": 1070
+    },
+    {
+      "epoch": 0.1601615074024226,
+      "grad_norm": 1.4168299418446093,
+      "learning_rate": 9.562590569252685e-06,
+      "loss": 0.1859,
+      "step": 1071
+    },
+    {
+      "epoch": 0.16031105129355466,
+      "grad_norm": 1.1574720210896832,
+      "learning_rate": 9.561599419350233e-06,
+      "loss": 0.2076,
+      "step": 1072
+    },
+    {
+      "epoch": 0.1604605951846867,
+      "grad_norm": 1.88010212201121,
+      "learning_rate": 9.560607199252594e-06,
+      "loss": 0.489,
+      "step": 1073
+    },
+    {
+      "epoch": 0.16061013907581875,
+      "grad_norm": 1.7557755514977327,
+      "learning_rate": 9.559613909192553e-06,
+      "loss": 0.2593,
+      "step": 1074
+    },
+    {
+      "epoch": 0.1607596829669508,
+      "grad_norm": 2.0354173981462864,
+      "learning_rate": 9.558619549403148e-06,
+      "loss": 0.266,
+      "step": 1075
+    },
+    {
+      "epoch": 0.16090922685808284,
+      "grad_norm": 1.984771028183608,
+      "learning_rate": 9.557624120117663e-06,
+      "loss": 0.1823,
+      "step": 1076
+    },
+    {
+      "epoch": 0.1610587707492149,
+      "grad_norm": 1.2435802855207878,
+      "learning_rate": 9.556627621569636e-06,
+      "loss": 0.2133,
+      "step": 1077
+    },
+    {
+      "epoch": 0.16120831464034693,
+      "grad_norm": 2.1628484109014603,
+      "learning_rate": 9.555630053992854e-06,
+      "loss": 0.6313,
+      "step": 1078
+    },
+    {
+      "epoch": 0.161357858531479,
+      "grad_norm": 1.3025632600056798,
+      "learning_rate": 9.554631417621358e-06,
+      "loss": 0.1861,
+      "step": 1079
+    },
+    {
+      "epoch": 0.16150740242261102,
+      "grad_norm": 1.2033276727813813,
+      "learning_rate": 9.553631712689437e-06,
+      "loss": 0.1411,
+      "step": 1080
+    },
+    {
+      "epoch": 0.16165694631374308,
+      "grad_norm": 1.7192459951334418,
+      "learning_rate": 9.55263093943163e-06,
+      "loss": 0.3415,
+      "step": 1081
+    },
+    {
+      "epoch": 0.16180649020487514,
+      "grad_norm": 2.195446146090599,
+      "learning_rate": 9.55162909808273e-06,
+      "loss": 0.5552,
+      "step": 1082
+    },
+    {
+      "epoch": 0.16195603409600717,
+      "grad_norm": 1.2518961650965623,
+      "learning_rate": 9.550626188877779e-06,
+      "loss": 0.1376,
+      "step": 1083
+    },
+    {
+      "epoch": 0.16210557798713923,
+      "grad_norm": 1.621475542449237,
+      "learning_rate": 9.549622212052067e-06,
+      "loss": 0.1918,
+      "step": 1084
+    },
+    {
+      "epoch": 0.16225512187827126,
+      "grad_norm": 1.6022576009951304,
+      "learning_rate": 9.548617167841139e-06,
+      "loss": 0.4754,
+      "step": 1085
+    },
+    {
+      "epoch": 0.16240466576940332,
+      "grad_norm": 2.140796057625746,
+      "learning_rate": 9.547611056480785e-06,
+      "loss": 0.3365,
+      "step": 1086
+    },
+    {
+      "epoch": 0.16255420966053535,
+      "grad_norm": 2.1352436057574415,
+      "learning_rate": 9.54660387820705e-06,
+      "loss": 0.2548,
+      "step": 1087
+    },
+    {
+      "epoch": 0.1627037535516674,
+      "grad_norm": 1.533857631085752,
+      "learning_rate": 9.54559563325623e-06,
+      "loss": 0.2295,
+      "step": 1088
+    },
+    {
+      "epoch": 0.16285329744279947,
+      "grad_norm": 1.3212471086223223,
+      "learning_rate": 9.544586321864865e-06,
+      "loss": 0.2047,
+      "step": 1089
+    },
+    {
+      "epoch": 0.1630028413339315,
+      "grad_norm": 1.0794095044455165,
+      "learning_rate": 9.543575944269752e-06,
+      "loss": 0.2824,
+      "step": 1090
+    },
+    {
+      "epoch": 0.16315238522506356,
+      "grad_norm": 1.7780256493338729,
+      "learning_rate": 9.542564500707934e-06,
+      "loss": 0.3763,
+      "step": 1091
+    },
+    {
+      "epoch": 0.1633019291161956,
+      "grad_norm": 1.1347608812400427,
+      "learning_rate": 9.541551991416704e-06,
+      "loss": 0.2046,
+      "step": 1092
+    },
+    {
+      "epoch": 0.16345147300732765,
+      "grad_norm": 1.7298153575861623,
+      "learning_rate": 9.540538416633611e-06,
+      "loss": 0.3394,
+      "step": 1093
+    },
+    {
+      "epoch": 0.1636010168984597,
+      "grad_norm": 1.822850678248627,
+      "learning_rate": 9.539523776596446e-06,
+      "loss": 0.4254,
+      "step": 1094
+    },
+    {
+      "epoch": 0.16375056078959174,
+      "grad_norm": 1.762563084427053,
+      "learning_rate": 9.538508071543253e-06,
+      "loss": 0.4083,
+      "step": 1095
+    },
+    {
+      "epoch": 0.1639001046807238,
+      "grad_norm": 1.653109483519857,
+      "learning_rate": 9.537491301712328e-06,
+      "loss": 0.3344,
+      "step": 1096
+    },
+    {
+      "epoch": 0.16404964857185583,
+      "grad_norm": 2.0381344440794713,
+      "learning_rate": 9.536473467342213e-06,
+      "loss": 0.6724,
+      "step": 1097
+    },
+    {
+      "epoch": 0.1641991924629879,
+      "grad_norm": 1.9754898794658875,
+      "learning_rate": 9.535454568671705e-06,
+      "loss": 0.5666,
+      "step": 1098
+    },
+    {
+      "epoch": 0.16434873635411992,
+      "grad_norm": 2.3447226731261344,
+      "learning_rate": 9.534434605939845e-06,
+      "loss": 0.3964,
+      "step": 1099
+    },
+    {
+      "epoch": 0.16449828024525198,
+      "grad_norm": 1.773763244389934,
+      "learning_rate": 9.533413579385925e-06,
+      "loss": 0.3883,
+      "step": 1100
+    },
+    {
+      "epoch": 0.16464782413638404,
+      "grad_norm": 1.8563640201434743,
+      "learning_rate": 9.532391489249489e-06,
+      "loss": 0.5295,
+      "step": 1101
+    },
+    {
+      "epoch": 0.16479736802751607,
+      "grad_norm": 1.610026424473214,
+      "learning_rate": 9.53136833577033e-06,
+      "loss": 0.2045,
+      "step": 1102
+    },
+    {
+      "epoch": 0.16494691191864813,
+      "grad_norm": 1.1595609057345122,
+      "learning_rate": 9.530344119188489e-06,
+      "loss": 0.196,
+      "step": 1103
+    },
+    {
+      "epoch": 0.16509645580978016,
+      "grad_norm": 1.380807492186586,
+      "learning_rate": 9.529318839744257e-06,
+      "loss": 0.1665,
+      "step": 1104
+    },
+    {
+      "epoch": 0.16524599970091222,
+      "grad_norm": 1.224405894645481,
+      "learning_rate": 9.528292497678175e-06,
+      "loss": 0.1904,
+      "step": 1105
+    },
+    {
+      "epoch": 0.16539554359204425,
+      "grad_norm": 1.6321539600343615,
+      "learning_rate": 9.527265093231031e-06,
+      "loss": 0.3776,
+      "step": 1106
+    },
+    {
+      "epoch": 0.1655450874831763,
+      "grad_norm": 1.8242961573055154,
+      "learning_rate": 9.526236626643867e-06,
+      "loss": 0.2581,
+      "step": 1107
+    },
+    {
+      "epoch": 0.16569463137430837,
+      "grad_norm": 1.3475980511137085,
+      "learning_rate": 9.525207098157968e-06,
+      "loss": 0.3415,
+      "step": 1108
+    },
+    {
+      "epoch": 0.1658441752654404,
+      "grad_norm": 1.9882440245015858,
+      "learning_rate": 9.524176508014873e-06,
+      "loss": 0.5573,
+      "step": 1109
+    },
+    {
+      "epoch": 0.16599371915657246,
+      "grad_norm": 1.5816342314828584,
+      "learning_rate": 9.523144856456367e-06,
+      "loss": 0.2691,
+      "step": 1110
+    },
+    {
+      "epoch": 0.1661432630477045,
+      "grad_norm": 1.3850164199795663,
+      "learning_rate": 9.522112143724489e-06,
+      "loss": 0.3378,
+      "step": 1111
+    },
+    {
+      "epoch": 0.16629280693883655,
+      "grad_norm": 1.4909627841603759,
+      "learning_rate": 9.52107837006152e-06,
+      "loss": 0.4023,
+      "step": 1112
+    },
+    {
+      "epoch": 0.16644235082996858,
+      "grad_norm": 1.338411558324243,
+      "learning_rate": 9.520043535709994e-06,
+      "loss": 0.225,
+      "step": 1113
+    },
+    {
+      "epoch": 0.16659189472110064,
+      "grad_norm": 1.5004852001929436,
+      "learning_rate": 9.519007640912691e-06,
+      "loss": 0.367,
+      "step": 1114
+    },
+    {
+      "epoch": 0.1667414386122327,
+      "grad_norm": 1.3522433759924148,
+      "learning_rate": 9.517970685912648e-06,
+      "loss": 0.3267,
+      "step": 1115
+    },
+    {
+      "epoch": 0.16689098250336473,
+      "grad_norm": 1.2771009278550414,
+      "learning_rate": 9.516932670953137e-06,
+      "loss": 0.2343,
+      "step": 1116
+    },
+    {
+      "epoch": 0.1670405263944968,
+      "grad_norm": 1.2558718520533958,
+      "learning_rate": 9.515893596277692e-06,
+      "loss": 0.2146,
+      "step": 1117
+    },
+    {
+      "epoch": 0.16719007028562882,
+      "grad_norm": 1.747427684722942,
+      "learning_rate": 9.514853462130087e-06,
+      "loss": 0.3897,
+      "step": 1118
+    },
+    {
+      "epoch": 0.16733961417676088,
+      "grad_norm": 1.350710739291567,
+      "learning_rate": 9.51381226875435e-06,
+      "loss": 0.1962,
+      "step": 1119
+    },
+    {
+      "epoch": 0.16748915806789294,
+      "grad_norm": 1.7048566484317351,
+      "learning_rate": 9.512770016394754e-06,
+      "loss": 0.4704,
+      "step": 1120
+    },
+    {
+      "epoch": 0.16763870195902497,
+      "grad_norm": 1.1820262059156426,
+      "learning_rate": 9.51172670529582e-06,
+      "loss": 0.2269,
+      "step": 1121
+    },
+    {
+      "epoch": 0.16778824585015703,
+      "grad_norm": 1.4968174934291172,
+      "learning_rate": 9.51068233570232e-06,
+      "loss": 0.2122,
+      "step": 1122
+    },
+    {
+      "epoch": 0.16793778974128906,
+      "grad_norm": 1.7978736217167008,
+      "learning_rate": 9.50963690785927e-06,
+      "loss": 0.4423,
+      "step": 1123
+    },
+    {
+      "epoch": 0.16808733363242112,
+      "grad_norm": 1.9086737935680094,
+      "learning_rate": 9.508590422011943e-06,
+      "loss": 0.5837,
+      "step": 1124
+    },
+    {
+      "epoch": 0.16823687752355315,
+      "grad_norm": 1.809798628211579,
+      "learning_rate": 9.507542878405853e-06,
+      "loss": 0.4956,
+      "step": 1125
+    },
+    {
+      "epoch": 0.1683864214146852,
+      "grad_norm": 2.256680269043329,
+      "learning_rate": 9.506494277286762e-06,
+      "loss": 0.3622,
+      "step": 1126
+    },
+    {
+      "epoch": 0.16853596530581727,
+      "grad_norm": 1.7846176870962676,
+      "learning_rate": 9.505444618900682e-06,
+      "loss": 0.3712,
+      "step": 1127
+    },
+    {
+      "epoch": 0.1686855091969493,
+      "grad_norm": 2.142681380142415,
+      "learning_rate": 9.504393903493874e-06,
+      "loss": 0.2523,
+      "step": 1128
+    },
+    {
+      "epoch": 0.16883505308808136,
+      "grad_norm": 1.993049487456574,
+      "learning_rate": 9.503342131312847e-06,
+      "loss": 0.4667,
+      "step": 1129
+    },
+    {
+      "epoch": 0.1689845969792134,
+      "grad_norm": 1.4929357757100867,
+      "learning_rate": 9.502289302604355e-06,
+      "loss": 0.2124,
+      "step": 1130
+    },
+    {
+      "epoch": 0.16913414087034545,
+      "grad_norm": 1.763381368002964,
+      "learning_rate": 9.501235417615402e-06,
+      "loss": 0.2563,
+      "step": 1131
+    },
+    {
+      "epoch": 0.16928368476147748,
+      "grad_norm": 1.879192084699487,
+      "learning_rate": 9.50018047659324e-06,
+      "loss": 0.6,
+      "step": 1132
+    },
+    {
+      "epoch": 0.16943322865260954,
+      "grad_norm": 1.7482598713089221,
+      "learning_rate": 9.49912447978537e-06,
+      "loss": 0.4743,
+      "step": 1133
+    },
+    {
+      "epoch": 0.1695827725437416,
+      "grad_norm": 2.413100609605461,
+      "learning_rate": 9.498067427439535e-06,
+      "loss": 0.5021,
+      "step": 1134
+    },
+    {
+      "epoch": 0.16973231643487363,
+      "grad_norm": 1.6675332056868586,
+      "learning_rate": 9.497009319803732e-06,
+      "loss": 0.3704,
+      "step": 1135
+    },
+    {
+      "epoch": 0.1698818603260057,
+      "grad_norm": 1.4103198339625334,
+      "learning_rate": 9.495950157126204e-06,
+      "loss": 0.3525,
+      "step": 1136
+    },
+    {
+      "epoch": 0.17003140421713772,
+      "grad_norm": 1.7592756830906924,
+      "learning_rate": 9.49488993965544e-06,
+      "loss": 0.2118,
+      "step": 1137
+    },
+    {
+      "epoch": 0.17018094810826978,
+      "grad_norm": 1.209600322516803,
+      "learning_rate": 9.493828667640179e-06,
+      "loss": 0.2156,
+      "step": 1138
+    },
+    {
+      "epoch": 0.1703304919994018,
+      "grad_norm": 1.8747303848334145,
+      "learning_rate": 9.492766341329402e-06,
+      "loss": 0.2998,
+      "step": 1139
+    },
+    {
+      "epoch": 0.17048003589053387,
+      "grad_norm": 1.9498503355521368,
+      "learning_rate": 9.491702960972343e-06,
+      "loss": 0.3723,
+      "step": 1140
+    },
+    {
+      "epoch": 0.17062957978166593,
+      "grad_norm": 1.6726039919291162,
+      "learning_rate": 9.490638526818482e-06,
+      "loss": 0.408,
+      "step": 1141
+    },
+    {
+      "epoch": 0.17077912367279796,
+      "grad_norm": 3.7262669891973386,
+      "learning_rate": 9.489573039117543e-06,
+      "loss": 0.4009,
+      "step": 1142
+    },
+    {
+      "epoch": 0.17092866756393002,
+      "grad_norm": 1.8934485182919263,
+      "learning_rate": 9.488506498119502e-06,
+      "loss": 0.5141,
+      "step": 1143
+    },
+    {
+      "epoch": 0.17107821145506205,
+      "grad_norm": 1.4710896960429016,
+      "learning_rate": 9.487438904074581e-06,
+      "loss": 0.3149,
+      "step": 1144
+    },
+    {
+      "epoch": 0.1712277553461941,
+      "grad_norm": 0.7650448318548279,
+      "learning_rate": 9.486370257233244e-06,
+      "loss": 0.1787,
+      "step": 1145
+    },
+    {
+      "epoch": 0.17137729923732614,
+      "grad_norm": 1.110992686058532,
+      "learning_rate": 9.48530055784621e-06,
+      "loss": 0.2058,
+      "step": 1146
+    },
+    {
+      "epoch": 0.1715268431284582,
+      "grad_norm": 1.7079056261573908,
+      "learning_rate": 9.484229806164435e-06,
+      "loss": 0.2944,
+      "step": 1147
+    },
+    {
+      "epoch": 0.17167638701959026,
+      "grad_norm": 1.5033153679810882,
+      "learning_rate": 9.483158002439134e-06,
+      "loss": 0.2221,
+      "step": 1148
+    },
+    {
+      "epoch": 0.1718259309107223,
+      "grad_norm": 1.672220461796242,
+      "learning_rate": 9.48208514692176e-06,
+      "loss": 0.3671,
+      "step": 1149
+    },
+    {
+      "epoch": 0.17197547480185435,
+      "grad_norm": 2.0626867491532885,
+      "learning_rate": 9.481011239864014e-06,
+      "loss": 0.3915,
+      "step": 1150
+    },
+    {
+      "epoch": 0.17212501869298638,
+      "grad_norm": 1.6104087564965828,
+      "learning_rate": 9.479936281517848e-06,
+      "loss": 0.3084,
+      "step": 1151
+    },
+    {
+      "epoch": 0.17227456258411844,
+      "grad_norm": 1.2158699033191027,
+      "learning_rate": 9.478860272135452e-06,
+      "loss": 0.2074,
+      "step": 1152
+    },
+    {
+      "epoch": 0.1724241064752505,
+      "grad_norm": 1.5670367135087604,
+      "learning_rate": 9.477783211969273e-06,
+      "loss": 0.3216,
+      "step": 1153
+    },
+    {
+      "epoch": 0.17257365036638253,
+      "grad_norm": 1.1456620290612602,
+      "learning_rate": 9.476705101272e-06,
+      "loss": 0.1631,
+      "step": 1154
+    },
+    {
+      "epoch": 0.1727231942575146,
+      "grad_norm": 2.240627806354505,
+      "learning_rate": 9.475625940296567e-06,
+      "loss": 0.3657,
+      "step": 1155
+    },
+    {
+      "epoch": 0.17287273814864662,
+      "grad_norm": 1.5745880507830639,
+      "learning_rate": 9.474545729296152e-06,
+      "loss": 0.2223,
+      "step": 1156
+    },
+    {
+      "epoch": 0.17302228203977868,
+      "grad_norm": 2.112259044674356,
+      "learning_rate": 9.47346446852419e-06,
+      "loss": 0.6911,
+      "step": 1157
+    },
+    {
+      "epoch": 0.1731718259309107,
+      "grad_norm": 1.5810443743533733,
+      "learning_rate": 9.472382158234349e-06,
+      "loss": 0.3099,
+      "step": 1158
+    },
+    {
+      "epoch": 0.17332136982204277,
+      "grad_norm": 1.2778440299252458,
+      "learning_rate": 9.47129879868055e-06,
+      "loss": 0.2136,
+      "step": 1159
+    },
+    {
+      "epoch": 0.17347091371317483,
+      "grad_norm": 1.4299544830347497,
+      "learning_rate": 9.470214390116965e-06,
+      "loss": 0.2145,
+      "step": 1160
+    },
+    {
+      "epoch": 0.17362045760430686,
+      "grad_norm": 1.3698311545901527,
+      "learning_rate": 9.469128932798e-06,
+      "loss": 0.2377,
+      "step": 1161
+    },
+    {
+      "epoch": 0.17377000149543892,
+      "grad_norm": 1.2952413351897127,
+      "learning_rate": 9.468042426978319e-06,
+      "loss": 0.2452,
+      "step": 1162
+    },
+    {
+      "epoch": 0.17391954538657095,
+      "grad_norm": 1.3794054233492332,
+      "learning_rate": 9.466954872912823e-06,
+      "loss": 0.1923,
+      "step": 1163
+    },
+    {
+      "epoch": 0.174069089277703,
+      "grad_norm": 1.1390407334512478,
+      "learning_rate": 9.465866270856665e-06,
+      "loss": 0.251,
+      "step": 1164
+    },
+    {
+      "epoch": 0.17421863316883504,
+      "grad_norm": 2.280807269259536,
+      "learning_rate": 9.46477662106524e-06,
+      "loss": 0.5433,
+      "step": 1165
+    },
+    {
+      "epoch": 0.1743681770599671,
+      "grad_norm": 1.4481272121275899,
+      "learning_rate": 9.463685923794191e-06,
+      "loss": 0.3209,
+      "step": 1166
+    },
+    {
+      "epoch": 0.17451772095109916,
+      "grad_norm": 1.3194380931708358,
+      "learning_rate": 9.462594179299408e-06,
+      "loss": 0.2208,
+      "step": 1167
+    },
+    {
+      "epoch": 0.1746672648422312,
+      "grad_norm": 2.0036222277986,
+      "learning_rate": 9.46150138783702e-06,
+      "loss": 0.2654,
+      "step": 1168
+    },
+    {
+      "epoch": 0.17481680873336325,
+      "grad_norm": 0.8359301847512149,
+      "learning_rate": 9.460407549663411e-06,
+      "loss": 0.2131,
+      "step": 1169
+    },
+    {
+      "epoch": 0.17496635262449528,
+      "grad_norm": 1.8746705630094216,
+      "learning_rate": 9.459312665035203e-06,
+      "loss": 0.2415,
+      "step": 1170
+    },
+    {
+      "epoch": 0.17511589651562734,
+      "grad_norm": 2.187475783350278,
+      "learning_rate": 9.458216734209269e-06,
+      "loss": 0.6378,
+      "step": 1171
+    },
+    {
+      "epoch": 0.17526544040675937,
+      "grad_norm": 1.3984177897039258,
+      "learning_rate": 9.457119757442723e-06,
+      "loss": 0.2034,
+      "step": 1172
+    },
+    {
+      "epoch": 0.17541498429789143,
+      "grad_norm": 1.776218225609686,
+      "learning_rate": 9.456021734992928e-06,
+      "loss": 0.3717,
+      "step": 1173
+    },
+    {
+      "epoch": 0.1755645281890235,
+      "grad_norm": 1.360440620305923,
+      "learning_rate": 9.45492266711749e-06,
+      "loss": 0.2499,
+      "step": 1174
+    },
+    {
+      "epoch": 0.17571407208015552,
+      "grad_norm": 1.11686981125786,
+      "learning_rate": 9.453822554074259e-06,
+      "loss": 0.1718,
+      "step": 1175
+    },
+    {
+      "epoch": 0.17586361597128758,
+      "grad_norm": 1.7801052355130151,
+      "learning_rate": 9.452721396121333e-06,
+      "loss": 0.392,
+      "step": 1176
+    },
+    {
+      "epoch": 0.1760131598624196,
+      "grad_norm": 1.4238435171417776,
+      "learning_rate": 9.451619193517057e-06,
+      "loss": 0.4248,
+      "step": 1177
+    },
+    {
+      "epoch": 0.17616270375355167,
+      "grad_norm": 1.4639661830539468,
+      "learning_rate": 9.450515946520016e-06,
+      "loss": 0.2049,
+      "step": 1178
+    },
+    {
+      "epoch": 0.17631224764468373,
+      "grad_norm": 1.5499038616483287,
+      "learning_rate": 9.449411655389042e-06,
+      "loss": 0.4062,
+      "step": 1179
+    },
+    {
+      "epoch": 0.17646179153581576,
+      "grad_norm": 1.7136605249554677,
+      "learning_rate": 9.448306320383215e-06,
+      "loss": 0.3911,
+      "step": 1180
+    },
+    {
+      "epoch": 0.17661133542694782,
+      "grad_norm": 1.4594560853866743,
+      "learning_rate": 9.447199941761852e-06,
+      "loss": 0.2117,
+      "step": 1181
+    },
+    {
+      "epoch": 0.17676087931807985,
+      "grad_norm": 1.6116818115006617,
+      "learning_rate": 9.446092519784525e-06,
+      "loss": 0.19,
+      "step": 1182
+    },
+    {
+      "epoch": 0.1769104232092119,
+      "grad_norm": 1.5420971786965234,
+      "learning_rate": 9.444984054711044e-06,
+      "loss": 0.3199,
+      "step": 1183
+    },
+    {
+      "epoch": 0.17705996710034394,
+      "grad_norm": 1.6968645734800356,
+      "learning_rate": 9.443874546801465e-06,
+      "loss": 0.3796,
+      "step": 1184
+    },
+    {
+      "epoch": 0.177209510991476,
+      "grad_norm": 1.3181328359871092,
+      "learning_rate": 9.442763996316093e-06,
+      "loss": 0.3955,
+      "step": 1185
+    },
+    {
+      "epoch": 0.17735905488260806,
+      "grad_norm": 2.8227798204695733,
+      "learning_rate": 9.44165240351547e-06,
+      "loss": 0.5894,
+      "step": 1186
+    },
+    {
+      "epoch": 0.1775085987737401,
+      "grad_norm": 1.9231832633740291,
+      "learning_rate": 9.440539768660386e-06,
+      "loss": 0.3844,
+      "step": 1187
+    },
+    {
+      "epoch": 0.17765814266487215,
+      "grad_norm": 1.4508084994855301,
+      "learning_rate": 9.439426092011877e-06,
+      "loss": 0.1983,
+      "step": 1188
+    },
+    {
+      "epoch": 0.17780768655600418,
+      "grad_norm": 1.838910674129457,
+      "learning_rate": 9.438311373831224e-06,
+      "loss": 0.5758,
+      "step": 1189
+    },
+    {
+      "epoch": 0.17795723044713624,
+      "grad_norm": 1.878745494476723,
+      "learning_rate": 9.437195614379947e-06,
+      "loss": 0.4892,
+      "step": 1190
+    },
+    {
+      "epoch": 0.17810677433826827,
+      "grad_norm": 1.5763295601740275,
+      "learning_rate": 9.436078813919818e-06,
+      "loss": 0.4209,
+      "step": 1191
+    },
+    {
+      "epoch": 0.17825631822940033,
+      "grad_norm": 1.939965363898736,
+      "learning_rate": 9.434960972712846e-06,
+      "loss": 0.4915,
+      "step": 1192
+    },
+    {
+      "epoch": 0.1784058621205324,
+      "grad_norm": 1.777010297411083,
+      "learning_rate": 9.433842091021287e-06,
+      "loss": 0.4445,
+      "step": 1193
+    },
+    {
+      "epoch": 0.17855540601166442,
+      "grad_norm": 1.8576886519979177,
+      "learning_rate": 9.432722169107647e-06,
+      "loss": 0.2065,
+      "step": 1194
+    },
+    {
+      "epoch": 0.17870494990279648,
+      "grad_norm": 1.6414559576032928,
+      "learning_rate": 9.431601207234663e-06,
+      "loss": 0.2878,
+      "step": 1195
+    },
+    {
+      "epoch": 0.1788544937939285,
+      "grad_norm": 1.6433452397947506,
+      "learning_rate": 9.430479205665329e-06,
+      "loss": 0.3933,
+      "step": 1196
+    },
+    {
+      "epoch": 0.17900403768506057,
+      "grad_norm": 1.3219617553631218,
+      "learning_rate": 9.429356164662872e-06,
+      "loss": 0.1886,
+      "step": 1197
+    },
+    {
+      "epoch": 0.1791535815761926,
+      "grad_norm": 1.3673182854086454,
+      "learning_rate": 9.428232084490774e-06,
+      "loss": 0.2098,
+      "step": 1198
+    },
+    {
+      "epoch": 0.17930312546732466,
+      "grad_norm": 1.4932716672657123,
+      "learning_rate": 9.427106965412752e-06,
+      "loss": 0.1868,
+      "step": 1199
+    },
+    {
+      "epoch": 0.17945266935845672,
+      "grad_norm": 1.2835655324809725,
+      "learning_rate": 9.425980807692771e-06,
+      "loss": 0.2841,
+      "step": 1200
+    },
+    {
+      "epoch": 0.17960221324958875,
+      "grad_norm": 1.5229676329003083,
+      "learning_rate": 9.424853611595037e-06,
+      "loss": 0.429,
+      "step": 1201
+    },
+    {
+      "epoch": 0.1797517571407208,
+      "grad_norm": 1.3834763754455093,
+      "learning_rate": 9.423725377384e-06,
+      "loss": 0.199,
+      "step": 1202
+    },
+    {
+      "epoch": 0.17990130103185284,
+      "grad_norm": 1.172574987366,
+      "learning_rate": 9.42259610532436e-06,
+      "loss": 0.2422,
+      "step": 1203
+    },
+    {
+      "epoch": 0.1800508449229849,
+      "grad_norm": 2.3677332220742753,
+      "learning_rate": 9.421465795681048e-06,
+      "loss": 0.6703,
+      "step": 1204
+    },
+    {
+      "epoch": 0.18020038881411693,
+      "grad_norm": 1.743670576433428,
+      "learning_rate": 9.420334448719251e-06,
+      "loss": 0.3879,
+      "step": 1205
+    },
+    {
+      "epoch": 0.180349932705249,
+      "grad_norm": 1.4354052350500734,
+      "learning_rate": 9.419202064704393e-06,
+      "loss": 0.2261,
+      "step": 1206
+    },
+    {
+      "epoch": 0.18049947659638105,
+      "grad_norm": 1.5355684537494616,
+      "learning_rate": 9.41806864390214e-06,
+      "loss": 0.2323,
+      "step": 1207
+    },
+    {
+      "epoch": 0.18064902048751308,
+      "grad_norm": 2.7749706919729067,
+      "learning_rate": 9.416934186578403e-06,
+      "loss": 0.2457,
+      "step": 1208
+    },
+    {
+      "epoch": 0.18079856437864514,
+      "grad_norm": 1.7998786782731084,
+      "learning_rate": 9.41579869299934e-06,
+      "loss": 0.5115,
+      "step": 1209
+    },
+    {
+      "epoch": 0.18094810826977717,
+      "grad_norm": 1.722249547477117,
+      "learning_rate": 9.414662163431347e-06,
+      "loss": 0.3978,
+      "step": 1210
+    },
+    {
+      "epoch": 0.18109765216090923,
+      "grad_norm": 1.470878100530038,
+      "learning_rate": 9.413524598141065e-06,
+      "loss": 0.3655,
+      "step": 1211
+    },
+    {
+      "epoch": 0.1812471960520413,
+      "grad_norm": 1.2321837235938764,
+      "learning_rate": 9.412385997395377e-06,
+      "loss": 0.206,
+      "step": 1212
+    },
+    {
+      "epoch": 0.18139673994317332,
+      "grad_norm": 1.71695513424398,
+      "learning_rate": 9.41124636146141e-06,
+      "loss": 0.2988,
+      "step": 1213
+    },
+    {
+      "epoch": 0.18154628383430538,
+      "grad_norm": 1.5665377801862033,
+      "learning_rate": 9.410105690606533e-06,
+      "loss": 0.435,
+      "step": 1214
+    },
+    {
+      "epoch": 0.1816958277254374,
+      "grad_norm": 1.9574571743722469,
+      "learning_rate": 9.40896398509836e-06,
+      "loss": 0.3844,
+      "step": 1215
+    },
+    {
+      "epoch": 0.18184537161656947,
+      "grad_norm": 1.732474617655161,
+      "learning_rate": 9.407821245204746e-06,
+      "loss": 0.4532,
+      "step": 1216
+    },
+    {
+      "epoch": 0.1819949155077015,
+      "grad_norm": 1.6282505343946028,
+      "learning_rate": 9.406677471193788e-06,
+      "loss": 0.346,
+      "step": 1217
+    },
+    {
+      "epoch": 0.18214445939883356,
+      "grad_norm": 1.7687288700904007,
+      "learning_rate": 9.405532663333826e-06,
+      "loss": 0.2398,
+      "step": 1218
+    },
+    {
+      "epoch": 0.18229400328996562,
+      "grad_norm": 1.710427334466053,
+      "learning_rate": 9.404386821893442e-06,
+      "loss": 0.2851,
+      "step": 1219
+    },
+    {
+      "epoch": 0.18244354718109765,
+      "grad_norm": 1.017825559673437,
+      "learning_rate": 9.403239947141467e-06,
+      "loss": 0.1898,
+      "step": 1220
+    },
+    {
+      "epoch": 0.1825930910722297,
+      "grad_norm": 1.7058191164095473,
+      "learning_rate": 9.402092039346961e-06,
+      "loss": 0.3391,
+      "step": 1221
+    },
+    {
+      "epoch": 0.18274263496336174,
+      "grad_norm": 1.2429292971081916,
+      "learning_rate": 9.40094309877924e-06,
+      "loss": 0.247,
+      "step": 1222
+    },
+    {
+      "epoch": 0.1828921788544938,
+      "grad_norm": 1.26527696992994,
+      "learning_rate": 9.399793125707853e-06,
+      "loss": 0.2229,
+      "step": 1223
+    },
+    {
+      "epoch": 0.18304172274562583,
+      "grad_norm": 1.150282472600963,
+      "learning_rate": 9.398642120402596e-06,
+      "loss": 0.2145,
+      "step": 1224
+    },
+    {
+      "epoch": 0.1831912666367579,
+      "grad_norm": 1.3914149403501497,
+      "learning_rate": 9.39749008313351e-06,
+      "loss": 0.231,
+      "step": 1225
+    },
+    {
+      "epoch": 0.18334081052788995,
+      "grad_norm": 1.3685090802839712,
+      "learning_rate": 9.396337014170866e-06,
+      "loss": 0.1872,
+      "step": 1226
+    },
+    {
+      "epoch": 0.18349035441902198,
+      "grad_norm": 1.6709772065779387,
+      "learning_rate": 9.395182913785192e-06,
+      "loss": 0.2055,
+      "step": 1227
+    },
+    {
+      "epoch": 0.18363989831015404,
+      "grad_norm": 2.0418194880673783,
+      "learning_rate": 9.394027782247247e-06,
+      "loss": 0.4888,
+      "step": 1228
+    },
+    {
+      "epoch": 0.18378944220128607,
+      "grad_norm": 1.5794839342981186,
+      "learning_rate": 9.392871619828036e-06,
+      "loss": 0.3355,
+      "step": 1229
+    },
+    {
+      "epoch": 0.18393898609241813,
+      "grad_norm": 2.365767436986478,
+      "learning_rate": 9.39171442679881e-06,
+      "loss": 0.4306,
+      "step": 1230
+    },
+    {
+      "epoch": 0.18408852998355016,
+      "grad_norm": 1.535735557296357,
+      "learning_rate": 9.390556203431053e-06,
+      "loss": 0.3454,
+      "step": 1231
+    },
+    {
+      "epoch": 0.18423807387468222,
+      "grad_norm": 2.0146640105762,
+      "learning_rate": 9.3893969499965e-06,
+      "loss": 0.5002,
+      "step": 1232
+    },
+    {
+      "epoch": 0.18438761776581428,
+      "grad_norm": 1.0888630229716356,
+      "learning_rate": 9.388236666767119e-06,
+      "loss": 0.1717,
+      "step": 1233
+    },
+    {
+      "epoch": 0.1845371616569463,
+      "grad_norm": 1.471926551369625,
+      "learning_rate": 9.387075354015125e-06,
+      "loss": 0.2728,
+      "step": 1234
+    },
+    {
+      "epoch": 0.18468670554807837,
+      "grad_norm": 1.2418392055984802,
+      "learning_rate": 9.385913012012972e-06,
+      "loss": 0.2338,
+      "step": 1235
+    },
+    {
+      "epoch": 0.1848362494392104,
+      "grad_norm": 1.1326547586847213,
+      "learning_rate": 9.384749641033358e-06,
+      "loss": 0.2014,
+      "step": 1236
+    },
+    {
+      "epoch": 0.18498579333034246,
+      "grad_norm": 1.2625669973249032,
+      "learning_rate": 9.383585241349223e-06,
+      "loss": 0.2257,
+      "step": 1237
+    },
+    {
+      "epoch": 0.18513533722147452,
+      "grad_norm": 2.0231610702160494,
+      "learning_rate": 9.382419813233741e-06,
+      "loss": 0.6136,
+      "step": 1238
+    },
+    {
+      "epoch": 0.18528488111260655,
+      "grad_norm": 2.2265632194384035,
+      "learning_rate": 9.381253356960339e-06,
+      "loss": 0.379,
+      "step": 1239
+    },
+    {
+      "epoch": 0.1854344250037386,
+      "grad_norm": 1.7519589257117685,
+      "learning_rate": 9.380085872802672e-06,
+      "loss": 0.4481,
+      "step": 1240
+    },
+    {
+      "epoch": 0.18558396889487064,
+      "grad_norm": 1.8894247538731719,
+      "learning_rate": 9.37891736103465e-06,
+      "loss": 0.2349,
+      "step": 1241
+    },
+    {
+      "epoch": 0.1857335127860027,
+      "grad_norm": 2.0840730120928153,
+      "learning_rate": 9.377747821930411e-06,
+      "loss": 0.386,
+      "step": 1242
+    },
+    {
+      "epoch": 0.18588305667713473,
+      "grad_norm": 1.2303960642463392,
+      "learning_rate": 9.376577255764346e-06,
+      "loss": 0.2138,
+      "step": 1243
+    },
+    {
+      "epoch": 0.1860326005682668,
+      "grad_norm": 1.0736052443136495,
+      "learning_rate": 9.375405662811076e-06,
+      "loss": 0.2919,
+      "step": 1244
+    },
+    {
+      "epoch": 0.18618214445939885,
+      "grad_norm": 2.1694546083973236,
+      "learning_rate": 9.37423304334547e-06,
+      "loss": 0.4716,
+      "step": 1245
+    },
+    {
+      "epoch": 0.18633168835053088,
+      "grad_norm": 1.7953994969561728,
+      "learning_rate": 9.373059397642637e-06,
+      "loss": 0.2303,
+      "step": 1246
+    },
+    {
+      "epoch": 0.18648123224166294,
+      "grad_norm": 1.1331346308690267,
+      "learning_rate": 9.371884725977924e-06,
+      "loss": 0.1681,
+      "step": 1247
+    },
+    {
+      "epoch": 0.18663077613279497,
+      "grad_norm": 1.8818511328803789,
+      "learning_rate": 9.370709028626921e-06,
+      "loss": 0.3736,
+      "step": 1248
+    },
+    {
+      "epoch": 0.18678032002392703,
+      "grad_norm": 1.080778831083804,
+      "learning_rate": 9.369532305865459e-06,
+      "loss": 0.2155,
+      "step": 1249
+    },
+    {
+      "epoch": 0.18692986391505906,
+      "grad_norm": 2.005915788639095,
+      "learning_rate": 9.368354557969606e-06,
+      "loss": 0.4026,
+      "step": 1250
+    },
+    {
+      "epoch": 0.18707940780619112,
+      "grad_norm": 0.8756048068588032,
+      "learning_rate": 9.367175785215674e-06,
+      "loss": 0.183,
+      "step": 1251
+    },
+    {
+      "epoch": 0.18722895169732318,
+      "grad_norm": 2.100099144522435,
+      "learning_rate": 9.365995987880216e-06,
+      "loss": 0.182,
+      "step": 1252
+    },
+    {
+      "epoch": 0.1873784955884552,
+      "grad_norm": 1.9441741117291806,
+      "learning_rate": 9.364815166240023e-06,
+      "loss": 0.3865,
+      "step": 1253
+    },
+    {
+      "epoch": 0.18752803947958727,
+      "grad_norm": 1.1762687046527927,
+      "learning_rate": 9.363633320572124e-06,
+      "loss": 0.2105,
+      "step": 1254
+    },
+    {
+      "epoch": 0.1876775833707193,
+      "grad_norm": 1.170057500642311,
+      "learning_rate": 9.362450451153795e-06,
+      "loss": 0.201,
+      "step": 1255
+    },
+    {
+      "epoch": 0.18782712726185136,
+      "grad_norm": 1.774823231357248,
+      "learning_rate": 9.36126655826255e-06,
+      "loss": 0.2958,
+      "step": 1256
+    },
+    {
+      "epoch": 0.1879766711529834,
+      "grad_norm": 1.847386943393164,
+      "learning_rate": 9.360081642176137e-06,
+      "loss": 0.2783,
+      "step": 1257
+    },
+    {
+      "epoch": 0.18812621504411545,
+      "grad_norm": 1.2317043700510546,
+      "learning_rate": 9.358895703172552e-06,
+      "loss": 0.2237,
+      "step": 1258
+    },
+    {
+      "epoch": 0.1882757589352475,
+      "grad_norm": 1.4962863598933458,
+      "learning_rate": 9.357708741530025e-06,
+      "loss": 0.208,
+      "step": 1259
+    },
+    {
+      "epoch": 0.18842530282637954,
+      "grad_norm": 1.753807685308467,
+      "learning_rate": 9.356520757527032e-06,
+      "loss": 0.513,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1885748467175116,
+      "grad_norm": 2.457507671022133,
+      "learning_rate": 9.355331751442284e-06,
+      "loss": 0.8743,
+      "step": 1261
+    },
+    {
+      "epoch": 0.18872439060864363,
+      "grad_norm": 1.7444181168119555,
+      "learning_rate": 9.354141723554734e-06,
+      "loss": 0.3346,
+      "step": 1262
+    },
+    {
+      "epoch": 0.1888739344997757,
+      "grad_norm": 1.971056965005781,
+      "learning_rate": 9.35295067414357e-06,
+      "loss": 0.2297,
+      "step": 1263
+    },
+    {
+      "epoch": 0.18902347839090772,
+      "grad_norm": 1.5052086349314306,
+      "learning_rate": 9.35175860348823e-06,
+      "loss": 0.2149,
+      "step": 1264
+    },
+    {
+      "epoch": 0.18917302228203978,
+      "grad_norm": 1.4361302390685748,
+      "learning_rate": 9.35056551186838e-06,
+      "loss": 0.3298,
+      "step": 1265
+    },
+    {
+      "epoch": 0.18932256617317184,
+      "grad_norm": 1.4206462492110938,
+      "learning_rate": 9.349371399563935e-06,
+      "loss": 0.1929,
+      "step": 1266
+    },
+    {
+      "epoch": 0.18947211006430387,
+      "grad_norm": 1.4913953281160535,
+      "learning_rate": 9.348176266855042e-06,
+      "loss": 0.2526,
+      "step": 1267
+    },
+    {
+      "epoch": 0.18962165395543593,
+      "grad_norm": 2.365270322972236,
+      "learning_rate": 9.346980114022092e-06,
+      "loss": 0.5066,
+      "step": 1268
+    },
+    {
+      "epoch": 0.18977119784656796,
+      "grad_norm": 1.8955954404187068,
+      "learning_rate": 9.345782941345714e-06,
+      "loss": 0.4404,
+      "step": 1269
+    },
+    {
+      "epoch": 0.18992074173770002,
+      "grad_norm": 1.1285554099802715,
+      "learning_rate": 9.344584749106775e-06,
+      "loss": 0.2001,
+      "step": 1270
+    },
+    {
+      "epoch": 0.19007028562883208,
+      "grad_norm": 1.6210163029014748,
+      "learning_rate": 9.343385537586385e-06,
+      "loss": 0.3274,
+      "step": 1271
+    },
+    {
+      "epoch": 0.1902198295199641,
+      "grad_norm": 1.4031028340124463,
+      "learning_rate": 9.342185307065888e-06,
+      "loss": 0.1922,
+      "step": 1272
+    },
+    {
+      "epoch": 0.19036937341109617,
+      "grad_norm": 1.5025368685887945,
+      "learning_rate": 9.340984057826872e-06,
+      "loss": 0.4106,
+      "step": 1273
+    },
+    {
+      "epoch": 0.1905189173022282,
+      "grad_norm": 1.3363573267962257,
+      "learning_rate": 9.339781790151159e-06,
+      "loss": 0.2906,
+      "step": 1274
+    },
+    {
+      "epoch": 0.19066846119336026,
+      "grad_norm": 2.2033082601743263,
+      "learning_rate": 9.338578504320815e-06,
+      "loss": 0.3913,
+      "step": 1275
+    },
+    {
+      "epoch": 0.1908180050844923,
+      "grad_norm": 1.5703985946217345,
+      "learning_rate": 9.337374200618141e-06,
+      "loss": 0.2363,
+      "step": 1276
+    },
+    {
+      "epoch": 0.19096754897562435,
+      "grad_norm": 1.8441964019229968,
+      "learning_rate": 9.336168879325678e-06,
+      "loss": 0.2193,
+      "step": 1277
+    },
+    {
+      "epoch": 0.1911170928667564,
+      "grad_norm": 1.9461041413852502,
+      "learning_rate": 9.334962540726208e-06,
+      "loss": 0.3327,
+      "step": 1278
+    },
+    {
+      "epoch": 0.19126663675788844,
+      "grad_norm": 1.987695873853033,
+      "learning_rate": 9.333755185102747e-06,
+      "loss": 0.5218,
+      "step": 1279
+    },
+    {
+      "epoch": 0.1914161806490205,
+      "grad_norm": 1.9784687710756435,
+      "learning_rate": 9.332546812738555e-06,
+      "loss": 0.4903,
+      "step": 1280
+    },
+    {
+      "epoch": 0.19156572454015253,
+      "grad_norm": 1.6276484190964966,
+      "learning_rate": 9.331337423917126e-06,
+      "loss": 0.3464,
+      "step": 1281
+    },
+    {
+      "epoch": 0.19171526843128459,
+      "grad_norm": 1.2032949092333924,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 0.1529,
+      "step": 1282
+    },
+    {
+      "epoch": 0.19186481232241662,
+      "grad_norm": 1.8200067255136916,
+      "learning_rate": 9.328915598037733e-06,
+      "loss": 0.4354,
+      "step": 1283
+    },
+    {
+      "epoch": 0.19201435621354868,
+      "grad_norm": 1.4794611609702433,
+      "learning_rate": 9.327703161547952e-06,
+      "loss": 0.2071,
+      "step": 1284
+    },
+    {
+      "epoch": 0.19216390010468073,
+      "grad_norm": 1.8032485542741101,
+      "learning_rate": 9.326489709737303e-06,
+      "loss": 0.3813,
+      "step": 1285
+    },
+    {
+      "epoch": 0.19231344399581277,
+      "grad_norm": 1.3455083426481262,
+      "learning_rate": 9.325275242890472e-06,
+      "loss": 0.1853,
+      "step": 1286
+    },
+    {
+      "epoch": 0.19246298788694483,
+      "grad_norm": 2.125860082250447,
+      "learning_rate": 9.324059761292385e-06,
+      "loss": 0.71,
+      "step": 1287
+    },
+    {
+      "epoch": 0.19261253177807686,
+      "grad_norm": 1.8089276946794224,
+      "learning_rate": 9.322843265228206e-06,
+      "loss": 0.3672,
+      "step": 1288
+    },
+    {
+      "epoch": 0.19276207566920892,
+      "grad_norm": 1.8361099385383872,
+      "learning_rate": 9.321625754983335e-06,
+      "loss": 0.3484,
+      "step": 1289
+    },
+    {
+      "epoch": 0.19291161956034095,
+      "grad_norm": 1.6363642315445044,
+      "learning_rate": 9.320407230843413e-06,
+      "loss": 0.3042,
+      "step": 1290
+    },
+    {
+      "epoch": 0.193061163451473,
+      "grad_norm": 2.021061654973304,
+      "learning_rate": 9.319187693094318e-06,
+      "loss": 0.5033,
+      "step": 1291
+    },
+    {
+      "epoch": 0.19321070734260506,
+      "grad_norm": 2.62527535830696,
+      "learning_rate": 9.317967142022163e-06,
+      "loss": 0.275,
+      "step": 1292
+    },
+    {
+      "epoch": 0.1933602512337371,
+      "grad_norm": 1.3806620083144838,
+      "learning_rate": 9.316745577913304e-06,
+      "loss": 0.2855,
+      "step": 1293
+    },
+    {
+      "epoch": 0.19350979512486916,
+      "grad_norm": 1.7655797800670596,
+      "learning_rate": 9.31552300105433e-06,
+      "loss": 0.3915,
+      "step": 1294
+    },
+    {
+      "epoch": 0.1936593390160012,
+      "grad_norm": 1.3465049850252158,
+      "learning_rate": 9.314299411732069e-06,
+      "loss": 0.249,
+      "step": 1295
+    },
+    {
+      "epoch": 0.19380888290713325,
+      "grad_norm": 0.8256201784583667,
+      "learning_rate": 9.313074810233589e-06,
+      "loss": 0.1543,
+      "step": 1296
+    },
+    {
+      "epoch": 0.1939584267982653,
+      "grad_norm": 1.9261313307855723,
+      "learning_rate": 9.31184919684619e-06,
+      "loss": 0.6008,
+      "step": 1297
+    },
+    {
+      "epoch": 0.19410797068939734,
+      "grad_norm": 1.519619673139573,
+      "learning_rate": 9.310622571857417e-06,
+      "loss": 0.239,
+      "step": 1298
+    },
+    {
+      "epoch": 0.1942575145805294,
+      "grad_norm": 2.0122490785681717,
+      "learning_rate": 9.309394935555042e-06,
+      "loss": 0.33,
+      "step": 1299
+    },
+    {
+      "epoch": 0.19440705847166143,
+      "grad_norm": 1.5990433489122537,
+      "learning_rate": 9.308166288227088e-06,
+      "loss": 0.4012,
+      "step": 1300
+    },
+    {
+      "epoch": 0.19455660236279348,
+      "grad_norm": 1.4956175342537672,
+      "learning_rate": 9.3069366301618e-06,
+      "loss": 0.5736,
+      "step": 1301
+    },
+    {
+      "epoch": 0.19470614625392552,
+      "grad_norm": 1.2896015261249874,
+      "learning_rate": 9.305705961647672e-06,
+      "loss": 0.1798,
+      "step": 1302
+    },
+    {
+      "epoch": 0.19485569014505758,
+      "grad_norm": 1.3006798401099697,
+      "learning_rate": 9.304474282973432e-06,
+      "loss": 0.3653,
+      "step": 1303
+    },
+    {
+      "epoch": 0.19500523403618963,
+      "grad_norm": 2.1249741454515054,
+      "learning_rate": 9.30324159442804e-06,
+      "loss": 0.7342,
+      "step": 1304
+    },
+    {
+      "epoch": 0.19515477792732167,
+      "grad_norm": 1.486234854981151,
+      "learning_rate": 9.302007896300697e-06,
+      "loss": 0.2874,
+      "step": 1305
+    },
+    {
+      "epoch": 0.19530432181845372,
+      "grad_norm": 1.486723968689139,
+      "learning_rate": 9.300773188880843e-06,
+      "loss": 0.2301,
+      "step": 1306
+    },
+    {
+      "epoch": 0.19545386570958576,
+      "grad_norm": 1.697087725237096,
+      "learning_rate": 9.29953747245815e-06,
+      "loss": 0.3191,
+      "step": 1307
+    },
+    {
+      "epoch": 0.19560340960071781,
+      "grad_norm": 1.9495679450825656,
+      "learning_rate": 9.29830074732253e-06,
+      "loss": 0.3954,
+      "step": 1308
+    },
+    {
+      "epoch": 0.19575295349184985,
+      "grad_norm": 1.2105140299371033,
+      "learning_rate": 9.29706301376413e-06,
+      "loss": 0.2557,
+      "step": 1309
+    },
+    {
+      "epoch": 0.1959024973829819,
+      "grad_norm": 1.547760007057872,
+      "learning_rate": 9.295824272073334e-06,
+      "loss": 0.2865,
+      "step": 1310
+    },
+    {
+      "epoch": 0.19605204127411396,
+      "grad_norm": 1.6181428409490188,
+      "learning_rate": 9.294584522540766e-06,
+      "loss": 0.3332,
+      "step": 1311
+    },
+    {
+      "epoch": 0.196201585165246,
+      "grad_norm": 1.664852192256293,
+      "learning_rate": 9.293343765457278e-06,
+      "loss": 0.3058,
+      "step": 1312
+    },
+    {
+      "epoch": 0.19635112905637805,
+      "grad_norm": 1.8608018825705885,
+      "learning_rate": 9.292102001113968e-06,
+      "loss": 0.3048,
+      "step": 1313
+    },
+    {
+      "epoch": 0.19650067294751009,
+      "grad_norm": 1.420503009424543,
+      "learning_rate": 9.290859229802162e-06,
+      "loss": 0.2283,
+      "step": 1314
+    },
+    {
+      "epoch": 0.19665021683864214,
+      "grad_norm": 1.996771180524021,
+      "learning_rate": 9.289615451813428e-06,
+      "loss": 0.4804,
+      "step": 1315
+    },
+    {
+      "epoch": 0.19679976072977418,
+      "grad_norm": 1.8295883207210475,
+      "learning_rate": 9.28837066743957e-06,
+      "loss": 0.3065,
+      "step": 1316
+    },
+    {
+      "epoch": 0.19694930462090623,
+      "grad_norm": 1.6508456488855519,
+      "learning_rate": 9.287124876972625e-06,
+      "loss": 0.2617,
+      "step": 1317
+    },
+    {
+      "epoch": 0.1970988485120383,
+      "grad_norm": 1.7646395203323395,
+      "learning_rate": 9.285878080704866e-06,
+      "loss": 0.3484,
+      "step": 1318
+    },
+    {
+      "epoch": 0.19724839240317033,
+      "grad_norm": 1.6976643786387164,
+      "learning_rate": 9.284630278928805e-06,
+      "loss": 0.4485,
+      "step": 1319
+    },
+    {
+      "epoch": 0.19739793629430238,
+      "grad_norm": 1.6867112197107144,
+      "learning_rate": 9.283381471937188e-06,
+      "loss": 0.381,
+      "step": 1320
+    },
+    {
+      "epoch": 0.19754748018543442,
+      "grad_norm": 1.650278888960391,
+      "learning_rate": 9.282131660022997e-06,
+      "loss": 0.2289,
+      "step": 1321
+    },
+    {
+      "epoch": 0.19769702407656647,
+      "grad_norm": 2.0028177208667977,
+      "learning_rate": 9.28088084347945e-06,
+      "loss": 0.5132,
+      "step": 1322
+    },
+    {
+      "epoch": 0.19784656796769853,
+      "grad_norm": 1.030996633782416,
+      "learning_rate": 9.279629022600002e-06,
+      "loss": 0.1764,
+      "step": 1323
+    },
+    {
+      "epoch": 0.19799611185883056,
+      "grad_norm": 1.448960209222983,
+      "learning_rate": 9.27837619767834e-06,
+      "loss": 0.2575,
+      "step": 1324
+    },
+    {
+      "epoch": 0.19814565574996262,
+      "grad_norm": 2.099657510604881,
+      "learning_rate": 9.27712236900839e-06,
+      "loss": 0.4117,
+      "step": 1325
+    },
+    {
+      "epoch": 0.19829519964109465,
+      "grad_norm": 1.1059240433107884,
+      "learning_rate": 9.27586753688431e-06,
+      "loss": 0.2064,
+      "step": 1326
+    },
+    {
+      "epoch": 0.19844474353222671,
+      "grad_norm": 0.9819176984170996,
+      "learning_rate": 9.274611701600502e-06,
+      "loss": 0.2357,
+      "step": 1327
+    },
+    {
+      "epoch": 0.19859428742335875,
+      "grad_norm": 1.5836465571763443,
+      "learning_rate": 9.273354863451589e-06,
+      "loss": 0.2478,
+      "step": 1328
+    },
+    {
+      "epoch": 0.1987438313144908,
+      "grad_norm": 1.8864795778055325,
+      "learning_rate": 9.272097022732444e-06,
+      "loss": 0.3705,
+      "step": 1329
+    },
+    {
+      "epoch": 0.19889337520562286,
+      "grad_norm": 1.5474710953559745,
+      "learning_rate": 9.270838179738164e-06,
+      "loss": 0.1888,
+      "step": 1330
+    },
+    {
+      "epoch": 0.1990429190967549,
+      "grad_norm": 1.6239105754270915,
+      "learning_rate": 9.269578334764087e-06,
+      "loss": 0.3698,
+      "step": 1331
+    },
+    {
+      "epoch": 0.19919246298788695,
+      "grad_norm": 1.033656840947032,
+      "learning_rate": 9.268317488105787e-06,
+      "loss": 0.1741,
+      "step": 1332
+    },
+    {
+      "epoch": 0.19934200687901898,
+      "grad_norm": 1.2644053752133695,
+      "learning_rate": 9.267055640059068e-06,
+      "loss": 0.3292,
+      "step": 1333
+    },
+    {
+      "epoch": 0.19949155077015104,
+      "grad_norm": 2.0362957566742224,
+      "learning_rate": 9.265792790919972e-06,
+      "loss": 0.3348,
+      "step": 1334
+    },
+    {
+      "epoch": 0.19964109466128308,
+      "grad_norm": 1.528100233715919,
+      "learning_rate": 9.264528940984777e-06,
+      "loss": 0.2456,
+      "step": 1335
+    },
+    {
+      "epoch": 0.19979063855241513,
+      "grad_norm": 1.420463259410139,
+      "learning_rate": 9.263264090549992e-06,
+      "loss": 0.3396,
+      "step": 1336
+    },
+    {
+      "epoch": 0.1999401824435472,
+      "grad_norm": 1.7091828149298964,
+      "learning_rate": 9.261998239912367e-06,
+      "loss": 0.2596,
+      "step": 1337
+    },
+    {
+      "epoch": 0.20008972633467922,
+      "grad_norm": 2.259480708170105,
+      "learning_rate": 9.26073138936888e-06,
+      "loss": 0.4451,
+      "step": 1338
+    },
+    {
+      "epoch": 0.20023927022581128,
+      "grad_norm": 1.8137629934472514,
+      "learning_rate": 9.259463539216746e-06,
+      "loss": 0.3549,
+      "step": 1339
+    },
+    {
+      "epoch": 0.20038881411694331,
+      "grad_norm": 1.1443459601678743,
+      "learning_rate": 9.258194689753417e-06,
+      "loss": 0.228,
+      "step": 1340
+    },
+    {
+      "epoch": 0.20053835800807537,
+      "grad_norm": 1.9827124471950097,
+      "learning_rate": 9.256924841276576e-06,
+      "loss": 0.2773,
+      "step": 1341
+    },
+    {
+      "epoch": 0.2006879018992074,
+      "grad_norm": 1.2970841523500378,
+      "learning_rate": 9.25565399408414e-06,
+      "loss": 0.2164,
+      "step": 1342
+    },
+    {
+      "epoch": 0.20083744579033946,
+      "grad_norm": 1.8476680005205972,
+      "learning_rate": 9.254382148474264e-06,
+      "loss": 0.3472,
+      "step": 1343
+    },
+    {
+      "epoch": 0.20098698968147152,
+      "grad_norm": 0.9750041200477407,
+      "learning_rate": 9.253109304745335e-06,
+      "loss": 0.1686,
+      "step": 1344
+    },
+    {
+      "epoch": 0.20113653357260355,
+      "grad_norm": 1.9603557255590418,
+      "learning_rate": 9.251835463195977e-06,
+      "loss": 0.5067,
+      "step": 1345
+    },
+    {
+      "epoch": 0.2012860774637356,
+      "grad_norm": 1.8365618678908586,
+      "learning_rate": 9.25056062412504e-06,
+      "loss": 0.3701,
+      "step": 1346
+    },
+    {
+      "epoch": 0.20143562135486764,
+      "grad_norm": 1.7722339453250509,
+      "learning_rate": 9.249284787831617e-06,
+      "loss": 0.4201,
+      "step": 1347
+    },
+    {
+      "epoch": 0.2015851652459997,
+      "grad_norm": 1.6882388762311373,
+      "learning_rate": 9.24800795461503e-06,
+      "loss": 0.3566,
+      "step": 1348
+    },
+    {
+      "epoch": 0.20173470913713173,
+      "grad_norm": 1.3408931481790511,
+      "learning_rate": 9.246730124774839e-06,
+      "loss": 0.3711,
+      "step": 1349
+    },
+    {
+      "epoch": 0.2018842530282638,
+      "grad_norm": 1.5904902378620338,
+      "learning_rate": 9.245451298610833e-06,
+      "loss": 0.2285,
+      "step": 1350
+    },
+    {
+      "epoch": 0.20203379691939585,
+      "grad_norm": 1.24091223117027,
+      "learning_rate": 9.244171476423037e-06,
+      "loss": 0.1831,
+      "step": 1351
+    },
+    {
+      "epoch": 0.20218334081052788,
+      "grad_norm": 1.8606049117343595,
+      "learning_rate": 9.24289065851171e-06,
+      "loss": 0.2355,
+      "step": 1352
+    },
+    {
+      "epoch": 0.20233288470165994,
+      "grad_norm": 1.3471226340640614,
+      "learning_rate": 9.241608845177344e-06,
+      "loss": 0.2296,
+      "step": 1353
+    },
+    {
+      "epoch": 0.20248242859279197,
+      "grad_norm": 1.8826806706856583,
+      "learning_rate": 9.240326036720665e-06,
+      "loss": 0.3794,
+      "step": 1354
+    },
+    {
+      "epoch": 0.20263197248392403,
+      "grad_norm": 1.6131643098193205,
+      "learning_rate": 9.239042233442632e-06,
+      "loss": 0.2487,
+      "step": 1355
+    },
+    {
+      "epoch": 0.2027815163750561,
+      "grad_norm": 1.250046228507658,
+      "learning_rate": 9.23775743564444e-06,
+      "loss": 0.362,
+      "step": 1356
+    },
+    {
+      "epoch": 0.20293106026618812,
+      "grad_norm": 1.680459140251771,
+      "learning_rate": 9.236471643627512e-06,
+      "loss": 0.218,
+      "step": 1357
+    },
+    {
+      "epoch": 0.20308060415732018,
+      "grad_norm": 3.806087040662127,
+      "learning_rate": 9.235184857693506e-06,
+      "loss": 0.3938,
+      "step": 1358
+    },
+    {
+      "epoch": 0.2032301480484522,
+      "grad_norm": 1.7243993707977876,
+      "learning_rate": 9.233897078144317e-06,
+      "loss": 0.2151,
+      "step": 1359
+    },
+    {
+      "epoch": 0.20337969193958427,
+      "grad_norm": 1.529478064822457,
+      "learning_rate": 9.23260830528207e-06,
+      "loss": 0.3384,
+      "step": 1360
+    },
+    {
+      "epoch": 0.2035292358307163,
+      "grad_norm": 1.4829330919870423,
+      "learning_rate": 9.231318539409124e-06,
+      "loss": 0.3183,
+      "step": 1361
+    },
+    {
+      "epoch": 0.20367877972184836,
+      "grad_norm": 1.332269401180316,
+      "learning_rate": 9.23002778082807e-06,
+      "loss": 0.1772,
+      "step": 1362
+    },
+    {
+      "epoch": 0.20382832361298042,
+      "grad_norm": 1.0915215645001923,
+      "learning_rate": 9.228736029841732e-06,
+      "loss": 0.1997,
+      "step": 1363
+    },
+    {
+      "epoch": 0.20397786750411245,
+      "grad_norm": 1.159153234988241,
+      "learning_rate": 9.227443286753167e-06,
+      "loss": 0.1916,
+      "step": 1364
+    },
+    {
+      "epoch": 0.2041274113952445,
+      "grad_norm": 1.7356076242142402,
+      "learning_rate": 9.226149551865665e-06,
+      "loss": 0.3072,
+      "step": 1365
+    },
+    {
+      "epoch": 0.20427695528637654,
+      "grad_norm": 1.4127677237560496,
+      "learning_rate": 9.224854825482752e-06,
+      "loss": 0.2194,
+      "step": 1366
+    },
+    {
+      "epoch": 0.2044264991775086,
+      "grad_norm": 1.5370329143503136,
+      "learning_rate": 9.223559107908178e-06,
+      "loss": 0.3882,
+      "step": 1367
+    },
+    {
+      "epoch": 0.20457604306864063,
+      "grad_norm": 1.231450996718572,
+      "learning_rate": 9.222262399445934e-06,
+      "loss": 0.2097,
+      "step": 1368
+    },
+    {
+      "epoch": 0.2047255869597727,
+      "grad_norm": 2.132300752161669,
+      "learning_rate": 9.22096470040024e-06,
+      "loss": 0.5167,
+      "step": 1369
+    },
+    {
+      "epoch": 0.20487513085090475,
+      "grad_norm": 1.6056336366949036,
+      "learning_rate": 9.219666011075548e-06,
+      "loss": 0.4712,
+      "step": 1370
+    },
+    {
+      "epoch": 0.20502467474203678,
+      "grad_norm": 1.0274444121008708,
+      "learning_rate": 9.218366331776543e-06,
+      "loss": 0.1489,
+      "step": 1371
+    },
+    {
+      "epoch": 0.20517421863316884,
+      "grad_norm": 1.7166617308295589,
+      "learning_rate": 9.217065662808143e-06,
+      "loss": 0.2071,
+      "step": 1372
+    },
+    {
+      "epoch": 0.20532376252430087,
+      "grad_norm": 1.678765123523581,
+      "learning_rate": 9.215764004475496e-06,
+      "loss": 0.3746,
+      "step": 1373
+    },
+    {
+      "epoch": 0.20547330641543293,
+      "grad_norm": 1.5537365180594542,
+      "learning_rate": 9.214461357083986e-06,
+      "loss": 0.1505,
+      "step": 1374
+    },
+    {
+      "epoch": 0.20562285030656496,
+      "grad_norm": 1.3435785882872704,
+      "learning_rate": 9.213157720939226e-06,
+      "loss": 0.1997,
+      "step": 1375
+    },
+    {
+      "epoch": 0.20577239419769702,
+      "grad_norm": 1.5875355486569163,
+      "learning_rate": 9.211853096347059e-06,
+      "loss": 0.1984,
+      "step": 1376
+    },
+    {
+      "epoch": 0.20592193808882908,
+      "grad_norm": 1.1101394692319788,
+      "learning_rate": 9.210547483613566e-06,
+      "loss": 0.1921,
+      "step": 1377
+    },
+    {
+      "epoch": 0.2060714819799611,
+      "grad_norm": 1.4017863241620425,
+      "learning_rate": 9.209240883045054e-06,
+      "loss": 0.2018,
+      "step": 1378
+    },
+    {
+      "epoch": 0.20622102587109317,
+      "grad_norm": 1.7309276589079257,
+      "learning_rate": 9.207933294948064e-06,
+      "loss": 0.2325,
+      "step": 1379
+    },
+    {
+      "epoch": 0.2063705697622252,
+      "grad_norm": 2.2693408682674647,
+      "learning_rate": 9.206624719629371e-06,
+      "loss": 0.6731,
+      "step": 1380
+    },
+    {
+      "epoch": 0.20652011365335726,
+      "grad_norm": 1.6617142222154033,
+      "learning_rate": 9.205315157395978e-06,
+      "loss": 0.2133,
+      "step": 1381
+    },
+    {
+      "epoch": 0.20666965754448932,
+      "grad_norm": 1.5651100627218977,
+      "learning_rate": 9.20400460855512e-06,
+      "loss": 0.3562,
+      "step": 1382
+    },
+    {
+      "epoch": 0.20681920143562135,
+      "grad_norm": 1.412248222099086,
+      "learning_rate": 9.202693073414267e-06,
+      "loss": 0.2117,
+      "step": 1383
+    },
+    {
+      "epoch": 0.2069687453267534,
+      "grad_norm": 2.058721397249849,
+      "learning_rate": 9.201380552281114e-06,
+      "loss": 0.4157,
+      "step": 1384
+    },
+    {
+      "epoch": 0.20711828921788544,
+      "grad_norm": 1.190322717831582,
+      "learning_rate": 9.200067045463594e-06,
+      "loss": 0.1468,
+      "step": 1385
+    },
+    {
+      "epoch": 0.2072678331090175,
+      "grad_norm": 1.6727715120686848,
+      "learning_rate": 9.198752553269867e-06,
+      "loss": 0.3737,
+      "step": 1386
+    },
+    {
+      "epoch": 0.20741737700014953,
+      "grad_norm": 1.6151445864315066,
+      "learning_rate": 9.197437076008328e-06,
+      "loss": 0.3504,
+      "step": 1387
+    },
+    {
+      "epoch": 0.2075669208912816,
+      "grad_norm": 1.8626136343488717,
+      "learning_rate": 9.196120613987596e-06,
+      "loss": 0.1897,
+      "step": 1388
+    },
+    {
+      "epoch": 0.20771646478241365,
+      "grad_norm": 2.1277783512048165,
+      "learning_rate": 9.19480316751653e-06,
+      "loss": 0.5398,
+      "step": 1389
+    },
+    {
+      "epoch": 0.20786600867354568,
+      "grad_norm": 1.6518405571968426,
+      "learning_rate": 9.193484736904214e-06,
+      "loss": 0.3226,
+      "step": 1390
+    },
+    {
+      "epoch": 0.20801555256467774,
+      "grad_norm": 1.680879037198174,
+      "learning_rate": 9.192165322459965e-06,
+      "loss": 0.2825,
+      "step": 1391
+    },
+    {
+      "epoch": 0.20816509645580977,
+      "grad_norm": 1.7490131840067196,
+      "learning_rate": 9.19084492449333e-06,
+      "loss": 0.4683,
+      "step": 1392
+    },
+    {
+      "epoch": 0.20831464034694183,
+      "grad_norm": 1.3086537105430815,
+      "learning_rate": 9.189523543314087e-06,
+      "loss": 0.2554,
+      "step": 1393
+    },
+    {
+      "epoch": 0.20846418423807386,
+      "grad_norm": 12.365317544881876,
+      "learning_rate": 9.188201179232243e-06,
+      "loss": 0.2051,
+      "step": 1394
+    },
+    {
+      "epoch": 0.20861372812920592,
+      "grad_norm": 1.6315580687346722,
+      "learning_rate": 9.18687783255804e-06,
+      "loss": 0.2119,
+      "step": 1395
+    },
+    {
+      "epoch": 0.20876327202033798,
+      "grad_norm": 1.7045742133633315,
+      "learning_rate": 9.185553503601948e-06,
+      "loss": 0.2251,
+      "step": 1396
+    },
+    {
+      "epoch": 0.20891281591147,
+      "grad_norm": 1.6188256816772855,
+      "learning_rate": 9.184228192674667e-06,
+      "loss": 0.2406,
+      "step": 1397
+    },
+    {
+      "epoch": 0.20906235980260207,
+      "grad_norm": 3.936101301721377,
+      "learning_rate": 9.182901900087124e-06,
+      "loss": 0.4307,
+      "step": 1398
+    },
+    {
+      "epoch": 0.2092119036937341,
+      "grad_norm": 2.5527420452443406,
+      "learning_rate": 9.181574626150486e-06,
+      "loss": 0.2308,
+      "step": 1399
+    },
+    {
+      "epoch": 0.20936144758486616,
+      "grad_norm": 1.6355845634467614,
+      "learning_rate": 9.180246371176141e-06,
+      "loss": 0.1983,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2095109914759982,
+      "grad_norm": 1.6442195705711569,
+      "learning_rate": 9.17891713547571e-06,
+      "loss": 0.4733,
+      "step": 1401
+    },
+    {
+      "epoch": 0.20966053536713025,
+      "grad_norm": 1.9389155277617691,
+      "learning_rate": 9.177586919361043e-06,
+      "loss": 0.3677,
+      "step": 1402
+    },
+    {
+      "epoch": 0.2098100792582623,
+      "grad_norm": 1.4858954046606485,
+      "learning_rate": 9.176255723144227e-06,
+      "loss": 0.2122,
+      "step": 1403
+    },
+    {
+      "epoch": 0.20995962314939434,
+      "grad_norm": 1.8989896403075806,
+      "learning_rate": 9.17492354713757e-06,
+      "loss": 0.5022,
+      "step": 1404
+    },
+    {
+      "epoch": 0.2101091670405264,
+      "grad_norm": 2.032698049116247,
+      "learning_rate": 9.173590391653612e-06,
+      "loss": 0.5431,
+      "step": 1405
+    },
+    {
+      "epoch": 0.21025871093165843,
+      "grad_norm": 1.5701626022875448,
+      "learning_rate": 9.172256257005127e-06,
+      "loss": 0.1862,
+      "step": 1406
+    },
+    {
+      "epoch": 0.2104082548227905,
+      "grad_norm": 1.9707343228520016,
+      "learning_rate": 9.170921143505114e-06,
+      "loss": 0.5544,
+      "step": 1407
+    },
+    {
+      "epoch": 0.21055779871392252,
+      "grad_norm": 1.3116097098782349,
+      "learning_rate": 9.169585051466804e-06,
+      "loss": 0.2191,
+      "step": 1408
+    },
+    {
+      "epoch": 0.21070734260505458,
+      "grad_norm": 1.4155318018990113,
+      "learning_rate": 9.168247981203657e-06,
+      "loss": 0.2867,
+      "step": 1409
+    },
+    {
+      "epoch": 0.21085688649618664,
+      "grad_norm": 1.5783535607698234,
+      "learning_rate": 9.166909933029365e-06,
+      "loss": 0.2115,
+      "step": 1410
+    },
+    {
+      "epoch": 0.21100643038731867,
+      "grad_norm": 2.1761826396642343,
+      "learning_rate": 9.16557090725784e-06,
+      "loss": 0.6031,
+      "step": 1411
+    },
+    {
+      "epoch": 0.21115597427845073,
+      "grad_norm": 1.5953173737260011,
+      "learning_rate": 9.16423090420324e-06,
+      "loss": 0.2972,
+      "step": 1412
+    },
+    {
+      "epoch": 0.21130551816958276,
+      "grad_norm": 1.859606383856977,
+      "learning_rate": 9.162889924179934e-06,
+      "loss": 0.3625,
+      "step": 1413
+    },
+    {
+      "epoch": 0.21145506206071482,
+      "grad_norm": 1.759605112492795,
+      "learning_rate": 9.161547967502536e-06,
+      "loss": 0.3425,
+      "step": 1414
+    },
+    {
+      "epoch": 0.21160460595184688,
+      "grad_norm": 1.2621512719829986,
+      "learning_rate": 9.160205034485875e-06,
+      "loss": 0.2218,
+      "step": 1415
+    },
+    {
+      "epoch": 0.2117541498429789,
+      "grad_norm": 1.4111741267825535,
+      "learning_rate": 9.158861125445022e-06,
+      "loss": 0.3193,
+      "step": 1416
+    },
+    {
+      "epoch": 0.21190369373411097,
+      "grad_norm": 1.2698781251191589,
+      "learning_rate": 9.157516240695266e-06,
+      "loss": 0.1909,
+      "step": 1417
+    },
+    {
+      "epoch": 0.212053237625243,
+      "grad_norm": 1.6960651115905963,
+      "learning_rate": 9.156170380552134e-06,
+      "loss": 0.3382,
+      "step": 1418
+    },
+    {
+      "epoch": 0.21220278151637506,
+      "grad_norm": 2.1154692881317434,
+      "learning_rate": 9.154823545331376e-06,
+      "loss": 0.4082,
+      "step": 1419
+    },
+    {
+      "epoch": 0.2123523254075071,
+      "grad_norm": 2.033684826459006,
+      "learning_rate": 9.153475735348973e-06,
+      "loss": 0.3867,
+      "step": 1420
+    },
+    {
+      "epoch": 0.21250186929863915,
+      "grad_norm": 1.8325048944634288,
+      "learning_rate": 9.152126950921135e-06,
+      "loss": 0.2373,
+      "step": 1421
+    },
+    {
+      "epoch": 0.2126514131897712,
+      "grad_norm": 2.0710673737402256,
+      "learning_rate": 9.150777192364297e-06,
+      "loss": 0.356,
+      "step": 1422
+    },
+    {
+      "epoch": 0.21280095708090324,
+      "grad_norm": 1.3874235860299686,
+      "learning_rate": 9.149426459995127e-06,
+      "loss": 0.3691,
+      "step": 1423
+    },
+    {
+      "epoch": 0.2129505009720353,
+      "grad_norm": 2.1249487008185803,
+      "learning_rate": 9.14807475413052e-06,
+      "loss": 0.3657,
+      "step": 1424
+    },
+    {
+      "epoch": 0.21310004486316733,
+      "grad_norm": 1.6076473156790783,
+      "learning_rate": 9.146722075087599e-06,
+      "loss": 0.2973,
+      "step": 1425
+    },
+    {
+      "epoch": 0.2132495887542994,
+      "grad_norm": 1.7585364821746667,
+      "learning_rate": 9.145368423183716e-06,
+      "loss": 0.3684,
+      "step": 1426
+    },
+    {
+      "epoch": 0.21339913264543142,
+      "grad_norm": 1.5724475968533793,
+      "learning_rate": 9.144013798736451e-06,
+      "loss": 0.3614,
+      "step": 1427
+    },
+    {
+      "epoch": 0.21354867653656348,
+      "grad_norm": 1.5213037170574124,
+      "learning_rate": 9.142658202063613e-06,
+      "loss": 0.331,
+      "step": 1428
+    },
+    {
+      "epoch": 0.21369822042769554,
+      "grad_norm": 1.7462862535903572,
+      "learning_rate": 9.141301633483233e-06,
+      "loss": 0.2972,
+      "step": 1429
+    },
+    {
+      "epoch": 0.21384776431882757,
+      "grad_norm": 1.3980558796390394,
+      "learning_rate": 9.139944093313582e-06,
+      "loss": 0.2375,
+      "step": 1430
+    },
+    {
+      "epoch": 0.21399730820995963,
+      "grad_norm": 1.655121853169513,
+      "learning_rate": 9.138585581873145e-06,
+      "loss": 0.1952,
+      "step": 1431
+    },
+    {
+      "epoch": 0.21414685210109166,
+      "grad_norm": 1.9616282412950032,
+      "learning_rate": 9.137226099480649e-06,
+      "loss": 0.3827,
+      "step": 1432
+    },
+    {
+      "epoch": 0.21429639599222372,
+      "grad_norm": 1.5101599924640094,
+      "learning_rate": 9.135865646455035e-06,
+      "loss": 0.2151,
+      "step": 1433
+    },
+    {
+      "epoch": 0.21444593988335575,
+      "grad_norm": 1.9492564570386166,
+      "learning_rate": 9.134504223115483e-06,
+      "loss": 0.5627,
+      "step": 1434
+    },
+    {
+      "epoch": 0.2145954837744878,
+      "grad_norm": 1.4722879693007511,
+      "learning_rate": 9.133141829781396e-06,
+      "loss": 0.3629,
+      "step": 1435
+    },
+    {
+      "epoch": 0.21474502766561987,
+      "grad_norm": 1.3336058131154513,
+      "learning_rate": 9.131778466772401e-06,
+      "loss": 0.2227,
+      "step": 1436
+    },
+    {
+      "epoch": 0.2148945715567519,
+      "grad_norm": 1.557620867064145,
+      "learning_rate": 9.130414134408358e-06,
+      "loss": 0.1922,
+      "step": 1437
+    },
+    {
+      "epoch": 0.21504411544788396,
+      "grad_norm": 1.5608371994510732,
+      "learning_rate": 9.129048833009354e-06,
+      "loss": 0.345,
+      "step": 1438
+    },
+    {
+      "epoch": 0.215193659339016,
+      "grad_norm": 2.0372244298598026,
+      "learning_rate": 9.127682562895701e-06,
+      "loss": 0.1943,
+      "step": 1439
+    },
+    {
+      "epoch": 0.21534320323014805,
+      "grad_norm": 1.771798287939486,
+      "learning_rate": 9.126315324387937e-06,
+      "loss": 0.2201,
+      "step": 1440
+    },
+    {
+      "epoch": 0.2154927471212801,
+      "grad_norm": 1.320918694554743,
+      "learning_rate": 9.124947117806833e-06,
+      "loss": 0.3557,
+      "step": 1441
+    },
+    {
+      "epoch": 0.21564229101241214,
+      "grad_norm": 1.2144354045037664,
+      "learning_rate": 9.12357794347338e-06,
+      "loss": 0.1677,
+      "step": 1442
+    },
+    {
+      "epoch": 0.2157918349035442,
+      "grad_norm": 1.7291539866895038,
+      "learning_rate": 9.122207801708802e-06,
+      "loss": 0.2361,
+      "step": 1443
+    },
+    {
+      "epoch": 0.21594137879467623,
+      "grad_norm": 1.525451199822209,
+      "learning_rate": 9.120836692834547e-06,
+      "loss": 0.2203,
+      "step": 1444
+    },
+    {
+      "epoch": 0.2160909226858083,
+      "grad_norm": 1.1450235200011314,
+      "learning_rate": 9.11946461717229e-06,
+      "loss": 0.2239,
+      "step": 1445
+    },
+    {
+      "epoch": 0.21624046657694032,
+      "grad_norm": 1.6249409995005226,
+      "learning_rate": 9.118091575043931e-06,
+      "loss": 0.2035,
+      "step": 1446
+    },
+    {
+      "epoch": 0.21639001046807238,
+      "grad_norm": 1.5911197134726236,
+      "learning_rate": 9.116717566771602e-06,
+      "loss": 0.3426,
+      "step": 1447
+    },
+    {
+      "epoch": 0.21653955435920444,
+      "grad_norm": 1.2768942365723772,
+      "learning_rate": 9.115342592677658e-06,
+      "loss": 0.3227,
+      "step": 1448
+    },
+    {
+      "epoch": 0.21668909825033647,
+      "grad_norm": 2.6013675708313646,
+      "learning_rate": 9.11396665308468e-06,
+      "loss": 0.876,
+      "step": 1449
+    },
+    {
+      "epoch": 0.21683864214146853,
+      "grad_norm": 1.5557280944558294,
+      "learning_rate": 9.112589748315477e-06,
+      "loss": 0.1862,
+      "step": 1450
+    },
+    {
+      "epoch": 0.21698818603260056,
+      "grad_norm": 1.9084211867181637,
+      "learning_rate": 9.111211878693084e-06,
+      "loss": 0.4744,
+      "step": 1451
+    },
+    {
+      "epoch": 0.21713772992373262,
+      "grad_norm": 1.9726349404572066,
+      "learning_rate": 9.109833044540766e-06,
+      "loss": 0.4964,
+      "step": 1452
+    },
+    {
+      "epoch": 0.21728727381486465,
+      "grad_norm": 1.7981363223292532,
+      "learning_rate": 9.108453246182005e-06,
+      "loss": 0.4617,
+      "step": 1453
+    },
+    {
+      "epoch": 0.2174368177059967,
+      "grad_norm": 1.4636175949338748,
+      "learning_rate": 9.10707248394052e-06,
+      "loss": 0.3562,
+      "step": 1454
+    },
+    {
+      "epoch": 0.21758636159712877,
+      "grad_norm": 1.5023091294876543,
+      "learning_rate": 9.105690758140247e-06,
+      "loss": 0.1974,
+      "step": 1455
+    },
+    {
+      "epoch": 0.2177359054882608,
+      "grad_norm": 1.5799772562807408,
+      "learning_rate": 9.104308069105355e-06,
+      "loss": 0.4225,
+      "step": 1456
+    },
+    {
+      "epoch": 0.21788544937939286,
+      "grad_norm": 1.697173297157486,
+      "learning_rate": 9.102924417160235e-06,
+      "loss": 0.2355,
+      "step": 1457
+    },
+    {
+      "epoch": 0.2180349932705249,
+      "grad_norm": 1.9831963966350166,
+      "learning_rate": 9.101539802629506e-06,
+      "loss": 0.6711,
+      "step": 1458
+    },
+    {
+      "epoch": 0.21818453716165695,
+      "grad_norm": 1.3049711330279425,
+      "learning_rate": 9.10015422583801e-06,
+      "loss": 0.2306,
+      "step": 1459
+    },
+    {
+      "epoch": 0.21833408105278898,
+      "grad_norm": 1.3485657787858605,
+      "learning_rate": 9.09876768711082e-06,
+      "loss": 0.3155,
+      "step": 1460
+    },
+    {
+      "epoch": 0.21848362494392104,
+      "grad_norm": 1.430084626746487,
+      "learning_rate": 9.097380186773225e-06,
+      "loss": 0.1803,
+      "step": 1461
+    },
+    {
+      "epoch": 0.2186331688350531,
+      "grad_norm": 1.042217037319963,
+      "learning_rate": 9.095991725150755e-06,
+      "loss": 0.2157,
+      "step": 1462
+    },
+    {
+      "epoch": 0.21878271272618513,
+      "grad_norm": 1.6038911196277919,
+      "learning_rate": 9.094602302569149e-06,
+      "loss": 0.3647,
+      "step": 1463
+    },
+    {
+      "epoch": 0.2189322566173172,
+      "grad_norm": 1.761061421005298,
+      "learning_rate": 9.093211919354384e-06,
+      "loss": 0.3914,
+      "step": 1464
+    },
+    {
+      "epoch": 0.21908180050844922,
+      "grad_norm": 1.780140598214583,
+      "learning_rate": 9.091820575832653e-06,
+      "loss": 0.5239,
+      "step": 1465
+    },
+    {
+      "epoch": 0.21923134439958128,
+      "grad_norm": 1.5056065150417166,
+      "learning_rate": 9.090428272330381e-06,
+      "loss": 0.3738,
+      "step": 1466
+    },
+    {
+      "epoch": 0.2193808882907133,
+      "grad_norm": 1.5037400785516828,
+      "learning_rate": 9.089035009174213e-06,
+      "loss": 0.3748,
+      "step": 1467
+    },
+    {
+      "epoch": 0.21953043218184537,
+      "grad_norm": 1.0034616725342143,
+      "learning_rate": 9.087640786691029e-06,
+      "loss": 0.1818,
+      "step": 1468
+    },
+    {
+      "epoch": 0.21967997607297743,
+      "grad_norm": 1.8826776828221814,
+      "learning_rate": 9.08624560520792e-06,
+      "loss": 0.5423,
+      "step": 1469
+    },
+    {
+      "epoch": 0.21982951996410946,
+      "grad_norm": 1.5762662138471795,
+      "learning_rate": 9.08484946505221e-06,
+      "loss": 0.362,
+      "step": 1470
+    },
+    {
+      "epoch": 0.21997906385524152,
+      "grad_norm": 1.459026627139429,
+      "learning_rate": 9.08345236655145e-06,
+      "loss": 0.2379,
+      "step": 1471
+    },
+    {
+      "epoch": 0.22012860774637355,
+      "grad_norm": 1.9767735560822688,
+      "learning_rate": 9.082054310033412e-06,
+      "loss": 0.2279,
+      "step": 1472
+    },
+    {
+      "epoch": 0.2202781516375056,
+      "grad_norm": 1.5033186272325123,
+      "learning_rate": 9.08065529582609e-06,
+      "loss": 0.3475,
+      "step": 1473
+    },
+    {
+      "epoch": 0.22042769552863767,
+      "grad_norm": 1.7347624363190264,
+      "learning_rate": 9.07925532425771e-06,
+      "loss": 0.2346,
+      "step": 1474
+    },
+    {
+      "epoch": 0.2205772394197697,
+      "grad_norm": 1.311368499943375,
+      "learning_rate": 9.077854395656719e-06,
+      "loss": 0.2136,
+      "step": 1475
+    },
+    {
+      "epoch": 0.22072678331090176,
+      "grad_norm": 2.079168631285795,
+      "learning_rate": 9.076452510351786e-06,
+      "loss": 0.5518,
+      "step": 1476
+    },
+    {
+      "epoch": 0.2208763272020338,
+      "grad_norm": 1.7471493510688818,
+      "learning_rate": 9.075049668671808e-06,
+      "loss": 0.4685,
+      "step": 1477
+    },
+    {
+      "epoch": 0.22102587109316585,
+      "grad_norm": 1.7503079479759727,
+      "learning_rate": 9.073645870945904e-06,
+      "loss": 0.3369,
+      "step": 1478
+    },
+    {
+      "epoch": 0.22117541498429788,
+      "grad_norm": 1.6001592489802432,
+      "learning_rate": 9.07224111750342e-06,
+      "loss": 0.2451,
+      "step": 1479
+    },
+    {
+      "epoch": 0.22132495887542994,
+      "grad_norm": 1.5995131942282934,
+      "learning_rate": 9.070835408673926e-06,
+      "loss": 0.2199,
+      "step": 1480
+    },
+    {
+      "epoch": 0.221474502766562,
+      "grad_norm": 1.4586326961701683,
+      "learning_rate": 9.06942874478721e-06,
+      "loss": 0.374,
+      "step": 1481
+    },
+    {
+      "epoch": 0.22162404665769403,
+      "grad_norm": 1.5381391273868525,
+      "learning_rate": 9.068021126173294e-06,
+      "loss": 0.2369,
+      "step": 1482
+    },
+    {
+      "epoch": 0.2217735905488261,
+      "grad_norm": 1.96520452808454,
+      "learning_rate": 9.066612553162417e-06,
+      "loss": 0.2214,
+      "step": 1483
+    },
+    {
+      "epoch": 0.22192313443995812,
+      "grad_norm": 1.3323608756885599,
+      "learning_rate": 9.065203026085041e-06,
+      "loss": 0.2335,
+      "step": 1484
+    },
+    {
+      "epoch": 0.22207267833109018,
+      "grad_norm": 1.725202352638352,
+      "learning_rate": 9.063792545271859e-06,
+      "loss": 0.4229,
+      "step": 1485
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 0.9584754362518301,
+      "learning_rate": 9.062381111053781e-06,
+      "loss": 0.1886,
+      "step": 1486
+    },
+    {
+      "epoch": 0.22237176611335427,
+      "grad_norm": 1.72245020942535,
+      "learning_rate": 9.060968723761945e-06,
+      "loss": 0.3353,
+      "step": 1487
+    },
+    {
+      "epoch": 0.22252131000448633,
+      "grad_norm": 1.6135392202141645,
+      "learning_rate": 9.05955538372771e-06,
+      "loss": 0.1953,
+      "step": 1488
+    },
+    {
+      "epoch": 0.22267085389561836,
+      "grad_norm": 1.3152021255142217,
+      "learning_rate": 9.058141091282656e-06,
+      "loss": 0.1742,
+      "step": 1489
+    },
+    {
+      "epoch": 0.22282039778675042,
+      "grad_norm": 2.024778297438531,
+      "learning_rate": 9.056725846758594e-06,
+      "loss": 0.4014,
+      "step": 1490
+    },
+    {
+      "epoch": 0.22296994167788245,
+      "grad_norm": 1.6698713908455967,
+      "learning_rate": 9.055309650487552e-06,
+      "loss": 0.3407,
+      "step": 1491
+    },
+    {
+      "epoch": 0.2231194855690145,
+      "grad_norm": 1.934590891729098,
+      "learning_rate": 9.053892502801783e-06,
+      "loss": 0.3714,
+      "step": 1492
+    },
+    {
+      "epoch": 0.22326902946014654,
+      "grad_norm": 1.8737430969121303,
+      "learning_rate": 9.052474404033764e-06,
+      "loss": 0.2273,
+      "step": 1493
+    },
+    {
+      "epoch": 0.2234185733512786,
+      "grad_norm": 2.1916032794360594,
+      "learning_rate": 9.051055354516195e-06,
+      "loss": 0.5141,
+      "step": 1494
+    },
+    {
+      "epoch": 0.22356811724241066,
+      "grad_norm": 1.3808614199352094,
+      "learning_rate": 9.049635354581998e-06,
+      "loss": 0.22,
+      "step": 1495
+    },
+    {
+      "epoch": 0.2237176611335427,
+      "grad_norm": 1.56662054094386,
+      "learning_rate": 9.048214404564319e-06,
+      "loss": 0.23,
+      "step": 1496
+    },
+    {
+      "epoch": 0.22386720502467475,
+      "grad_norm": 1.3046782458999366,
+      "learning_rate": 9.046792504796526e-06,
+      "loss": 0.2448,
+      "step": 1497
+    },
+    {
+      "epoch": 0.22401674891580678,
+      "grad_norm": 1.2799576312369438,
+      "learning_rate": 9.045369655612212e-06,
+      "loss": 0.222,
+      "step": 1498
+    },
+    {
+      "epoch": 0.22416629280693884,
+      "grad_norm": 1.3122547692582682,
+      "learning_rate": 9.043945857345189e-06,
+      "loss": 0.2261,
+      "step": 1499
+    },
+    {
+      "epoch": 0.2243158366980709,
+      "grad_norm": 1.644275160578279,
+      "learning_rate": 9.042521110329497e-06,
+      "loss": 0.3581,
+      "step": 1500
+    },
+    {
+      "epoch": 0.22446538058920293,
+      "grad_norm": 1.3566851036629488,
+      "learning_rate": 9.04109541489939e-06,
+      "loss": 0.3989,
+      "step": 1501
+    },
+    {
+      "epoch": 0.224614924480335,
+      "grad_norm": 1.1272465309050348,
+      "learning_rate": 9.039668771389356e-06,
+      "loss": 0.2049,
+      "step": 1502
+    },
+    {
+      "epoch": 0.22476446837146702,
+      "grad_norm": 1.2545829340424604,
+      "learning_rate": 9.038241180134095e-06,
+      "loss": 0.2186,
+      "step": 1503
+    },
+    {
+      "epoch": 0.22491401226259908,
+      "grad_norm": 2.458444877022837,
+      "learning_rate": 9.036812641468535e-06,
+      "loss": 0.9375,
+      "step": 1504
+    },
+    {
+      "epoch": 0.2250635561537311,
+      "grad_norm": 1.7631503038372196,
+      "learning_rate": 9.035383155727826e-06,
+      "loss": 0.5211,
+      "step": 1505
+    },
+    {
+      "epoch": 0.22521310004486317,
+      "grad_norm": 1.2773941362574963,
+      "learning_rate": 9.03395272324734e-06,
+      "loss": 0.1924,
+      "step": 1506
+    },
+    {
+      "epoch": 0.22536264393599523,
+      "grad_norm": 0.8285607729871721,
+      "learning_rate": 9.032521344362665e-06,
+      "loss": 0.1591,
+      "step": 1507
+    },
+    {
+      "epoch": 0.22551218782712726,
+      "grad_norm": 1.4527549144198617,
+      "learning_rate": 9.031089019409622e-06,
+      "loss": 0.3534,
+      "step": 1508
+    },
+    {
+      "epoch": 0.22566173171825932,
+      "grad_norm": 1.5654074226457095,
+      "learning_rate": 9.029655748724245e-06,
+      "loss": 0.3845,
+      "step": 1509
+    },
+    {
+      "epoch": 0.22581127560939135,
+      "grad_norm": 1.7456103899692976,
+      "learning_rate": 9.028221532642793e-06,
+      "loss": 0.3186,
+      "step": 1510
+    },
+    {
+      "epoch": 0.2259608195005234,
+      "grad_norm": 1.9962184366897087,
+      "learning_rate": 9.02678637150175e-06,
+      "loss": 0.4627,
+      "step": 1511
+    },
+    {
+      "epoch": 0.22611036339165544,
+      "grad_norm": 1.1115007945174045,
+      "learning_rate": 9.025350265637816e-06,
+      "loss": 0.2216,
+      "step": 1512
+    },
+    {
+      "epoch": 0.2262599072827875,
+      "grad_norm": 1.1002289464689736,
+      "learning_rate": 9.023913215387914e-06,
+      "loss": 0.26,
+      "step": 1513
+    },
+    {
+      "epoch": 0.22640945117391956,
+      "grad_norm": 4.707452449710372,
+      "learning_rate": 9.022475221089194e-06,
+      "loss": 0.2086,
+      "step": 1514
+    },
+    {
+      "epoch": 0.2265589950650516,
+      "grad_norm": 1.1907478811651444,
+      "learning_rate": 9.021036283079022e-06,
+      "loss": 0.1712,
+      "step": 1515
+    },
+    {
+      "epoch": 0.22670853895618365,
+      "grad_norm": 1.4943756417337315,
+      "learning_rate": 9.019596401694984e-06,
+      "loss": 0.3612,
+      "step": 1516
+    },
+    {
+      "epoch": 0.22685808284731568,
+      "grad_norm": 1.8328029211172867,
+      "learning_rate": 9.018155577274891e-06,
+      "loss": 0.5014,
+      "step": 1517
+    },
+    {
+      "epoch": 0.22700762673844774,
+      "grad_norm": 1.5769600203549734,
+      "learning_rate": 9.016713810156777e-06,
+      "loss": 0.3982,
+      "step": 1518
+    },
+    {
+      "epoch": 0.22715717062957977,
+      "grad_norm": 1.033998264540829,
+      "learning_rate": 9.015271100678892e-06,
+      "loss": 0.1698,
+      "step": 1519
+    },
+    {
+      "epoch": 0.22730671452071183,
+      "grad_norm": 1.4450333274104865,
+      "learning_rate": 9.01382744917971e-06,
+      "loss": 0.2241,
+      "step": 1520
+    },
+    {
+      "epoch": 0.2274562584118439,
+      "grad_norm": 1.4009561553962124,
+      "learning_rate": 9.012382855997925e-06,
+      "loss": 0.318,
+      "step": 1521
+    },
+    {
+      "epoch": 0.22760580230297592,
+      "grad_norm": 1.6380239076287348,
+      "learning_rate": 9.010937321472454e-06,
+      "loss": 0.3649,
+      "step": 1522
+    },
+    {
+      "epoch": 0.22775534619410798,
+      "grad_norm": 1.8718410186357894,
+      "learning_rate": 9.009490845942433e-06,
+      "loss": 0.3845,
+      "step": 1523
+    },
+    {
+      "epoch": 0.22790489008524,
+      "grad_norm": 1.474697832902752,
+      "learning_rate": 9.00804342974722e-06,
+      "loss": 0.3287,
+      "step": 1524
+    },
+    {
+      "epoch": 0.22805443397637207,
+      "grad_norm": 1.9033242818165663,
+      "learning_rate": 9.006595073226387e-06,
+      "loss": 0.3614,
+      "step": 1525
+    },
+    {
+      "epoch": 0.2282039778675041,
+      "grad_norm": 1.7576573706097778,
+      "learning_rate": 9.005145776719742e-06,
+      "loss": 0.4287,
+      "step": 1526
+    },
+    {
+      "epoch": 0.22835352175863616,
+      "grad_norm": 1.646435656567765,
+      "learning_rate": 9.003695540567294e-06,
+      "loss": 0.3619,
+      "step": 1527
+    },
+    {
+      "epoch": 0.22850306564976822,
+      "grad_norm": 1.6955358560275535,
+      "learning_rate": 9.00224436510929e-06,
+      "loss": 0.371,
+      "step": 1528
+    },
+    {
+      "epoch": 0.22865260954090025,
+      "grad_norm": 1.658825841648886,
+      "learning_rate": 9.000792250686186e-06,
+      "loss": 0.3939,
+      "step": 1529
+    },
+    {
+      "epoch": 0.2288021534320323,
+      "grad_norm": 1.5326543749267825,
+      "learning_rate": 8.999339197638664e-06,
+      "loss": 0.2482,
+      "step": 1530
+    },
+    {
+      "epoch": 0.22895169732316434,
+      "grad_norm": 2.4642466957027174,
+      "learning_rate": 8.99788520630762e-06,
+      "loss": 0.4219,
+      "step": 1531
+    },
+    {
+      "epoch": 0.2291012412142964,
+      "grad_norm": 1.532408533906101,
+      "learning_rate": 8.99643027703418e-06,
+      "loss": 0.3121,
+      "step": 1532
+    },
+    {
+      "epoch": 0.22925078510542846,
+      "grad_norm": 1.068785213809154,
+      "learning_rate": 8.994974410159682e-06,
+      "loss": 0.2221,
+      "step": 1533
+    },
+    {
+      "epoch": 0.2294003289965605,
+      "grad_norm": 1.6812794615422513,
+      "learning_rate": 8.993517606025686e-06,
+      "loss": 0.36,
+      "step": 1534
+    },
+    {
+      "epoch": 0.22954987288769255,
+      "grad_norm": 1.3397673899297329,
+      "learning_rate": 8.992059864973972e-06,
+      "loss": 0.1707,
+      "step": 1535
+    },
+    {
+      "epoch": 0.22969941677882458,
+      "grad_norm": 2.2086842094358445,
+      "learning_rate": 8.990601187346542e-06,
+      "loss": 0.2141,
+      "step": 1536
+    },
+    {
+      "epoch": 0.22984896066995664,
+      "grad_norm": 1.9813196209128292,
+      "learning_rate": 8.989141573485614e-06,
+      "loss": 0.3664,
+      "step": 1537
+    },
+    {
+      "epoch": 0.22999850456108867,
+      "grad_norm": 1.4968536528735594,
+      "learning_rate": 8.987681023733628e-06,
+      "loss": 0.212,
+      "step": 1538
+    },
+    {
+      "epoch": 0.23014804845222073,
+      "grad_norm": 2.005340468193955,
+      "learning_rate": 8.986219538433243e-06,
+      "loss": 0.5171,
+      "step": 1539
+    },
+    {
+      "epoch": 0.23029759234335279,
+      "grad_norm": 1.6056668375146956,
+      "learning_rate": 8.984757117927337e-06,
+      "loss": 0.367,
+      "step": 1540
+    },
+    {
+      "epoch": 0.23044713623448482,
+      "grad_norm": 1.736972659173659,
+      "learning_rate": 8.983293762559009e-06,
+      "loss": 0.3054,
+      "step": 1541
+    },
+    {
+      "epoch": 0.23059668012561688,
+      "grad_norm": 2.3554690220264893,
+      "learning_rate": 8.981829472671576e-06,
+      "loss": 0.48,
+      "step": 1542
+    },
+    {
+      "epoch": 0.2307462240167489,
+      "grad_norm": 2.441414392076267,
+      "learning_rate": 8.980364248608576e-06,
+      "loss": 0.5507,
+      "step": 1543
+    },
+    {
+      "epoch": 0.23089576790788097,
+      "grad_norm": 1.6179923206419418,
+      "learning_rate": 8.97889809071376e-06,
+      "loss": 0.1799,
+      "step": 1544
+    },
+    {
+      "epoch": 0.231045311799013,
+      "grad_norm": 1.3522900238471818,
+      "learning_rate": 8.977430999331108e-06,
+      "loss": 0.2205,
+      "step": 1545
+    },
+    {
+      "epoch": 0.23119485569014506,
+      "grad_norm": 1.5829130211417721,
+      "learning_rate": 8.97596297480481e-06,
+      "loss": 0.3255,
+      "step": 1546
+    },
+    {
+      "epoch": 0.23134439958127712,
+      "grad_norm": 1.5826039702225925,
+      "learning_rate": 8.974494017479281e-06,
+      "loss": 0.2593,
+      "step": 1547
+    },
+    {
+      "epoch": 0.23149394347240915,
+      "grad_norm": 1.498668084288196,
+      "learning_rate": 8.973024127699152e-06,
+      "loss": 0.3489,
+      "step": 1548
+    },
+    {
+      "epoch": 0.2316434873635412,
+      "grad_norm": 1.753546348637437,
+      "learning_rate": 8.971553305809274e-06,
+      "loss": 0.1926,
+      "step": 1549
+    },
+    {
+      "epoch": 0.23179303125467324,
+      "grad_norm": 1.434969286299199,
+      "learning_rate": 8.970081552154714e-06,
+      "loss": 0.3408,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2319425751458053,
+      "grad_norm": 0.8478352471965336,
+      "learning_rate": 8.968608867080761e-06,
+      "loss": 0.2296,
+      "step": 1551
+    },
+    {
+      "epoch": 0.23209211903693733,
+      "grad_norm": 1.5740700313091822,
+      "learning_rate": 8.967135250932921e-06,
+      "loss": 0.2804,
+      "step": 1552
+    },
+    {
+      "epoch": 0.2322416629280694,
+      "grad_norm": 1.5528025294287224,
+      "learning_rate": 8.96566070405692e-06,
+      "loss": 0.3515,
+      "step": 1553
+    },
+    {
+      "epoch": 0.23239120681920145,
+      "grad_norm": 1.6156309602690202,
+      "learning_rate": 8.964185226798696e-06,
+      "loss": 0.3846,
+      "step": 1554
+    },
+    {
+      "epoch": 0.23254075071033348,
+      "grad_norm": 1.900324205764928,
+      "learning_rate": 8.962708819504415e-06,
+      "loss": 0.3893,
+      "step": 1555
+    },
+    {
+      "epoch": 0.23269029460146554,
+      "grad_norm": 1.6871831737787155,
+      "learning_rate": 8.961231482520456e-06,
+      "loss": 0.4995,
+      "step": 1556
+    },
+    {
+      "epoch": 0.23283983849259757,
+      "grad_norm": 1.7659270863655478,
+      "learning_rate": 8.959753216193415e-06,
+      "loss": 0.3834,
+      "step": 1557
+    },
+    {
+      "epoch": 0.23298938238372963,
+      "grad_norm": 1.4355628771765856,
+      "learning_rate": 8.958274020870107e-06,
+      "loss": 0.2458,
+      "step": 1558
+    },
+    {
+      "epoch": 0.23313892627486169,
+      "grad_norm": 1.091740260422094,
+      "learning_rate": 8.956793896897566e-06,
+      "loss": 0.1625,
+      "step": 1559
+    },
+    {
+      "epoch": 0.23328847016599372,
+      "grad_norm": 1.1925103613752697,
+      "learning_rate": 8.955312844623045e-06,
+      "loss": 0.1767,
+      "step": 1560
+    },
+    {
+      "epoch": 0.23343801405712578,
+      "grad_norm": 1.8795430835063092,
+      "learning_rate": 8.953830864394012e-06,
+      "loss": 0.3983,
+      "step": 1561
+    },
+    {
+      "epoch": 0.2335875579482578,
+      "grad_norm": 1.4373449366320268,
+      "learning_rate": 8.952347956558152e-06,
+      "loss": 0.2384,
+      "step": 1562
+    },
+    {
+      "epoch": 0.23373710183938987,
+      "grad_norm": 1.4903754933686846,
+      "learning_rate": 8.950864121463374e-06,
+      "loss": 0.2191,
+      "step": 1563
+    },
+    {
+      "epoch": 0.2338866457305219,
+      "grad_norm": 1.8744944645931436,
+      "learning_rate": 8.949379359457795e-06,
+      "loss": 0.4788,
+      "step": 1564
+    },
+    {
+      "epoch": 0.23403618962165396,
+      "grad_norm": 2.30369719872738,
+      "learning_rate": 8.947893670889756e-06,
+      "loss": 0.5508,
+      "step": 1565
+    },
+    {
+      "epoch": 0.23418573351278602,
+      "grad_norm": 1.5228147309739217,
+      "learning_rate": 8.946407056107815e-06,
+      "loss": 0.2744,
+      "step": 1566
+    },
+    {
+      "epoch": 0.23433527740391805,
+      "grad_norm": 0.9855113226049245,
+      "learning_rate": 8.944919515460746e-06,
+      "loss": 0.1762,
+      "step": 1567
+    },
+    {
+      "epoch": 0.2344848212950501,
+      "grad_norm": 2.0244671473740476,
+      "learning_rate": 8.943431049297542e-06,
+      "loss": 0.4977,
+      "step": 1568
+    },
+    {
+      "epoch": 0.23463436518618214,
+      "grad_norm": 1.9053040838783037,
+      "learning_rate": 8.941941657967408e-06,
+      "loss": 0.3878,
+      "step": 1569
+    },
+    {
+      "epoch": 0.2347839090773142,
+      "grad_norm": 1.2593264871558218,
+      "learning_rate": 8.94045134181977e-06,
+      "loss": 0.2085,
+      "step": 1570
+    },
+    {
+      "epoch": 0.23493345296844623,
+      "grad_norm": 1.4409087053787786,
+      "learning_rate": 8.938960101204273e-06,
+      "loss": 0.392,
+      "step": 1571
+    },
+    {
+      "epoch": 0.23508299685957829,
+      "grad_norm": 0.9533079750475656,
+      "learning_rate": 8.937467936470773e-06,
+      "loss": 0.161,
+      "step": 1572
+    },
+    {
+      "epoch": 0.23523254075071034,
+      "grad_norm": 1.3187507218092258,
+      "learning_rate": 8.935974847969352e-06,
+      "loss": 0.228,
+      "step": 1573
+    },
+    {
+      "epoch": 0.23538208464184238,
+      "grad_norm": 1.4711338301621406,
+      "learning_rate": 8.934480836050297e-06,
+      "loss": 0.2065,
+      "step": 1574
+    },
+    {
+      "epoch": 0.23553162853297444,
+      "grad_norm": 1.6346673332024881,
+      "learning_rate": 8.93298590106412e-06,
+      "loss": 0.3639,
+      "step": 1575
+    },
+    {
+      "epoch": 0.23568117242410647,
+      "grad_norm": 1.519040249043761,
+      "learning_rate": 8.931490043361546e-06,
+      "loss": 0.3451,
+      "step": 1576
+    },
+    {
+      "epoch": 0.23583071631523853,
+      "grad_norm": 1.5939490670813292,
+      "learning_rate": 8.929993263293519e-06,
+      "loss": 0.3313,
+      "step": 1577
+    },
+    {
+      "epoch": 0.23598026020637056,
+      "grad_norm": 2.4862093840935584,
+      "learning_rate": 8.928495561211199e-06,
+      "loss": 0.5844,
+      "step": 1578
+    },
+    {
+      "epoch": 0.23612980409750262,
+      "grad_norm": 0.9227521449231808,
+      "learning_rate": 8.92699693746596e-06,
+      "loss": 0.1858,
+      "step": 1579
+    },
+    {
+      "epoch": 0.23627934798863467,
+      "grad_norm": 1.7904495917281569,
+      "learning_rate": 8.925497392409392e-06,
+      "loss": 0.4007,
+      "step": 1580
+    },
+    {
+      "epoch": 0.2364288918797667,
+      "grad_norm": 1.6104758349435173,
+      "learning_rate": 8.923996926393306e-06,
+      "loss": 0.2506,
+      "step": 1581
+    },
+    {
+      "epoch": 0.23657843577089877,
+      "grad_norm": 1.162466771602011,
+      "learning_rate": 8.922495539769722e-06,
+      "loss": 0.2251,
+      "step": 1582
+    },
+    {
+      "epoch": 0.2367279796620308,
+      "grad_norm": 1.4044418047951504,
+      "learning_rate": 8.920993232890885e-06,
+      "loss": 0.3529,
+      "step": 1583
+    },
+    {
+      "epoch": 0.23687752355316286,
+      "grad_norm": 1.3430242838204132,
+      "learning_rate": 8.919490006109243e-06,
+      "loss": 0.213,
+      "step": 1584
+    },
+    {
+      "epoch": 0.23702706744429491,
+      "grad_norm": 2.047091159617954,
+      "learning_rate": 8.917985859777477e-06,
+      "loss": 0.4145,
+      "step": 1585
+    },
+    {
+      "epoch": 0.23717661133542695,
+      "grad_norm": 2.0721630966321247,
+      "learning_rate": 8.916480794248465e-06,
+      "loss": 0.5021,
+      "step": 1586
+    },
+    {
+      "epoch": 0.237326155226559,
+      "grad_norm": 1.5816352819652573,
+      "learning_rate": 8.914974809875317e-06,
+      "loss": 0.3858,
+      "step": 1587
+    },
+    {
+      "epoch": 0.23747569911769104,
+      "grad_norm": 1.8550669988064394,
+      "learning_rate": 8.913467907011349e-06,
+      "loss": 0.4832,
+      "step": 1588
+    },
+    {
+      "epoch": 0.2376252430088231,
+      "grad_norm": 1.6366187681919886,
+      "learning_rate": 8.911960086010091e-06,
+      "loss": 0.3524,
+      "step": 1589
+    },
+    {
+      "epoch": 0.23777478689995513,
+      "grad_norm": 1.4416460700943408,
+      "learning_rate": 8.910451347225298e-06,
+      "loss": 0.3926,
+      "step": 1590
+    },
+    {
+      "epoch": 0.23792433079108719,
+      "grad_norm": 1.564579376384169,
+      "learning_rate": 8.908941691010933e-06,
+      "loss": 0.2312,
+      "step": 1591
+    },
+    {
+      "epoch": 0.23807387468221924,
+      "grad_norm": 1.2944939224073606,
+      "learning_rate": 8.907431117721175e-06,
+      "loss": 0.2245,
+      "step": 1592
+    },
+    {
+      "epoch": 0.23822341857335128,
+      "grad_norm": 1.6975131715816902,
+      "learning_rate": 8.905919627710419e-06,
+      "loss": 0.3363,
+      "step": 1593
+    },
+    {
+      "epoch": 0.23837296246448333,
+      "grad_norm": 1.596361094662525,
+      "learning_rate": 8.904407221333275e-06,
+      "loss": 0.3658,
+      "step": 1594
+    },
+    {
+      "epoch": 0.23852250635561537,
+      "grad_norm": 1.4858572761749074,
+      "learning_rate": 8.902893898944571e-06,
+      "loss": 0.2827,
+      "step": 1595
+    },
+    {
+      "epoch": 0.23867205024674742,
+      "grad_norm": 2.4979339586019464,
+      "learning_rate": 8.901379660899343e-06,
+      "loss": 0.5078,
+      "step": 1596
+    },
+    {
+      "epoch": 0.23882159413787946,
+      "grad_norm": 1.647641401011452,
+      "learning_rate": 8.899864507552846e-06,
+      "loss": 0.3907,
+      "step": 1597
+    },
+    {
+      "epoch": 0.23897113802901152,
+      "grad_norm": 1.604466675493429,
+      "learning_rate": 8.898348439260553e-06,
+      "loss": 0.4351,
+      "step": 1598
+    },
+    {
+      "epoch": 0.23912068192014357,
+      "grad_norm": 1.158380693426154,
+      "learning_rate": 8.896831456378148e-06,
+      "loss": 0.2045,
+      "step": 1599
+    },
+    {
+      "epoch": 0.2392702258112756,
+      "grad_norm": 1.6498746618436277,
+      "learning_rate": 8.895313559261525e-06,
+      "loss": 0.3361,
+      "step": 1600
+    },
+    {
+      "epoch": 0.23941976970240766,
+      "grad_norm": 1.237277087089687,
+      "learning_rate": 8.893794748266801e-06,
+      "loss": 0.2327,
+      "step": 1601
+    },
+    {
+      "epoch": 0.2395693135935397,
+      "grad_norm": 1.7276131272814355,
+      "learning_rate": 8.892275023750305e-06,
+      "loss": 0.3676,
+      "step": 1602
+    },
+    {
+      "epoch": 0.23971885748467175,
+      "grad_norm": 1.0737115286717451,
+      "learning_rate": 8.890754386068577e-06,
+      "loss": 0.1695,
+      "step": 1603
+    },
+    {
+      "epoch": 0.23986840137580379,
+      "grad_norm": 1.7893197216212124,
+      "learning_rate": 8.889232835578372e-06,
+      "loss": 0.2075,
+      "step": 1604
+    },
+    {
+      "epoch": 0.24001794526693584,
+      "grad_norm": 1.0316754586933763,
+      "learning_rate": 8.887710372636662e-06,
+      "loss": 0.1818,
+      "step": 1605
+    },
+    {
+      "epoch": 0.2401674891580679,
+      "grad_norm": 1.908936676793948,
+      "learning_rate": 8.886186997600633e-06,
+      "loss": 0.4014,
+      "step": 1606
+    },
+    {
+      "epoch": 0.24031703304919994,
+      "grad_norm": 1.8121753558090572,
+      "learning_rate": 8.884662710827679e-06,
+      "loss": 0.5739,
+      "step": 1607
+    },
+    {
+      "epoch": 0.240466576940332,
+      "grad_norm": 1.5081453030204104,
+      "learning_rate": 8.883137512675417e-06,
+      "loss": 0.2058,
+      "step": 1608
+    },
+    {
+      "epoch": 0.24061612083146403,
+      "grad_norm": 1.6968786810383643,
+      "learning_rate": 8.88161140350167e-06,
+      "loss": 0.3992,
+      "step": 1609
+    },
+    {
+      "epoch": 0.24076566472259608,
+      "grad_norm": 1.0441105843798988,
+      "learning_rate": 8.880084383664481e-06,
+      "loss": 0.1736,
+      "step": 1610
+    },
+    {
+      "epoch": 0.24091520861372812,
+      "grad_norm": 1.4531784381064643,
+      "learning_rate": 8.8785564535221e-06,
+      "loss": 0.2213,
+      "step": 1611
+    },
+    {
+      "epoch": 0.24106475250486017,
+      "grad_norm": 1.6179328389513437,
+      "learning_rate": 8.877027613432997e-06,
+      "loss": 0.4397,
+      "step": 1612
+    },
+    {
+      "epoch": 0.24121429639599223,
+      "grad_norm": 1.3779127226717234,
+      "learning_rate": 8.875497863755851e-06,
+      "loss": 0.2095,
+      "step": 1613
+    },
+    {
+      "epoch": 0.24136384028712426,
+      "grad_norm": 1.4488714989705522,
+      "learning_rate": 8.873967204849556e-06,
+      "loss": 0.3495,
+      "step": 1614
+    },
+    {
+      "epoch": 0.24151338417825632,
+      "grad_norm": 1.699118467638372,
+      "learning_rate": 8.872435637073223e-06,
+      "loss": 0.2042,
+      "step": 1615
+    },
+    {
+      "epoch": 0.24166292806938836,
+      "grad_norm": 1.5724198212446476,
+      "learning_rate": 8.870903160786166e-06,
+      "loss": 0.358,
+      "step": 1616
+    },
+    {
+      "epoch": 0.24181247196052041,
+      "grad_norm": 1.8735086152337022,
+      "learning_rate": 8.869369776347923e-06,
+      "loss": 0.352,
+      "step": 1617
+    },
+    {
+      "epoch": 0.24196201585165247,
+      "grad_norm": 2.071130867923683,
+      "learning_rate": 8.867835484118241e-06,
+      "loss": 0.369,
+      "step": 1618
+    },
+    {
+      "epoch": 0.2421115597427845,
+      "grad_norm": 1.00009701768564,
+      "learning_rate": 8.866300284457078e-06,
+      "loss": 0.1362,
+      "step": 1619
+    },
+    {
+      "epoch": 0.24226110363391656,
+      "grad_norm": 1.2342660250726913,
+      "learning_rate": 8.864764177724607e-06,
+      "loss": 0.228,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2424106475250486,
+      "grad_norm": 1.181191216278982,
+      "learning_rate": 8.863227164281214e-06,
+      "loss": 0.185,
+      "step": 1621
+    },
+    {
+      "epoch": 0.24256019141618065,
+      "grad_norm": 1.600687165751174,
+      "learning_rate": 8.861689244487497e-06,
+      "loss": 0.198,
+      "step": 1622
+    },
+    {
+      "epoch": 0.24270973530731269,
+      "grad_norm": 1.8325486581074766,
+      "learning_rate": 8.860150418704268e-06,
+      "loss": 0.3872,
+      "step": 1623
+    },
+    {
+      "epoch": 0.24285927919844474,
+      "grad_norm": 1.8934954141642377,
+      "learning_rate": 8.858610687292548e-06,
+      "loss": 0.4214,
+      "step": 1624
+    },
+    {
+      "epoch": 0.2430088230895768,
+      "grad_norm": 2.2505887638767184,
+      "learning_rate": 8.857070050613573e-06,
+      "loss": 0.2427,
+      "step": 1625
+    },
+    {
+      "epoch": 0.24315836698070883,
+      "grad_norm": 1.7156119494211053,
+      "learning_rate": 8.855528509028793e-06,
+      "loss": 0.3129,
+      "step": 1626
+    },
+    {
+      "epoch": 0.2433079108718409,
+      "grad_norm": 1.1926433487919907,
+      "learning_rate": 8.853986062899869e-06,
+      "loss": 0.1944,
+      "step": 1627
+    },
+    {
+      "epoch": 0.24345745476297292,
+      "grad_norm": 1.4955836838428689,
+      "learning_rate": 8.852442712588671e-06,
+      "loss": 0.4259,
+      "step": 1628
+    },
+    {
+      "epoch": 0.24360699865410498,
+      "grad_norm": 1.5899010145338972,
+      "learning_rate": 8.850898458457284e-06,
+      "loss": 0.1922,
+      "step": 1629
+    },
+    {
+      "epoch": 0.24375654254523701,
+      "grad_norm": 1.7661527081371415,
+      "learning_rate": 8.849353300868007e-06,
+      "loss": 0.3413,
+      "step": 1630
+    },
+    {
+      "epoch": 0.24390608643636907,
+      "grad_norm": 1.0197540346721858,
+      "learning_rate": 8.847807240183349e-06,
+      "loss": 0.2392,
+      "step": 1631
+    },
+    {
+      "epoch": 0.24405563032750113,
+      "grad_norm": 2.304126832944234,
+      "learning_rate": 8.84626027676603e-06,
+      "loss": 0.6652,
+      "step": 1632
+    },
+    {
+      "epoch": 0.24420517421863316,
+      "grad_norm": 1.730517756971552,
+      "learning_rate": 8.844712410978981e-06,
+      "loss": 0.2769,
+      "step": 1633
+    },
+    {
+      "epoch": 0.24435471810976522,
+      "grad_norm": 1.7941803588719758,
+      "learning_rate": 8.843163643185347e-06,
+      "loss": 0.5527,
+      "step": 1634
+    },
+    {
+      "epoch": 0.24450426200089725,
+      "grad_norm": 1.5034412891428997,
+      "learning_rate": 8.841613973748486e-06,
+      "loss": 0.2425,
+      "step": 1635
+    },
+    {
+      "epoch": 0.2446538058920293,
+      "grad_norm": 1.8387720627371986,
+      "learning_rate": 8.840063403031962e-06,
+      "loss": 0.4814,
+      "step": 1636
+    },
+    {
+      "epoch": 0.24480334978316134,
+      "grad_norm": 1.6058048815286128,
+      "learning_rate": 8.838511931399557e-06,
+      "loss": 0.205,
+      "step": 1637
+    },
+    {
+      "epoch": 0.2449528936742934,
+      "grad_norm": 1.6535570636461936,
+      "learning_rate": 8.83695955921526e-06,
+      "loss": 0.3842,
+      "step": 1638
+    },
+    {
+      "epoch": 0.24510243756542546,
+      "grad_norm": 2.0495436648956677,
+      "learning_rate": 8.83540628684327e-06,
+      "loss": 0.588,
+      "step": 1639
+    },
+    {
+      "epoch": 0.2452519814565575,
+      "grad_norm": 1.932080426112819,
+      "learning_rate": 8.833852114648006e-06,
+      "loss": 0.2566,
+      "step": 1640
+    },
+    {
+      "epoch": 0.24540152534768955,
+      "grad_norm": 1.9660531070561371,
+      "learning_rate": 8.832297042994083e-06,
+      "loss": 0.5196,
+      "step": 1641
+    },
+    {
+      "epoch": 0.24555106923882158,
+      "grad_norm": 1.4040557060017893,
+      "learning_rate": 8.830741072246343e-06,
+      "loss": 0.2041,
+      "step": 1642
+    },
+    {
+      "epoch": 0.24570061312995364,
+      "grad_norm": 2.1028771117617637,
+      "learning_rate": 8.829184202769828e-06,
+      "loss": 0.2795,
+      "step": 1643
+    },
+    {
+      "epoch": 0.2458501570210857,
+      "grad_norm": 1.379854611664084,
+      "learning_rate": 8.827626434929796e-06,
+      "loss": 0.1597,
+      "step": 1644
+    },
+    {
+      "epoch": 0.24599970091221773,
+      "grad_norm": 1.5617312208997327,
+      "learning_rate": 8.826067769091715e-06,
+      "loss": 0.3839,
+      "step": 1645
+    },
+    {
+      "epoch": 0.2461492448033498,
+      "grad_norm": 1.7647416850681912,
+      "learning_rate": 8.824508205621263e-06,
+      "loss": 0.373,
+      "step": 1646
+    },
+    {
+      "epoch": 0.24629878869448182,
+      "grad_norm": 2.0755613511333406,
+      "learning_rate": 8.822947744884326e-06,
+      "loss": 0.5092,
+      "step": 1647
+    },
+    {
+      "epoch": 0.24644833258561388,
+      "grad_norm": 2.235044525778246,
+      "learning_rate": 8.821386387247006e-06,
+      "loss": 0.6294,
+      "step": 1648
+    },
+    {
+      "epoch": 0.24659787647674591,
+      "grad_norm": 1.6548207522042184,
+      "learning_rate": 8.81982413307561e-06,
+      "loss": 0.2872,
+      "step": 1649
+    },
+    {
+      "epoch": 0.24674742036787797,
+      "grad_norm": 0.9460460785889475,
+      "learning_rate": 8.818260982736662e-06,
+      "loss": 0.2059,
+      "step": 1650
+    },
+    {
+      "epoch": 0.24689696425901003,
+      "grad_norm": 1.596549339690058,
+      "learning_rate": 8.816696936596887e-06,
+      "loss": 0.2146,
+      "step": 1651
+    },
+    {
+      "epoch": 0.24704650815014206,
+      "grad_norm": 1.7405302714522426,
+      "learning_rate": 8.815131995023228e-06,
+      "loss": 0.2358,
+      "step": 1652
+    },
+    {
+      "epoch": 0.24719605204127412,
+      "grad_norm": 1.71942999566363,
+      "learning_rate": 8.813566158382835e-06,
+      "loss": 0.4195,
+      "step": 1653
+    },
+    {
+      "epoch": 0.24734559593240615,
+      "grad_norm": 1.8118502915807313,
+      "learning_rate": 8.81199942704307e-06,
+      "loss": 0.3486,
+      "step": 1654
+    },
+    {
+      "epoch": 0.2474951398235382,
+      "grad_norm": 1.0992755654244406,
+      "learning_rate": 8.810431801371501e-06,
+      "loss": 0.2063,
+      "step": 1655
+    },
+    {
+      "epoch": 0.24764468371467024,
+      "grad_norm": 3.017300017050166,
+      "learning_rate": 8.80886328173591e-06,
+      "loss": 0.4337,
+      "step": 1656
+    },
+    {
+      "epoch": 0.2477942276058023,
+      "grad_norm": 1.6022895744312384,
+      "learning_rate": 8.807293868504282e-06,
+      "loss": 0.3729,
+      "step": 1657
+    },
+    {
+      "epoch": 0.24794377149693436,
+      "grad_norm": 1.6149891917484116,
+      "learning_rate": 8.805723562044825e-06,
+      "loss": 0.4056,
+      "step": 1658
+    },
+    {
+      "epoch": 0.2480933153880664,
+      "grad_norm": 1.6145150498338217,
+      "learning_rate": 8.80415236272594e-06,
+      "loss": 0.3016,
+      "step": 1659
+    },
+    {
+      "epoch": 0.24824285927919845,
+      "grad_norm": 1.1459038449035153,
+      "learning_rate": 8.80258027091625e-06,
+      "loss": 0.2218,
+      "step": 1660
+    },
+    {
+      "epoch": 0.24839240317033048,
+      "grad_norm": 1.591564655332456,
+      "learning_rate": 8.801007286984581e-06,
+      "loss": 0.245,
+      "step": 1661
+    },
+    {
+      "epoch": 0.24854194706146254,
+      "grad_norm": 1.7569520327773172,
+      "learning_rate": 8.799433411299971e-06,
+      "loss": 0.4247,
+      "step": 1662
+    },
+    {
+      "epoch": 0.24869149095259457,
+      "grad_norm": 1.064692259533788,
+      "learning_rate": 8.797858644231666e-06,
+      "loss": 0.2029,
+      "step": 1663
+    },
+    {
+      "epoch": 0.24884103484372663,
+      "grad_norm": 1.7207168422676138,
+      "learning_rate": 8.796282986149123e-06,
+      "loss": 0.4056,
+      "step": 1664
+    },
+    {
+      "epoch": 0.2489905787348587,
+      "grad_norm": 1.8670534021913596,
+      "learning_rate": 8.794706437422004e-06,
+      "loss": 0.4252,
+      "step": 1665
+    },
+    {
+      "epoch": 0.24914012262599072,
+      "grad_norm": 1.1847022204061899,
+      "learning_rate": 8.793128998420183e-06,
+      "loss": 0.2108,
+      "step": 1666
+    },
+    {
+      "epoch": 0.24928966651712278,
+      "grad_norm": 1.7975419112991038,
+      "learning_rate": 8.791550669513746e-06,
+      "loss": 0.335,
+      "step": 1667
+    },
+    {
+      "epoch": 0.2494392104082548,
+      "grad_norm": 1.5925842023054706,
+      "learning_rate": 8.789971451072979e-06,
+      "loss": 0.3249,
+      "step": 1668
+    },
+    {
+      "epoch": 0.24958875429938687,
+      "grad_norm": 1.6010634947496465,
+      "learning_rate": 8.788391343468385e-06,
+      "loss": 0.395,
+      "step": 1669
+    },
+    {
+      "epoch": 0.2497382981905189,
+      "grad_norm": 1.6989646502396285,
+      "learning_rate": 8.78681034707067e-06,
+      "loss": 0.2221,
+      "step": 1670
+    },
+    {
+      "epoch": 0.24988784208165096,
+      "grad_norm": 1.1982003401309609,
+      "learning_rate": 8.785228462250755e-06,
+      "loss": 0.1606,
+      "step": 1671
+    },
+    {
+      "epoch": 0.250037385972783,
+      "grad_norm": 1.504124635470649,
+      "learning_rate": 8.783645689379763e-06,
+      "loss": 0.2759,
+      "step": 1672
+    },
+    {
+      "epoch": 0.25018692986391505,
+      "grad_norm": 1.1944052488188541,
+      "learning_rate": 8.782062028829028e-06,
+      "loss": 0.1768,
+      "step": 1673
+    },
+    {
+      "epoch": 0.2503364737550471,
+      "grad_norm": 2.329098645094478,
+      "learning_rate": 8.78047748097009e-06,
+      "loss": 0.8662,
+      "step": 1674
+    },
+    {
+      "epoch": 0.25048601764617917,
+      "grad_norm": 1.0556820313003359,
+      "learning_rate": 8.778892046174703e-06,
+      "loss": 0.1784,
+      "step": 1675
+    },
+    {
+      "epoch": 0.2506355615373112,
+      "grad_norm": 2.158513791608413,
+      "learning_rate": 8.777305724814823e-06,
+      "loss": 0.2676,
+      "step": 1676
+    },
+    {
+      "epoch": 0.25078510542844323,
+      "grad_norm": 1.4537227026901378,
+      "learning_rate": 8.775718517262616e-06,
+      "loss": 0.3745,
+      "step": 1677
+    },
+    {
+      "epoch": 0.2509346493195753,
+      "grad_norm": 0.950798082633677,
+      "learning_rate": 8.774130423890457e-06,
+      "loss": 0.1824,
+      "step": 1678
+    },
+    {
+      "epoch": 0.25108419321070735,
+      "grad_norm": 1.545457884426641,
+      "learning_rate": 8.77254144507093e-06,
+      "loss": 0.3852,
+      "step": 1679
+    },
+    {
+      "epoch": 0.2512337371018394,
+      "grad_norm": 1.319903070406757,
+      "learning_rate": 8.770951581176819e-06,
+      "loss": 0.1928,
+      "step": 1680
+    },
+    {
+      "epoch": 0.2513832809929714,
+      "grad_norm": 1.6348677618098524,
+      "learning_rate": 8.769360832581127e-06,
+      "loss": 0.2614,
+      "step": 1681
+    },
+    {
+      "epoch": 0.2515328248841035,
+      "grad_norm": 1.2973642919936317,
+      "learning_rate": 8.767769199657056e-06,
+      "loss": 0.1926,
+      "step": 1682
+    },
+    {
+      "epoch": 0.25168236877523553,
+      "grad_norm": 2.0805063872419214,
+      "learning_rate": 8.766176682778021e-06,
+      "loss": 0.6252,
+      "step": 1683
+    },
+    {
+      "epoch": 0.25183191266636756,
+      "grad_norm": 1.5452777534643873,
+      "learning_rate": 8.76458328231764e-06,
+      "loss": 0.3641,
+      "step": 1684
+    },
+    {
+      "epoch": 0.25198145655749965,
+      "grad_norm": 2.031081999653712,
+      "learning_rate": 8.76298899864974e-06,
+      "loss": 0.6777,
+      "step": 1685
+    },
+    {
+      "epoch": 0.2521310004486317,
+      "grad_norm": 1.4412028187219246,
+      "learning_rate": 8.761393832148355e-06,
+      "loss": 0.3482,
+      "step": 1686
+    },
+    {
+      "epoch": 0.2522805443397637,
+      "grad_norm": 1.756325567561018,
+      "learning_rate": 8.759797783187728e-06,
+      "loss": 0.486,
+      "step": 1687
+    },
+    {
+      "epoch": 0.25243008823089574,
+      "grad_norm": 1.4179914068088109,
+      "learning_rate": 8.758200852142306e-06,
+      "loss": 0.3271,
+      "step": 1688
+    },
+    {
+      "epoch": 0.25257963212202783,
+      "grad_norm": 1.5841241720765276,
+      "learning_rate": 8.756603039386744e-06,
+      "loss": 0.2038,
+      "step": 1689
+    },
+    {
+      "epoch": 0.25272917601315986,
+      "grad_norm": 2.0246116663194687,
+      "learning_rate": 8.755004345295906e-06,
+      "loss": 0.5168,
+      "step": 1690
+    },
+    {
+      "epoch": 0.2528787199042919,
+      "grad_norm": 1.4410735414416371,
+      "learning_rate": 8.753404770244861e-06,
+      "loss": 0.2331,
+      "step": 1691
+    },
+    {
+      "epoch": 0.253028263795424,
+      "grad_norm": 1.6406911200868926,
+      "learning_rate": 8.751804314608885e-06,
+      "loss": 0.3777,
+      "step": 1692
+    },
+    {
+      "epoch": 0.253177807686556,
+      "grad_norm": 1.4803860988544733,
+      "learning_rate": 8.750202978763455e-06,
+      "loss": 0.3677,
+      "step": 1693
+    },
+    {
+      "epoch": 0.25332735157768804,
+      "grad_norm": 1.1771068735919579,
+      "learning_rate": 8.748600763084267e-06,
+      "loss": 0.3354,
+      "step": 1694
+    },
+    {
+      "epoch": 0.2534768954688201,
+      "grad_norm": 1.364747322095615,
+      "learning_rate": 8.746997667947215e-06,
+      "loss": 0.2101,
+      "step": 1695
+    },
+    {
+      "epoch": 0.25362643935995216,
+      "grad_norm": 1.0147659263108189,
+      "learning_rate": 8.745393693728395e-06,
+      "loss": 0.2104,
+      "step": 1696
+    },
+    {
+      "epoch": 0.2537759832510842,
+      "grad_norm": 1.3008901060796498,
+      "learning_rate": 8.74378884080412e-06,
+      "loss": 0.2224,
+      "step": 1697
+    },
+    {
+      "epoch": 0.2539255271422162,
+      "grad_norm": 1.5410622225873059,
+      "learning_rate": 8.742183109550906e-06,
+      "loss": 0.3175,
+      "step": 1698
+    },
+    {
+      "epoch": 0.2540750710333483,
+      "grad_norm": 1.1271133540339682,
+      "learning_rate": 8.740576500345465e-06,
+      "loss": 0.2078,
+      "step": 1699
+    },
+    {
+      "epoch": 0.25422461492448034,
+      "grad_norm": 1.5002212907557977,
+      "learning_rate": 8.73896901356473e-06,
+      "loss": 0.3457,
+      "step": 1700
+    },
+    {
+      "epoch": 0.2543741588156124,
+      "grad_norm": 1.5194450466740133,
+      "learning_rate": 8.737360649585831e-06,
+      "loss": 0.2974,
+      "step": 1701
+    },
+    {
+      "epoch": 0.2545237027067444,
+      "grad_norm": 1.541411347823068,
+      "learning_rate": 8.735751408786106e-06,
+      "loss": 0.1909,
+      "step": 1702
+    },
+    {
+      "epoch": 0.2546732465978765,
+      "grad_norm": 1.0383311616312054,
+      "learning_rate": 8.734141291543096e-06,
+      "loss": 0.1968,
+      "step": 1703
+    },
+    {
+      "epoch": 0.2548227904890085,
+      "grad_norm": 1.1554526724480272,
+      "learning_rate": 8.732530298234551e-06,
+      "loss": 0.3622,
+      "step": 1704
+    },
+    {
+      "epoch": 0.25497233438014055,
+      "grad_norm": 1.5935676507788412,
+      "learning_rate": 8.730918429238429e-06,
+      "loss": 0.4022,
+      "step": 1705
+    },
+    {
+      "epoch": 0.25512187827127264,
+      "grad_norm": 1.7826101439006568,
+      "learning_rate": 8.729305684932884e-06,
+      "loss": 0.4635,
+      "step": 1706
+    },
+    {
+      "epoch": 0.25527142216240467,
+      "grad_norm": 1.9360511884951532,
+      "learning_rate": 8.727692065696286e-06,
+      "loss": 0.5279,
+      "step": 1707
+    },
+    {
+      "epoch": 0.2554209660535367,
+      "grad_norm": 2.2688771091405355,
+      "learning_rate": 8.726077571907205e-06,
+      "loss": 0.4807,
+      "step": 1708
+    },
+    {
+      "epoch": 0.25557050994466873,
+      "grad_norm": 1.1500886235737136,
+      "learning_rate": 8.724462203944417e-06,
+      "loss": 0.1742,
+      "step": 1709
+    },
+    {
+      "epoch": 0.2557200538358008,
+      "grad_norm": 2.1581902257611505,
+      "learning_rate": 8.7228459621869e-06,
+      "loss": 0.2518,
+      "step": 1710
+    },
+    {
+      "epoch": 0.25586959772693285,
+      "grad_norm": 1.5158538301505384,
+      "learning_rate": 8.721228847013844e-06,
+      "loss": 0.4283,
+      "step": 1711
+    },
+    {
+      "epoch": 0.2560191416180649,
+      "grad_norm": 1.1144613603300342,
+      "learning_rate": 8.719610858804634e-06,
+      "loss": 0.1865,
+      "step": 1712
+    },
+    {
+      "epoch": 0.25616868550919697,
+      "grad_norm": 1.712020727078079,
+      "learning_rate": 8.717991997938872e-06,
+      "loss": 0.4059,
+      "step": 1713
+    },
+    {
+      "epoch": 0.256318229400329,
+      "grad_norm": 1.3838237147007102,
+      "learning_rate": 8.716372264796355e-06,
+      "loss": 0.3562,
+      "step": 1714
+    },
+    {
+      "epoch": 0.25646777329146103,
+      "grad_norm": 2.009760245861866,
+      "learning_rate": 8.71475165975709e-06,
+      "loss": 0.6606,
+      "step": 1715
+    },
+    {
+      "epoch": 0.2566173171825931,
+      "grad_norm": 1.4629059301831144,
+      "learning_rate": 8.713130183201283e-06,
+      "loss": 0.204,
+      "step": 1716
+    },
+    {
+      "epoch": 0.25676686107372515,
+      "grad_norm": 1.712741449487023,
+      "learning_rate": 8.711507835509352e-06,
+      "loss": 0.3339,
+      "step": 1717
+    },
+    {
+      "epoch": 0.2569164049648572,
+      "grad_norm": 1.7719128226224596,
+      "learning_rate": 8.709884617061912e-06,
+      "loss": 0.4942,
+      "step": 1718
+    },
+    {
+      "epoch": 0.2570659488559892,
+      "grad_norm": 2.5501496729214645,
+      "learning_rate": 8.708260528239788e-06,
+      "loss": 0.1606,
+      "step": 1719
+    },
+    {
+      "epoch": 0.2572154927471213,
+      "grad_norm": 1.3379881043667554,
+      "learning_rate": 8.706635569424007e-06,
+      "loss": 0.1771,
+      "step": 1720
+    },
+    {
+      "epoch": 0.25736503663825333,
+      "grad_norm": 1.3113682840451848,
+      "learning_rate": 8.705009740995798e-06,
+      "loss": 0.1974,
+      "step": 1721
+    },
+    {
+      "epoch": 0.25751458052938536,
+      "grad_norm": 1.3252114693451582,
+      "learning_rate": 8.703383043336598e-06,
+      "loss": 0.1857,
+      "step": 1722
+    },
+    {
+      "epoch": 0.25766412442051745,
+      "grad_norm": 1.2740861405038273,
+      "learning_rate": 8.701755476828045e-06,
+      "loss": 0.179,
+      "step": 1723
+    },
+    {
+      "epoch": 0.2578136683116495,
+      "grad_norm": 1.106234026109271,
+      "learning_rate": 8.700127041851983e-06,
+      "loss": 0.2017,
+      "step": 1724
+    },
+    {
+      "epoch": 0.2579632122027815,
+      "grad_norm": 1.6172632969033804,
+      "learning_rate": 8.698497738790456e-06,
+      "loss": 0.3627,
+      "step": 1725
+    },
+    {
+      "epoch": 0.25811275609391354,
+      "grad_norm": 1.1499075369100074,
+      "learning_rate": 8.696867568025715e-06,
+      "loss": 0.2216,
+      "step": 1726
+    },
+    {
+      "epoch": 0.25826229998504563,
+      "grad_norm": 1.360793130358255,
+      "learning_rate": 8.695236529940217e-06,
+      "loss": 0.2168,
+      "step": 1727
+    },
+    {
+      "epoch": 0.25841184387617766,
+      "grad_norm": 1.5474318214998557,
+      "learning_rate": 8.693604624916614e-06,
+      "loss": 0.3872,
+      "step": 1728
+    },
+    {
+      "epoch": 0.2585613877673097,
+      "grad_norm": 1.603910943390027,
+      "learning_rate": 8.691971853337772e-06,
+      "loss": 0.3828,
+      "step": 1729
+    },
+    {
+      "epoch": 0.2587109316584418,
+      "grad_norm": 1.6171942954911174,
+      "learning_rate": 8.69033821558675e-06,
+      "loss": 0.3922,
+      "step": 1730
+    },
+    {
+      "epoch": 0.2588604755495738,
+      "grad_norm": 1.2280233394489264,
+      "learning_rate": 8.68870371204682e-06,
+      "loss": 0.234,
+      "step": 1731
+    },
+    {
+      "epoch": 0.25901001944070584,
+      "grad_norm": 1.495515188592798,
+      "learning_rate": 8.687068343101449e-06,
+      "loss": 0.3597,
+      "step": 1732
+    },
+    {
+      "epoch": 0.25915956333183787,
+      "grad_norm": 1.2004438690731185,
+      "learning_rate": 8.685432109134309e-06,
+      "loss": 0.3891,
+      "step": 1733
+    },
+    {
+      "epoch": 0.25930910722296996,
+      "grad_norm": 1.1612103655288195,
+      "learning_rate": 8.68379501052928e-06,
+      "loss": 0.1674,
+      "step": 1734
+    },
+    {
+      "epoch": 0.259458651114102,
+      "grad_norm": 1.5379960884614503,
+      "learning_rate": 8.682157047670439e-06,
+      "loss": 0.2165,
+      "step": 1735
+    },
+    {
+      "epoch": 0.259608195005234,
+      "grad_norm": 1.256481316250901,
+      "learning_rate": 8.68051822094207e-06,
+      "loss": 0.1956,
+      "step": 1736
+    },
+    {
+      "epoch": 0.2597577388963661,
+      "grad_norm": 1.150195060615946,
+      "learning_rate": 8.678878530728653e-06,
+      "loss": 0.2046,
+      "step": 1737
+    },
+    {
+      "epoch": 0.25990728278749814,
+      "grad_norm": 0.9645721264599928,
+      "learning_rate": 8.677237977414879e-06,
+      "loss": 0.2169,
+      "step": 1738
+    },
+    {
+      "epoch": 0.26005682667863017,
+      "grad_norm": 2.055210192087077,
+      "learning_rate": 8.675596561385637e-06,
+      "loss": 0.5667,
+      "step": 1739
+    },
+    {
+      "epoch": 0.2602063705697622,
+      "grad_norm": 1.6323123777977344,
+      "learning_rate": 8.67395428302602e-06,
+      "loss": 0.2085,
+      "step": 1740
+    },
+    {
+      "epoch": 0.2603559144608943,
+      "grad_norm": 2.0705229310389734,
+      "learning_rate": 8.67231114272132e-06,
+      "loss": 0.6948,
+      "step": 1741
+    },
+    {
+      "epoch": 0.2605054583520263,
+      "grad_norm": 1.4253846439937559,
+      "learning_rate": 8.670667140857034e-06,
+      "loss": 0.2156,
+      "step": 1742
+    },
+    {
+      "epoch": 0.26065500224315835,
+      "grad_norm": 1.2432278896167281,
+      "learning_rate": 8.669022277818861e-06,
+      "loss": 0.1668,
+      "step": 1743
+    },
+    {
+      "epoch": 0.26080454613429044,
+      "grad_norm": 2.1626989048511396,
+      "learning_rate": 8.6673765539927e-06,
+      "loss": 0.5859,
+      "step": 1744
+    },
+    {
+      "epoch": 0.26095409002542247,
+      "grad_norm": 1.3212936003748748,
+      "learning_rate": 8.66572996976466e-06,
+      "loss": 0.2208,
+      "step": 1745
+    },
+    {
+      "epoch": 0.2611036339165545,
+      "grad_norm": 1.6949012011313151,
+      "learning_rate": 8.66408252552104e-06,
+      "loss": 0.3465,
+      "step": 1746
+    },
+    {
+      "epoch": 0.26125317780768653,
+      "grad_norm": 1.6334321794556426,
+      "learning_rate": 8.662434221648344e-06,
+      "loss": 0.3536,
+      "step": 1747
+    },
+    {
+      "epoch": 0.2614027216988186,
+      "grad_norm": 1.4554313973278277,
+      "learning_rate": 8.660785058533288e-06,
+      "loss": 0.1853,
+      "step": 1748
+    },
+    {
+      "epoch": 0.26155226558995065,
+      "grad_norm": 1.4889399004698576,
+      "learning_rate": 8.659135036562774e-06,
+      "loss": 0.3361,
+      "step": 1749
+    },
+    {
+      "epoch": 0.2617018094810827,
+      "grad_norm": 1.2820685580565203,
+      "learning_rate": 8.657484156123916e-06,
+      "loss": 0.2098,
+      "step": 1750
+    },
+    {
+      "epoch": 0.26185135337221477,
+      "grad_norm": 1.4528196588291855,
+      "learning_rate": 8.655832417604028e-06,
+      "loss": 0.3716,
+      "step": 1751
+    },
+    {
+      "epoch": 0.2620008972633468,
+      "grad_norm": 1.1976317675054606,
+      "learning_rate": 8.65417982139062e-06,
+      "loss": 0.2427,
+      "step": 1752
+    },
+    {
+      "epoch": 0.26215044115447883,
+      "grad_norm": 1.5994876354129555,
+      "learning_rate": 8.652526367871412e-06,
+      "loss": 0.4616,
+      "step": 1753
+    },
+    {
+      "epoch": 0.26229998504561086,
+      "grad_norm": 1.4225347188398327,
+      "learning_rate": 8.650872057434316e-06,
+      "loss": 0.2385,
+      "step": 1754
+    },
+    {
+      "epoch": 0.26244952893674295,
+      "grad_norm": 1.6636967962332678,
+      "learning_rate": 8.64921689046745e-06,
+      "loss": 0.2447,
+      "step": 1755
+    },
+    {
+      "epoch": 0.262599072827875,
+      "grad_norm": 1.5225788924725037,
+      "learning_rate": 8.647560867359133e-06,
+      "loss": 0.3694,
+      "step": 1756
+    },
+    {
+      "epoch": 0.262748616719007,
+      "grad_norm": 1.2202525703548277,
+      "learning_rate": 8.645903988497884e-06,
+      "loss": 0.187,
+      "step": 1757
+    },
+    {
+      "epoch": 0.2628981606101391,
+      "grad_norm": 1.9853373281678817,
+      "learning_rate": 8.644246254272423e-06,
+      "loss": 0.2257,
+      "step": 1758
+    },
+    {
+      "epoch": 0.26304770450127113,
+      "grad_norm": 1.1859244322025415,
+      "learning_rate": 8.64258766507167e-06,
+      "loss": 0.2107,
+      "step": 1759
+    },
+    {
+      "epoch": 0.26319724839240316,
+      "grad_norm": 1.610055608255526,
+      "learning_rate": 8.640928221284744e-06,
+      "loss": 0.3033,
+      "step": 1760
+    },
+    {
+      "epoch": 0.2633467922835352,
+      "grad_norm": 1.322382799898529,
+      "learning_rate": 8.63926792330097e-06,
+      "loss": 0.219,
+      "step": 1761
+    },
+    {
+      "epoch": 0.2634963361746673,
+      "grad_norm": 1.5892475425535593,
+      "learning_rate": 8.637606771509868e-06,
+      "loss": 0.3678,
+      "step": 1762
+    },
+    {
+      "epoch": 0.2636458800657993,
+      "grad_norm": 1.6038277957567881,
+      "learning_rate": 8.635944766301158e-06,
+      "loss": 0.3819,
+      "step": 1763
+    },
+    {
+      "epoch": 0.26379542395693134,
+      "grad_norm": 1.7817364162394536,
+      "learning_rate": 8.634281908064767e-06,
+      "loss": 0.3632,
+      "step": 1764
+    },
+    {
+      "epoch": 0.2639449678480634,
+      "grad_norm": 1.8816182001166262,
+      "learning_rate": 8.632618197190817e-06,
+      "loss": 0.4874,
+      "step": 1765
+    },
+    {
+      "epoch": 0.26409451173919546,
+      "grad_norm": 1.7637033301301739,
+      "learning_rate": 8.630953634069627e-06,
+      "loss": 0.2109,
+      "step": 1766
+    },
+    {
+      "epoch": 0.2642440556303275,
+      "grad_norm": 1.4820279840766182,
+      "learning_rate": 8.629288219091722e-06,
+      "loss": 0.2482,
+      "step": 1767
+    },
+    {
+      "epoch": 0.2643935995214595,
+      "grad_norm": 1.3257383394968065,
+      "learning_rate": 8.627621952647825e-06,
+      "loss": 0.2066,
+      "step": 1768
+    },
+    {
+      "epoch": 0.2645431434125916,
+      "grad_norm": 1.6804248433426907,
+      "learning_rate": 8.625954835128856e-06,
+      "loss": 0.2436,
+      "step": 1769
+    },
+    {
+      "epoch": 0.26469268730372364,
+      "grad_norm": 1.8362570342532276,
+      "learning_rate": 8.624286866925938e-06,
+      "loss": 0.3694,
+      "step": 1770
+    },
+    {
+      "epoch": 0.26484223119485567,
+      "grad_norm": 1.5661865351400834,
+      "learning_rate": 8.622618048430391e-06,
+      "loss": 0.3591,
+      "step": 1771
+    },
+    {
+      "epoch": 0.26499177508598776,
+      "grad_norm": 2.1673241284131044,
+      "learning_rate": 8.62094838003374e-06,
+      "loss": 0.5737,
+      "step": 1772
+    },
+    {
+      "epoch": 0.2651413189771198,
+      "grad_norm": 1.0699484499241998,
+      "learning_rate": 8.619277862127702e-06,
+      "loss": 0.2189,
+      "step": 1773
+    },
+    {
+      "epoch": 0.2652908628682518,
+      "grad_norm": 1.736379723484864,
+      "learning_rate": 8.617606495104198e-06,
+      "loss": 0.4754,
+      "step": 1774
+    },
+    {
+      "epoch": 0.2654404067593839,
+      "grad_norm": 1.583421637766384,
+      "learning_rate": 8.615934279355345e-06,
+      "loss": 0.3366,
+      "step": 1775
+    },
+    {
+      "epoch": 0.26558995065051594,
+      "grad_norm": 1.6478010096837548,
+      "learning_rate": 8.614261215273462e-06,
+      "loss": 0.3672,
+      "step": 1776
+    },
+    {
+      "epoch": 0.26573949454164797,
+      "grad_norm": 1.9102927551172686,
+      "learning_rate": 8.612587303251065e-06,
+      "loss": 0.4308,
+      "step": 1777
+    },
+    {
+      "epoch": 0.26588903843278,
+      "grad_norm": 1.013647409926638,
+      "learning_rate": 8.610912543680872e-06,
+      "loss": 0.2259,
+      "step": 1778
+    },
+    {
+      "epoch": 0.2660385823239121,
+      "grad_norm": 0.9074429437734283,
+      "learning_rate": 8.609236936955797e-06,
+      "loss": 0.1942,
+      "step": 1779
+    },
+    {
+      "epoch": 0.2661881262150441,
+      "grad_norm": 1.1128832323090567,
+      "learning_rate": 8.607560483468953e-06,
+      "loss": 0.2495,
+      "step": 1780
+    },
+    {
+      "epoch": 0.26633767010617615,
+      "grad_norm": 1.4209004510295011,
+      "learning_rate": 8.60588318361365e-06,
+      "loss": 0.2153,
+      "step": 1781
+    },
+    {
+      "epoch": 0.26648721399730824,
+      "grad_norm": 1.5238417321433448,
+      "learning_rate": 8.604205037783404e-06,
+      "loss": 0.1742,
+      "step": 1782
+    },
+    {
+      "epoch": 0.26663675788844027,
+      "grad_norm": 1.2906114880614232,
+      "learning_rate": 8.60252604637192e-06,
+      "loss": 0.3304,
+      "step": 1783
+    },
+    {
+      "epoch": 0.2667863017795723,
+      "grad_norm": 2.0207728584980726,
+      "learning_rate": 8.600846209773107e-06,
+      "loss": 0.7203,
+      "step": 1784
+    },
+    {
+      "epoch": 0.26693584567070433,
+      "grad_norm": 1.0286990588537865,
+      "learning_rate": 8.599165528381068e-06,
+      "loss": 0.3377,
+      "step": 1785
+    },
+    {
+      "epoch": 0.2670853895618364,
+      "grad_norm": 1.6105436646556477,
+      "learning_rate": 8.597484002590113e-06,
+      "loss": 0.3509,
+      "step": 1786
+    },
+    {
+      "epoch": 0.26723493345296845,
+      "grad_norm": 1.655849264356554,
+      "learning_rate": 8.595801632794738e-06,
+      "loss": 0.3233,
+      "step": 1787
+    },
+    {
+      "epoch": 0.2673844773441005,
+      "grad_norm": 1.411449436450616,
+      "learning_rate": 8.594118419389648e-06,
+      "loss": 0.2504,
+      "step": 1788
+    },
+    {
+      "epoch": 0.26753402123523257,
+      "grad_norm": 1.3172823398923945,
+      "learning_rate": 8.592434362769738e-06,
+      "loss": 0.3616,
+      "step": 1789
+    },
+    {
+      "epoch": 0.2676835651263646,
+      "grad_norm": 1.4726126876884136,
+      "learning_rate": 8.590749463330105e-06,
+      "loss": 0.1986,
+      "step": 1790
+    },
+    {
+      "epoch": 0.26783310901749663,
+      "grad_norm": 1.7217144976129872,
+      "learning_rate": 8.589063721466041e-06,
+      "loss": 0.3317,
+      "step": 1791
+    },
+    {
+      "epoch": 0.26798265290862866,
+      "grad_norm": 1.5253687117024657,
+      "learning_rate": 8.58737713757304e-06,
+      "loss": 0.1819,
+      "step": 1792
+    },
+    {
+      "epoch": 0.26813219679976075,
+      "grad_norm": 1.0839801985857815,
+      "learning_rate": 8.585689712046792e-06,
+      "loss": 0.1985,
+      "step": 1793
+    },
+    {
+      "epoch": 0.2682817406908928,
+      "grad_norm": 1.4726576160050187,
+      "learning_rate": 8.58400144528318e-06,
+      "loss": 0.3811,
+      "step": 1794
+    },
+    {
+      "epoch": 0.2684312845820248,
+      "grad_norm": 1.628036147711573,
+      "learning_rate": 8.582312337678286e-06,
+      "loss": 0.4639,
+      "step": 1795
+    },
+    {
+      "epoch": 0.2685808284731569,
+      "grad_norm": 1.8952018740494194,
+      "learning_rate": 8.580622389628395e-06,
+      "loss": 0.3552,
+      "step": 1796
+    },
+    {
+      "epoch": 0.2687303723642889,
+      "grad_norm": 2.001984029391798,
+      "learning_rate": 8.578931601529983e-06,
+      "loss": 0.3726,
+      "step": 1797
+    },
+    {
+      "epoch": 0.26887991625542096,
+      "grad_norm": 1.2603899746687859,
+      "learning_rate": 8.577239973779727e-06,
+      "loss": 0.3207,
+      "step": 1798
+    },
+    {
+      "epoch": 0.269029460146553,
+      "grad_norm": 1.4854613896182869,
+      "learning_rate": 8.575547506774498e-06,
+      "loss": 0.2007,
+      "step": 1799
+    },
+    {
+      "epoch": 0.2691790040376851,
+      "grad_norm": 1.8906751578747407,
+      "learning_rate": 8.573854200911365e-06,
+      "loss": 0.6032,
+      "step": 1800
+    },
+    {
+      "epoch": 0.2693285479288171,
+      "grad_norm": 1.9613152098406168,
+      "learning_rate": 8.572160056587592e-06,
+      "loss": 0.3775,
+      "step": 1801
+    },
+    {
+      "epoch": 0.26947809181994914,
+      "grad_norm": 0.9889564582361468,
+      "learning_rate": 8.570465074200645e-06,
+      "loss": 0.1862,
+      "step": 1802
+    },
+    {
+      "epoch": 0.2696276357110812,
+      "grad_norm": 2.0931504871597015,
+      "learning_rate": 8.568769254148182e-06,
+      "loss": 0.5499,
+      "step": 1803
+    },
+    {
+      "epoch": 0.26977717960221326,
+      "grad_norm": 1.5128120195103958,
+      "learning_rate": 8.56707259682806e-06,
+      "loss": 0.3961,
+      "step": 1804
+    },
+    {
+      "epoch": 0.2699267234933453,
+      "grad_norm": 1.4113868405534056,
+      "learning_rate": 8.565375102638327e-06,
+      "loss": 0.332,
+      "step": 1805
+    },
+    {
+      "epoch": 0.2700762673844773,
+      "grad_norm": 1.9594938693881467,
+      "learning_rate": 8.563676771977234e-06,
+      "loss": 0.5145,
+      "step": 1806
+    },
+    {
+      "epoch": 0.2702258112756094,
+      "grad_norm": 1.7464884113562893,
+      "learning_rate": 8.561977605243228e-06,
+      "loss": 0.3668,
+      "step": 1807
+    },
+    {
+      "epoch": 0.27037535516674144,
+      "grad_norm": 1.1626732860512075,
+      "learning_rate": 8.560277602834945e-06,
+      "loss": 0.168,
+      "step": 1808
+    },
+    {
+      "epoch": 0.27052489905787347,
+      "grad_norm": 1.6748170212716917,
+      "learning_rate": 8.558576765151227e-06,
+      "loss": 0.2433,
+      "step": 1809
+    },
+    {
+      "epoch": 0.27067444294900556,
+      "grad_norm": 2.025570121175929,
+      "learning_rate": 8.556875092591104e-06,
+      "loss": 0.2798,
+      "step": 1810
+    },
+    {
+      "epoch": 0.2708239868401376,
+      "grad_norm": 2.374954242014533,
+      "learning_rate": 8.555172585553804e-06,
+      "loss": 0.4171,
+      "step": 1811
+    },
+    {
+      "epoch": 0.2709735307312696,
+      "grad_norm": 1.7745787943498124,
+      "learning_rate": 8.553469244438754e-06,
+      "loss": 0.3595,
+      "step": 1812
+    },
+    {
+      "epoch": 0.27112307462240165,
+      "grad_norm": 2.431602477728119,
+      "learning_rate": 8.551765069645574e-06,
+      "loss": 0.472,
+      "step": 1813
+    },
+    {
+      "epoch": 0.27127261851353374,
+      "grad_norm": 1.5428274790156662,
+      "learning_rate": 8.55006006157408e-06,
+      "loss": 0.3913,
+      "step": 1814
+    },
+    {
+      "epoch": 0.27142216240466577,
+      "grad_norm": 1.4868080151451115,
+      "learning_rate": 8.548354220624278e-06,
+      "loss": 0.2358,
+      "step": 1815
+    },
+    {
+      "epoch": 0.2715717062957978,
+      "grad_norm": 1.5637252724169068,
+      "learning_rate": 8.546647547196383e-06,
+      "loss": 0.1934,
+      "step": 1816
+    },
+    {
+      "epoch": 0.2717212501869299,
+      "grad_norm": 1.4784784501314225,
+      "learning_rate": 8.544940041690792e-06,
+      "loss": 0.5315,
+      "step": 1817
+    },
+    {
+      "epoch": 0.2718707940780619,
+      "grad_norm": 1.6049147134883024,
+      "learning_rate": 8.543231704508102e-06,
+      "loss": 0.2277,
+      "step": 1818
+    },
+    {
+      "epoch": 0.27202033796919395,
+      "grad_norm": 1.1639185886904326,
+      "learning_rate": 8.541522536049108e-06,
+      "loss": 0.2021,
+      "step": 1819
+    },
+    {
+      "epoch": 0.272169881860326,
+      "grad_norm": 1.5080421265491435,
+      "learning_rate": 8.539812536714796e-06,
+      "loss": 0.1978,
+      "step": 1820
+    },
+    {
+      "epoch": 0.27231942575145807,
+      "grad_norm": 0.880232094927403,
+      "learning_rate": 8.538101706906347e-06,
+      "loss": 0.1639,
+      "step": 1821
+    },
+    {
+      "epoch": 0.2724689696425901,
+      "grad_norm": 1.4359482262535581,
+      "learning_rate": 8.536390047025143e-06,
+      "loss": 0.2078,
+      "step": 1822
+    },
+    {
+      "epoch": 0.27261851353372213,
+      "grad_norm": 1.3134455382926835,
+      "learning_rate": 8.53467755747275e-06,
+      "loss": 0.3189,
+      "step": 1823
+    },
+    {
+      "epoch": 0.2727680574248542,
+      "grad_norm": 1.4404618987021778,
+      "learning_rate": 8.532964238650938e-06,
+      "loss": 0.3408,
+      "step": 1824
+    },
+    {
+      "epoch": 0.27291760131598625,
+      "grad_norm": 1.3234925937557052,
+      "learning_rate": 8.531250090961666e-06,
+      "loss": 0.2465,
+      "step": 1825
+    },
+    {
+      "epoch": 0.2730671452071183,
+      "grad_norm": 1.678127185102635,
+      "learning_rate": 8.52953511480709e-06,
+      "loss": 0.5327,
+      "step": 1826
+    },
+    {
+      "epoch": 0.2732166890982503,
+      "grad_norm": 1.4928358926152048,
+      "learning_rate": 8.527819310589564e-06,
+      "loss": 0.3706,
+      "step": 1827
+    },
+    {
+      "epoch": 0.2733662329893824,
+      "grad_norm": 1.7549470816483508,
+      "learning_rate": 8.526102678711626e-06,
+      "loss": 0.4288,
+      "step": 1828
+    },
+    {
+      "epoch": 0.2735157768805144,
+      "grad_norm": 1.5327378751610867,
+      "learning_rate": 8.524385219576019e-06,
+      "loss": 0.4647,
+      "step": 1829
+    },
+    {
+      "epoch": 0.27366532077164646,
+      "grad_norm": 1.7197750676087826,
+      "learning_rate": 8.522666933585672e-06,
+      "loss": 0.3523,
+      "step": 1830
+    },
+    {
+      "epoch": 0.27381486466277855,
+      "grad_norm": 1.174817852571183,
+      "learning_rate": 8.520947821143714e-06,
+      "loss": 0.2134,
+      "step": 1831
+    },
+    {
+      "epoch": 0.2739644085539106,
+      "grad_norm": 2.0070860843733302,
+      "learning_rate": 8.519227882653465e-06,
+      "loss": 0.4656,
+      "step": 1832
+    },
+    {
+      "epoch": 0.2741139524450426,
+      "grad_norm": 1.7275106234713773,
+      "learning_rate": 8.517507118518436e-06,
+      "loss": 0.5513,
+      "step": 1833
+    },
+    {
+      "epoch": 0.2742634963361747,
+      "grad_norm": 1.2685207625970671,
+      "learning_rate": 8.515785529142339e-06,
+      "loss": 0.2491,
+      "step": 1834
+    },
+    {
+      "epoch": 0.2744130402273067,
+      "grad_norm": 1.2419112651105808,
+      "learning_rate": 8.514063114929077e-06,
+      "loss": 0.1897,
+      "step": 1835
+    },
+    {
+      "epoch": 0.27456258411843876,
+      "grad_norm": 1.1605015571551742,
+      "learning_rate": 8.512339876282738e-06,
+      "loss": 0.2282,
+      "step": 1836
+    },
+    {
+      "epoch": 0.2747121280095708,
+      "grad_norm": 1.2889897008048643,
+      "learning_rate": 8.510615813607617e-06,
+      "loss": 0.3405,
+      "step": 1837
+    },
+    {
+      "epoch": 0.2748616719007029,
+      "grad_norm": 1.8490938154917704,
+      "learning_rate": 8.508890927308191e-06,
+      "loss": 0.3821,
+      "step": 1838
+    },
+    {
+      "epoch": 0.2750112157918349,
+      "grad_norm": 1.3689509544740683,
+      "learning_rate": 8.507165217789137e-06,
+      "loss": 0.3832,
+      "step": 1839
+    },
+    {
+      "epoch": 0.27516075968296694,
+      "grad_norm": 1.6403361270537955,
+      "learning_rate": 8.505438685455325e-06,
+      "loss": 0.3848,
+      "step": 1840
+    },
+    {
+      "epoch": 0.275310303574099,
+      "grad_norm": 1.8182842779990729,
+      "learning_rate": 8.503711330711813e-06,
+      "loss": 0.2549,
+      "step": 1841
+    },
+    {
+      "epoch": 0.27545984746523106,
+      "grad_norm": 1.7148274136332127,
+      "learning_rate": 8.501983153963855e-06,
+      "loss": 0.5105,
+      "step": 1842
+    },
+    {
+      "epoch": 0.2756093913563631,
+      "grad_norm": 1.537038227068178,
+      "learning_rate": 8.500254155616901e-06,
+      "loss": 0.3514,
+      "step": 1843
+    },
+    {
+      "epoch": 0.2757589352474951,
+      "grad_norm": 1.641983169416554,
+      "learning_rate": 8.498524336076587e-06,
+      "loss": 0.6491,
+      "step": 1844
+    },
+    {
+      "epoch": 0.2759084791386272,
+      "grad_norm": 1.250729132131747,
+      "learning_rate": 8.49679369574875e-06,
+      "loss": 0.2051,
+      "step": 1845
+    },
+    {
+      "epoch": 0.27605802302975924,
+      "grad_norm": 1.579592841261978,
+      "learning_rate": 8.49506223503941e-06,
+      "loss": 0.3549,
+      "step": 1846
+    },
+    {
+      "epoch": 0.27620756692089127,
+      "grad_norm": 1.1166297247129173,
+      "learning_rate": 8.493329954354788e-06,
+      "loss": 0.2016,
+      "step": 1847
+    },
+    {
+      "epoch": 0.27635711081202335,
+      "grad_norm": 1.457594575978762,
+      "learning_rate": 8.491596854101292e-06,
+      "loss": 0.3752,
+      "step": 1848
+    },
+    {
+      "epoch": 0.2765066547031554,
+      "grad_norm": 1.6194515688957392,
+      "learning_rate": 8.489862934685523e-06,
+      "loss": 0.2791,
+      "step": 1849
+    },
+    {
+      "epoch": 0.2766561985942874,
+      "grad_norm": 1.189645094014293,
+      "learning_rate": 8.488128196514279e-06,
+      "loss": 0.1892,
+      "step": 1850
+    },
+    {
+      "epoch": 0.27680574248541945,
+      "grad_norm": 1.6414546776027312,
+      "learning_rate": 8.486392639994541e-06,
+      "loss": 0.4263,
+      "step": 1851
+    },
+    {
+      "epoch": 0.27695528637655153,
+      "grad_norm": 1.1030338687374446,
+      "learning_rate": 8.48465626553349e-06,
+      "loss": 0.2174,
+      "step": 1852
+    },
+    {
+      "epoch": 0.27710483026768357,
+      "grad_norm": 2.062987335024284,
+      "learning_rate": 8.482919073538498e-06,
+      "loss": 0.5137,
+      "step": 1853
+    },
+    {
+      "epoch": 0.2772543741588156,
+      "grad_norm": 1.9265333026489615,
+      "learning_rate": 8.481181064417124e-06,
+      "loss": 0.5053,
+      "step": 1854
+    },
+    {
+      "epoch": 0.2774039180499477,
+      "grad_norm": 1.2857660275288423,
+      "learning_rate": 8.479442238577123e-06,
+      "loss": 0.2356,
+      "step": 1855
+    },
+    {
+      "epoch": 0.2775534619410797,
+      "grad_norm": 1.4308427978938358,
+      "learning_rate": 8.477702596426441e-06,
+      "loss": 0.3124,
+      "step": 1856
+    },
+    {
+      "epoch": 0.27770300583221175,
+      "grad_norm": 2.112210621284666,
+      "learning_rate": 8.475962138373212e-06,
+      "loss": 0.5583,
+      "step": 1857
+    },
+    {
+      "epoch": 0.2778525497233438,
+      "grad_norm": 1.5648504781268902,
+      "learning_rate": 8.474220864825768e-06,
+      "loss": 0.2046,
+      "step": 1858
+    },
+    {
+      "epoch": 0.27800209361447586,
+      "grad_norm": 1.0861658989898926,
+      "learning_rate": 8.472478776192624e-06,
+      "loss": 0.1941,
+      "step": 1859
+    },
+    {
+      "epoch": 0.2781516375056079,
+      "grad_norm": 1.7593139339710724,
+      "learning_rate": 8.470735872882498e-06,
+      "loss": 0.3911,
+      "step": 1860
+    },
+    {
+      "epoch": 0.2783011813967399,
+      "grad_norm": 1.9051060527282064,
+      "learning_rate": 8.468992155304285e-06,
+      "loss": 0.3441,
+      "step": 1861
+    },
+    {
+      "epoch": 0.278450725287872,
+      "grad_norm": 1.2726234923585076,
+      "learning_rate": 8.46724762386708e-06,
+      "loss": 0.1966,
+      "step": 1862
+    },
+    {
+      "epoch": 0.27860026917900405,
+      "grad_norm": 1.6275578962742665,
+      "learning_rate": 8.465502278980168e-06,
+      "loss": 0.2015,
+      "step": 1863
+    },
+    {
+      "epoch": 0.2787498130701361,
+      "grad_norm": 1.1894457179498164,
+      "learning_rate": 8.463756121053024e-06,
+      "loss": 0.2064,
+      "step": 1864
+    },
+    {
+      "epoch": 0.2788993569612681,
+      "grad_norm": 0.8970585928811048,
+      "learning_rate": 8.462009150495311e-06,
+      "loss": 0.2077,
+      "step": 1865
+    },
+    {
+      "epoch": 0.2790489008524002,
+      "grad_norm": 1.4979876938013277,
+      "learning_rate": 8.460261367716888e-06,
+      "loss": 0.1868,
+      "step": 1866
+    },
+    {
+      "epoch": 0.2791984447435322,
+      "grad_norm": 1.5952489152766243,
+      "learning_rate": 8.458512773127801e-06,
+      "loss": 0.3924,
+      "step": 1867
+    },
+    {
+      "epoch": 0.27934798863466426,
+      "grad_norm": 1.3568222258261973,
+      "learning_rate": 8.456763367138283e-06,
+      "loss": 0.3057,
+      "step": 1868
+    },
+    {
+      "epoch": 0.27949753252579634,
+      "grad_norm": 1.430069295266679,
+      "learning_rate": 8.455013150158767e-06,
+      "loss": 0.2027,
+      "step": 1869
+    },
+    {
+      "epoch": 0.2796470764169284,
+      "grad_norm": 1.835468064313153,
+      "learning_rate": 8.453262122599871e-06,
+      "loss": 0.203,
+      "step": 1870
+    },
+    {
+      "epoch": 0.2797966203080604,
+      "grad_norm": 1.16217124036999,
+      "learning_rate": 8.451510284872397e-06,
+      "loss": 0.2065,
+      "step": 1871
+    },
+    {
+      "epoch": 0.27994616419919244,
+      "grad_norm": 1.2439706377831612,
+      "learning_rate": 8.449757637387349e-06,
+      "loss": 0.1854,
+      "step": 1872
+    },
+    {
+      "epoch": 0.2800957080903245,
+      "grad_norm": 1.771165875925341,
+      "learning_rate": 8.448004180555912e-06,
+      "loss": 0.2118,
+      "step": 1873
+    },
+    {
+      "epoch": 0.28024525198145656,
+      "grad_norm": 1.448237639694636,
+      "learning_rate": 8.446249914789464e-06,
+      "loss": 0.329,
+      "step": 1874
+    },
+    {
+      "epoch": 0.2803947958725886,
+      "grad_norm": 1.9446859154931548,
+      "learning_rate": 8.444494840499573e-06,
+      "loss": 0.1968,
+      "step": 1875
+    },
+    {
+      "epoch": 0.2805443397637207,
+      "grad_norm": 1.1282582980944778,
+      "learning_rate": 8.442738958097998e-06,
+      "loss": 0.1929,
+      "step": 1876
+    },
+    {
+      "epoch": 0.2806938836548527,
+      "grad_norm": 0.9295466099898506,
+      "learning_rate": 8.440982267996683e-06,
+      "loss": 0.176,
+      "step": 1877
+    },
+    {
+      "epoch": 0.28084342754598474,
+      "grad_norm": 1.6067454611327383,
+      "learning_rate": 8.439224770607768e-06,
+      "loss": 0.2216,
+      "step": 1878
+    },
+    {
+      "epoch": 0.28099297143711677,
+      "grad_norm": 1.506148146117863,
+      "learning_rate": 8.437466466343573e-06,
+      "loss": 0.3125,
+      "step": 1879
+    },
+    {
+      "epoch": 0.28114251532824885,
+      "grad_norm": 1.1300992074338925,
+      "learning_rate": 8.43570735561662e-06,
+      "loss": 0.2173,
+      "step": 1880
+    },
+    {
+      "epoch": 0.2812920592193809,
+      "grad_norm": 1.0011833423935625,
+      "learning_rate": 8.43394743883961e-06,
+      "loss": 0.2908,
+      "step": 1881
+    },
+    {
+      "epoch": 0.2814416031105129,
+      "grad_norm": 1.8315268271499454,
+      "learning_rate": 8.432186716425438e-06,
+      "loss": 0.3737,
+      "step": 1882
+    },
+    {
+      "epoch": 0.281591147001645,
+      "grad_norm": 1.7995305921177922,
+      "learning_rate": 8.430425188787181e-06,
+      "loss": 0.3303,
+      "step": 1883
+    },
+    {
+      "epoch": 0.28174069089277703,
+      "grad_norm": 1.2710141324651125,
+      "learning_rate": 8.428662856338116e-06,
+      "loss": 0.2441,
+      "step": 1884
+    },
+    {
+      "epoch": 0.28189023478390907,
+      "grad_norm": 1.3946588107946323,
+      "learning_rate": 8.426899719491703e-06,
+      "loss": 0.2108,
+      "step": 1885
+    },
+    {
+      "epoch": 0.2820397786750411,
+      "grad_norm": 1.1149474892337983,
+      "learning_rate": 8.42513577866159e-06,
+      "loss": 0.2215,
+      "step": 1886
+    },
+    {
+      "epoch": 0.2821893225661732,
+      "grad_norm": 2.008735822937365,
+      "learning_rate": 8.423371034261612e-06,
+      "loss": 0.4968,
+      "step": 1887
+    },
+    {
+      "epoch": 0.2823388664573052,
+      "grad_norm": 1.935748092708562,
+      "learning_rate": 8.4216054867058e-06,
+      "loss": 0.1889,
+      "step": 1888
+    },
+    {
+      "epoch": 0.28248841034843725,
+      "grad_norm": 1.7520452919556104,
+      "learning_rate": 8.419839136408361e-06,
+      "loss": 0.3447,
+      "step": 1889
+    },
+    {
+      "epoch": 0.28263795423956933,
+      "grad_norm": 0.8268783166552869,
+      "learning_rate": 8.418071983783707e-06,
+      "loss": 0.2028,
+      "step": 1890
+    },
+    {
+      "epoch": 0.28278749813070136,
+      "grad_norm": 1.2060134973326222,
+      "learning_rate": 8.416304029246422e-06,
+      "loss": 0.2033,
+      "step": 1891
+    },
+    {
+      "epoch": 0.2829370420218334,
+      "grad_norm": 1.081395124056579,
+      "learning_rate": 8.414535273211286e-06,
+      "loss": 0.2197,
+      "step": 1892
+    },
+    {
+      "epoch": 0.2830865859129655,
+      "grad_norm": 1.6424700307942506,
+      "learning_rate": 8.412765716093273e-06,
+      "loss": 0.3939,
+      "step": 1893
+    },
+    {
+      "epoch": 0.2832361298040975,
+      "grad_norm": 1.3786288095772328,
+      "learning_rate": 8.410995358307528e-06,
+      "loss": 0.187,
+      "step": 1894
+    },
+    {
+      "epoch": 0.28338567369522955,
+      "grad_norm": 1.1978351240462277,
+      "learning_rate": 8.4092242002694e-06,
+      "loss": 0.2215,
+      "step": 1895
+    },
+    {
+      "epoch": 0.2835352175863616,
+      "grad_norm": 1.2274171257303508,
+      "learning_rate": 8.40745224239442e-06,
+      "loss": 0.2047,
+      "step": 1896
+    },
+    {
+      "epoch": 0.28368476147749366,
+      "grad_norm": 1.5902928460844958,
+      "learning_rate": 8.405679485098304e-06,
+      "loss": 0.3429,
+      "step": 1897
+    },
+    {
+      "epoch": 0.2838343053686257,
+      "grad_norm": 1.8154747327907406,
+      "learning_rate": 8.403905928796961e-06,
+      "loss": 0.4344,
+      "step": 1898
+    },
+    {
+      "epoch": 0.2839838492597577,
+      "grad_norm": 1.626710770634783,
+      "learning_rate": 8.402131573906479e-06,
+      "loss": 0.4395,
+      "step": 1899
+    },
+    {
+      "epoch": 0.2841333931508898,
+      "grad_norm": 1.2949990843572656,
+      "learning_rate": 8.400356420843144e-06,
+      "loss": 0.2326,
+      "step": 1900
+    },
+    {
+      "epoch": 0.28428293704202184,
+      "grad_norm": 1.5453342232271752,
+      "learning_rate": 8.398580470023422e-06,
+      "loss": 0.3238,
+      "step": 1901
+    },
+    {
+      "epoch": 0.2844324809331539,
+      "grad_norm": 1.4286541357110913,
+      "learning_rate": 8.396803721863969e-06,
+      "loss": 0.2427,
+      "step": 1902
+    },
+    {
+      "epoch": 0.2845820248242859,
+      "grad_norm": 1.830298456304548,
+      "learning_rate": 8.395026176781627e-06,
+      "loss": 0.3781,
+      "step": 1903
+    },
+    {
+      "epoch": 0.284731568715418,
+      "grad_norm": 0.95541693717727,
+      "learning_rate": 8.393247835193424e-06,
+      "loss": 0.2069,
+      "step": 1904
+    },
+    {
+      "epoch": 0.28488111260655,
+      "grad_norm": 1.4752609480755603,
+      "learning_rate": 8.391468697516575e-06,
+      "loss": 0.2127,
+      "step": 1905
+    },
+    {
+      "epoch": 0.28503065649768206,
+      "grad_norm": 1.4768738670643429,
+      "learning_rate": 8.389688764168487e-06,
+      "loss": 0.3605,
+      "step": 1906
+    },
+    {
+      "epoch": 0.28518020038881414,
+      "grad_norm": 1.2979854503198742,
+      "learning_rate": 8.387908035566747e-06,
+      "loss": 0.1915,
+      "step": 1907
+    },
+    {
+      "epoch": 0.2853297442799462,
+      "grad_norm": 0.9902824337524958,
+      "learning_rate": 8.38612651212913e-06,
+      "loss": 0.2028,
+      "step": 1908
+    },
+    {
+      "epoch": 0.2854792881710782,
+      "grad_norm": 1.6167134405609127,
+      "learning_rate": 8.384344194273602e-06,
+      "loss": 0.2436,
+      "step": 1909
+    },
+    {
+      "epoch": 0.28562883206221024,
+      "grad_norm": 1.901370964967397,
+      "learning_rate": 8.38256108241831e-06,
+      "loss": 0.5823,
+      "step": 1910
+    },
+    {
+      "epoch": 0.2857783759533423,
+      "grad_norm": 1.8663326764685,
+      "learning_rate": 8.380777176981586e-06,
+      "loss": 0.296,
+      "step": 1911
+    },
+    {
+      "epoch": 0.28592791984447435,
+      "grad_norm": 2.010032346949138,
+      "learning_rate": 8.378992478381957e-06,
+      "loss": 0.3593,
+      "step": 1912
+    },
+    {
+      "epoch": 0.2860774637356064,
+      "grad_norm": 1.4476840280584171,
+      "learning_rate": 8.377206987038128e-06,
+      "loss": 0.2342,
+      "step": 1913
+    },
+    {
+      "epoch": 0.28622700762673847,
+      "grad_norm": 1.5594572863531158,
+      "learning_rate": 8.375420703368993e-06,
+      "loss": 0.3667,
+      "step": 1914
+    },
+    {
+      "epoch": 0.2863765515178705,
+      "grad_norm": 1.6185881957264394,
+      "learning_rate": 8.37363362779363e-06,
+      "loss": 0.3414,
+      "step": 1915
+    },
+    {
+      "epoch": 0.28652609540900253,
+      "grad_norm": 1.6321798860422365,
+      "learning_rate": 8.371845760731305e-06,
+      "loss": 0.4336,
+      "step": 1916
+    },
+    {
+      "epoch": 0.28667563930013457,
+      "grad_norm": 1.391635511553069,
+      "learning_rate": 8.370057102601467e-06,
+      "loss": 0.2641,
+      "step": 1917
+    },
+    {
+      "epoch": 0.28682518319126665,
+      "grad_norm": 1.6939840429561068,
+      "learning_rate": 8.368267653823758e-06,
+      "loss": 0.3579,
+      "step": 1918
+    },
+    {
+      "epoch": 0.2869747270823987,
+      "grad_norm": 1.6529503491198996,
+      "learning_rate": 8.366477414817993e-06,
+      "loss": 0.3893,
+      "step": 1919
+    },
+    {
+      "epoch": 0.2871242709735307,
+      "grad_norm": 1.750622438533098,
+      "learning_rate": 8.364686386004184e-06,
+      "loss": 0.3065,
+      "step": 1920
+    },
+    {
+      "epoch": 0.2872738148646628,
+      "grad_norm": 1.5641156236161282,
+      "learning_rate": 8.362894567802522e-06,
+      "loss": 0.1891,
+      "step": 1921
+    },
+    {
+      "epoch": 0.28742335875579483,
+      "grad_norm": 1.3418086188856035,
+      "learning_rate": 8.361101960633384e-06,
+      "loss": 0.2246,
+      "step": 1922
+    },
+    {
+      "epoch": 0.28757290264692686,
+      "grad_norm": 1.5465668819355392,
+      "learning_rate": 8.359308564917335e-06,
+      "loss": 0.2342,
+      "step": 1923
+    },
+    {
+      "epoch": 0.2877224465380589,
+      "grad_norm": 1.6114344336379887,
+      "learning_rate": 8.357514381075123e-06,
+      "loss": 0.1817,
+      "step": 1924
+    },
+    {
+      "epoch": 0.287871990429191,
+      "grad_norm": 1.3668429785293057,
+      "learning_rate": 8.355719409527676e-06,
+      "loss": 0.3416,
+      "step": 1925
+    },
+    {
+      "epoch": 0.288021534320323,
+      "grad_norm": 1.0266243878235821,
+      "learning_rate": 8.353923650696119e-06,
+      "loss": 0.1685,
+      "step": 1926
+    },
+    {
+      "epoch": 0.28817107821145505,
+      "grad_norm": 1.7588874793595086,
+      "learning_rate": 8.352127105001748e-06,
+      "loss": 0.3868,
+      "step": 1927
+    },
+    {
+      "epoch": 0.28832062210258713,
+      "grad_norm": 1.8522510535939594,
+      "learning_rate": 8.350329772866054e-06,
+      "loss": 0.2917,
+      "step": 1928
+    },
+    {
+      "epoch": 0.28847016599371916,
+      "grad_norm": 1.7432437696852978,
+      "learning_rate": 8.348531654710706e-06,
+      "loss": 0.2314,
+      "step": 1929
+    },
+    {
+      "epoch": 0.2886197098848512,
+      "grad_norm": 1.8048485192800683,
+      "learning_rate": 8.34673275095756e-06,
+      "loss": 0.421,
+      "step": 1930
+    },
+    {
+      "epoch": 0.2887692537759832,
+      "grad_norm": 1.254687110691794,
+      "learning_rate": 8.344933062028659e-06,
+      "loss": 0.3086,
+      "step": 1931
+    },
+    {
+      "epoch": 0.2889187976671153,
+      "grad_norm": 1.5516562911782632,
+      "learning_rate": 8.343132588346223e-06,
+      "loss": 0.243,
+      "step": 1932
+    },
+    {
+      "epoch": 0.28906834155824734,
+      "grad_norm": 1.5234323648219494,
+      "learning_rate": 8.341331330332665e-06,
+      "loss": 0.3708,
+      "step": 1933
+    },
+    {
+      "epoch": 0.2892178854493794,
+      "grad_norm": 1.6869191874454617,
+      "learning_rate": 8.339529288410575e-06,
+      "loss": 0.3664,
+      "step": 1934
+    },
+    {
+      "epoch": 0.28936742934051146,
+      "grad_norm": 1.6174375460534882,
+      "learning_rate": 8.337726463002728e-06,
+      "loss": 0.3548,
+      "step": 1935
+    },
+    {
+      "epoch": 0.2895169732316435,
+      "grad_norm": 1.3712125139885123,
+      "learning_rate": 8.335922854532087e-06,
+      "loss": 0.258,
+      "step": 1936
+    },
+    {
+      "epoch": 0.2896665171227755,
+      "grad_norm": 1.782853458880192,
+      "learning_rate": 8.334118463421795e-06,
+      "loss": 0.549,
+      "step": 1937
+    },
+    {
+      "epoch": 0.28981606101390756,
+      "grad_norm": 1.620144816801684,
+      "learning_rate": 8.33231329009518e-06,
+      "loss": 0.3349,
+      "step": 1938
+    },
+    {
+      "epoch": 0.28996560490503964,
+      "grad_norm": 1.333232350703901,
+      "learning_rate": 8.33050733497575e-06,
+      "loss": 0.2262,
+      "step": 1939
+    },
+    {
+      "epoch": 0.2901151487961717,
+      "grad_norm": 1.6837289583297483,
+      "learning_rate": 8.328700598487203e-06,
+      "loss": 0.4787,
+      "step": 1940
+    },
+    {
+      "epoch": 0.2902646926873037,
+      "grad_norm": 1.0591641715642133,
+      "learning_rate": 8.326893081053417e-06,
+      "loss": 0.2385,
+      "step": 1941
+    },
+    {
+      "epoch": 0.2904142365784358,
+      "grad_norm": 1.129840571438994,
+      "learning_rate": 8.325084783098452e-06,
+      "loss": 0.247,
+      "step": 1942
+    },
+    {
+      "epoch": 0.2905637804695678,
+      "grad_norm": 1.432313552747029,
+      "learning_rate": 8.32327570504655e-06,
+      "loss": 0.191,
+      "step": 1943
+    },
+    {
+      "epoch": 0.29071332436069985,
+      "grad_norm": 1.5356970591574979,
+      "learning_rate": 8.32146584732214e-06,
+      "loss": 0.187,
+      "step": 1944
+    },
+    {
+      "epoch": 0.2908628682518319,
+      "grad_norm": 1.0595724079445366,
+      "learning_rate": 8.319655210349832e-06,
+      "loss": 0.1802,
+      "step": 1945
+    },
+    {
+      "epoch": 0.29101241214296397,
+      "grad_norm": 1.6123333459960774,
+      "learning_rate": 8.31784379455442e-06,
+      "loss": 0.2293,
+      "step": 1946
+    },
+    {
+      "epoch": 0.291161956034096,
+      "grad_norm": 1.3341178548001948,
+      "learning_rate": 8.31603160036088e-06,
+      "loss": 0.2359,
+      "step": 1947
+    },
+    {
+      "epoch": 0.29131149992522803,
+      "grad_norm": 1.2376522352882886,
+      "learning_rate": 8.314218628194365e-06,
+      "loss": 0.1896,
+      "step": 1948
+    },
+    {
+      "epoch": 0.2914610438163601,
+      "grad_norm": 1.679702490362655,
+      "learning_rate": 8.312404878480222e-06,
+      "loss": 0.3223,
+      "step": 1949
+    },
+    {
+      "epoch": 0.29161058770749215,
+      "grad_norm": 1.2597407375509204,
+      "learning_rate": 8.31059035164397e-06,
+      "loss": 0.2175,
+      "step": 1950
+    },
+    {
+      "epoch": 0.2917601315986242,
+      "grad_norm": 1.624231610744949,
+      "learning_rate": 8.308775048111318e-06,
+      "loss": 0.3432,
+      "step": 1951
+    },
+    {
+      "epoch": 0.29190967548975627,
+      "grad_norm": 1.9330093174874856,
+      "learning_rate": 8.306958968308152e-06,
+      "loss": 0.2494,
+      "step": 1952
+    },
+    {
+      "epoch": 0.2920592193808883,
+      "grad_norm": 1.7103715792411738,
+      "learning_rate": 8.30514211266054e-06,
+      "loss": 0.3095,
+      "step": 1953
+    },
+    {
+      "epoch": 0.29220876327202033,
+      "grad_norm": 1.27605011963153,
+      "learning_rate": 8.303324481594737e-06,
+      "loss": 0.2158,
+      "step": 1954
+    },
+    {
+      "epoch": 0.29235830716315236,
+      "grad_norm": 1.2814850533429023,
+      "learning_rate": 8.301506075537173e-06,
+      "loss": 0.2314,
+      "step": 1955
+    },
+    {
+      "epoch": 0.29250785105428445,
+      "grad_norm": 1.954217300152197,
+      "learning_rate": 8.299686894914467e-06,
+      "loss": 0.567,
+      "step": 1956
+    },
+    {
+      "epoch": 0.2926573949454165,
+      "grad_norm": 0.9650685264754246,
+      "learning_rate": 8.297866940153416e-06,
+      "loss": 0.2588,
+      "step": 1957
+    },
+    {
+      "epoch": 0.2928069388365485,
+      "grad_norm": 1.5166783973503064,
+      "learning_rate": 8.296046211681e-06,
+      "loss": 0.369,
+      "step": 1958
+    },
+    {
+      "epoch": 0.2929564827276806,
+      "grad_norm": 1.7977899668272896,
+      "learning_rate": 8.294224709924373e-06,
+      "loss": 0.5176,
+      "step": 1959
+    },
+    {
+      "epoch": 0.29310602661881263,
+      "grad_norm": 1.5915279488901846,
+      "learning_rate": 8.292402435310883e-06,
+      "loss": 0.3614,
+      "step": 1960
+    },
+    {
+      "epoch": 0.29325557050994466,
+      "grad_norm": 1.8933979183970016,
+      "learning_rate": 8.290579388268054e-06,
+      "loss": 0.5366,
+      "step": 1961
+    },
+    {
+      "epoch": 0.2934051144010767,
+      "grad_norm": 1.8812762399598735,
+      "learning_rate": 8.288755569223586e-06,
+      "loss": 0.5443,
+      "step": 1962
+    },
+    {
+      "epoch": 0.2935546582922088,
+      "grad_norm": 1.6643902182805455,
+      "learning_rate": 8.286930978605366e-06,
+      "loss": 0.3375,
+      "step": 1963
+    },
+    {
+      "epoch": 0.2937042021833408,
+      "grad_norm": 1.762005285741713,
+      "learning_rate": 8.285105616841463e-06,
+      "loss": 0.5123,
+      "step": 1964
+    },
+    {
+      "epoch": 0.29385374607447284,
+      "grad_norm": 1.6916688314394261,
+      "learning_rate": 8.283279484360119e-06,
+      "loss": 0.4779,
+      "step": 1965
+    },
+    {
+      "epoch": 0.29400328996560493,
+      "grad_norm": 1.0908816823199547,
+      "learning_rate": 8.281452581589768e-06,
+      "loss": 0.3536,
+      "step": 1966
+    },
+    {
+      "epoch": 0.29415283385673696,
+      "grad_norm": 1.4434211599007758,
+      "learning_rate": 8.279624908959018e-06,
+      "loss": 0.1613,
+      "step": 1967
+    },
+    {
+      "epoch": 0.294302377747869,
+      "grad_norm": 1.8196704493501261,
+      "learning_rate": 8.277796466896657e-06,
+      "loss": 0.4989,
+      "step": 1968
+    },
+    {
+      "epoch": 0.294451921639001,
+      "grad_norm": 1.3656918969234806,
+      "learning_rate": 8.275967255831655e-06,
+      "loss": 0.1918,
+      "step": 1969
+    },
+    {
+      "epoch": 0.2946014655301331,
+      "grad_norm": 1.4834473723024195,
+      "learning_rate": 8.274137276193162e-06,
+      "loss": 0.2105,
+      "step": 1970
+    },
+    {
+      "epoch": 0.29475100942126514,
+      "grad_norm": 2.0922406506784745,
+      "learning_rate": 8.272306528410511e-06,
+      "loss": 0.3878,
+      "step": 1971
+    },
+    {
+      "epoch": 0.2949005533123972,
+      "grad_norm": 1.6479894472086396,
+      "learning_rate": 8.270475012913212e-06,
+      "loss": 0.3405,
+      "step": 1972
+    },
+    {
+      "epoch": 0.29505009720352926,
+      "grad_norm": 1.816314957488247,
+      "learning_rate": 8.268642730130956e-06,
+      "loss": 0.3493,
+      "step": 1973
+    },
+    {
+      "epoch": 0.2951996410946613,
+      "grad_norm": 1.7886487407400398,
+      "learning_rate": 8.266809680493615e-06,
+      "loss": 0.4509,
+      "step": 1974
+    },
+    {
+      "epoch": 0.2953491849857933,
+      "grad_norm": 1.4476130090504609,
+      "learning_rate": 8.264975864431239e-06,
+      "loss": 0.3609,
+      "step": 1975
+    },
+    {
+      "epoch": 0.29549872887692535,
+      "grad_norm": 1.6461788405344846,
+      "learning_rate": 8.263141282374058e-06,
+      "loss": 0.3404,
+      "step": 1976
+    },
+    {
+      "epoch": 0.29564827276805744,
+      "grad_norm": 1.5013235196228902,
+      "learning_rate": 8.261305934752486e-06,
+      "loss": 0.2106,
+      "step": 1977
+    },
+    {
+      "epoch": 0.29579781665918947,
+      "grad_norm": 1.4062496662139496,
+      "learning_rate": 8.259469821997111e-06,
+      "loss": 0.3559,
+      "step": 1978
+    },
+    {
+      "epoch": 0.2959473605503215,
+      "grad_norm": 1.1293811953986368,
+      "learning_rate": 8.257632944538704e-06,
+      "loss": 0.2078,
+      "step": 1979
+    },
+    {
+      "epoch": 0.2960969044414536,
+      "grad_norm": 1.4191657864487062,
+      "learning_rate": 8.255795302808212e-06,
+      "loss": 0.3335,
+      "step": 1980
+    },
+    {
+      "epoch": 0.2962464483325856,
+      "grad_norm": 1.2818891440117588,
+      "learning_rate": 8.253956897236765e-06,
+      "loss": 0.1667,
+      "step": 1981
+    },
+    {
+      "epoch": 0.29639599222371765,
+      "grad_norm": 2.032954982306054,
+      "learning_rate": 8.25211772825567e-06,
+      "loss": 0.3959,
+      "step": 1982
+    },
+    {
+      "epoch": 0.2965455361148497,
+      "grad_norm": 2.0441581298148606,
+      "learning_rate": 8.250277796296412e-06,
+      "loss": 0.5914,
+      "step": 1983
+    },
+    {
+      "epoch": 0.29669508000598177,
+      "grad_norm": 1.0550678485707448,
+      "learning_rate": 8.248437101790663e-06,
+      "loss": 0.2187,
+      "step": 1984
+    },
+    {
+      "epoch": 0.2968446238971138,
+      "grad_norm": 1.714428768910598,
+      "learning_rate": 8.24659564517026e-06,
+      "loss": 0.3521,
+      "step": 1985
+    },
+    {
+      "epoch": 0.29699416778824583,
+      "grad_norm": 1.495931419848694,
+      "learning_rate": 8.244753426867233e-06,
+      "loss": 0.3141,
+      "step": 1986
+    },
+    {
+      "epoch": 0.2971437116793779,
+      "grad_norm": 1.4595624143758643,
+      "learning_rate": 8.24291044731378e-06,
+      "loss": 0.207,
+      "step": 1987
+    },
+    {
+      "epoch": 0.29729325557050995,
+      "grad_norm": 1.4893368784443541,
+      "learning_rate": 8.241066706942282e-06,
+      "loss": 0.2096,
+      "step": 1988
+    },
+    {
+      "epoch": 0.297442799461642,
+      "grad_norm": 1.5298733994754299,
+      "learning_rate": 8.239222206185303e-06,
+      "loss": 0.2963,
+      "step": 1989
+    },
+    {
+      "epoch": 0.297592343352774,
+      "grad_norm": 1.7110744142837584,
+      "learning_rate": 8.237376945475573e-06,
+      "loss": 0.4879,
+      "step": 1990
+    },
+    {
+      "epoch": 0.2977418872439061,
+      "grad_norm": 1.3646893921129908,
+      "learning_rate": 8.235530925246013e-06,
+      "loss": 0.2199,
+      "step": 1991
+    },
+    {
+      "epoch": 0.29789143113503813,
+      "grad_norm": 1.5519467713508963,
+      "learning_rate": 8.233684145929714e-06,
+      "loss": 0.1567,
+      "step": 1992
+    },
+    {
+      "epoch": 0.29804097502617016,
+      "grad_norm": 1.2022227133419618,
+      "learning_rate": 8.231836607959953e-06,
+      "loss": 0.3048,
+      "step": 1993
+    },
+    {
+      "epoch": 0.29819051891730225,
+      "grad_norm": 1.0528429237271446,
+      "learning_rate": 8.229988311770176e-06,
+      "loss": 0.2089,
+      "step": 1994
+    },
+    {
+      "epoch": 0.2983400628084343,
+      "grad_norm": 2.0589676323841233,
+      "learning_rate": 8.228139257794012e-06,
+      "loss": 0.4729,
+      "step": 1995
+    },
+    {
+      "epoch": 0.2984896066995663,
+      "grad_norm": 1.699439689917369,
+      "learning_rate": 8.226289446465269e-06,
+      "loss": 0.197,
+      "step": 1996
+    },
+    {
+      "epoch": 0.29863915059069834,
+      "grad_norm": 2.1597718626316214,
+      "learning_rate": 8.224438878217928e-06,
+      "loss": 0.5298,
+      "step": 1997
+    },
+    {
+      "epoch": 0.29878869448183043,
+      "grad_norm": 1.5026245898132045,
+      "learning_rate": 8.22258755348615e-06,
+      "loss": 0.4905,
+      "step": 1998
+    },
+    {
+      "epoch": 0.29893823837296246,
+      "grad_norm": 1.5917682855765136,
+      "learning_rate": 8.220735472704278e-06,
+      "loss": 0.2223,
+      "step": 1999
+    },
+    {
+      "epoch": 0.2990877822640945,
+      "grad_norm": 1.4766123425180093,
+      "learning_rate": 8.218882636306823e-06,
+      "loss": 0.2059,
+      "step": 2000
+    },
+    {
+      "epoch": 0.2992373261552266,
+      "grad_norm": 1.0856069277789302,
+      "learning_rate": 8.217029044728478e-06,
+      "loss": 0.193,
+      "step": 2001
+    },
+    {
+      "epoch": 0.2993868700463586,
+      "grad_norm": 1.2329171435469741,
+      "learning_rate": 8.215174698404118e-06,
+      "loss": 0.2372,
+      "step": 2002
+    },
+    {
+      "epoch": 0.29953641393749064,
+      "grad_norm": 1.7452583821562881,
+      "learning_rate": 8.213319597768785e-06,
+      "loss": 0.2098,
+      "step": 2003
+    },
+    {
+      "epoch": 0.2996859578286227,
+      "grad_norm": 1.0905531929281826,
+      "learning_rate": 8.21146374325771e-06,
+      "loss": 0.2187,
+      "step": 2004
+    },
+    {
+      "epoch": 0.29983550171975476,
+      "grad_norm": 1.816609674773371,
+      "learning_rate": 8.209607135306287e-06,
+      "loss": 0.3709,
+      "step": 2005
+    },
+    {
+      "epoch": 0.2999850456108868,
+      "grad_norm": 1.7738216295020472,
+      "learning_rate": 8.2077497743501e-06,
+      "loss": 0.4486,
+      "step": 2006
+    },
+    {
+      "epoch": 0.3001345895020188,
+      "grad_norm": 1.1263320970303174,
+      "learning_rate": 8.205891660824903e-06,
+      "loss": 0.2091,
+      "step": 2007
+    },
+    {
+      "epoch": 0.3002841333931509,
+      "grad_norm": 1.501664539486306,
+      "learning_rate": 8.204032795166625e-06,
+      "loss": 0.2009,
+      "step": 2008
+    },
+    {
+      "epoch": 0.30043367728428294,
+      "grad_norm": 2.6199919066049335,
+      "learning_rate": 8.202173177811374e-06,
+      "loss": 0.203,
+      "step": 2009
+    },
+    {
+      "epoch": 0.30058322117541497,
+      "grad_norm": 1.625937686489783,
+      "learning_rate": 8.200312809195436e-06,
+      "loss": 0.3831,
+      "step": 2010
+    },
+    {
+      "epoch": 0.30073276506654706,
+      "grad_norm": 1.8974675970112806,
+      "learning_rate": 8.198451689755269e-06,
+      "loss": 0.5418,
+      "step": 2011
+    },
+    {
+      "epoch": 0.3008823089576791,
+      "grad_norm": 1.8748455619787359,
+      "learning_rate": 8.196589819927512e-06,
+      "loss": 0.4425,
+      "step": 2012
+    },
+    {
+      "epoch": 0.3010318528488111,
+      "grad_norm": 1.6094667931646591,
+      "learning_rate": 8.194727200148978e-06,
+      "loss": 0.3389,
+      "step": 2013
+    },
+    {
+      "epoch": 0.30118139673994315,
+      "grad_norm": 0.9707433761295292,
+      "learning_rate": 8.192863830856652e-06,
+      "loss": 0.1763,
+      "step": 2014
+    },
+    {
+      "epoch": 0.30133094063107524,
+      "grad_norm": 1.3903941059235092,
+      "learning_rate": 8.1909997124877e-06,
+      "loss": 0.2339,
+      "step": 2015
+    },
+    {
+      "epoch": 0.30148048452220727,
+      "grad_norm": 1.3528795750135822,
+      "learning_rate": 8.189134845479462e-06,
+      "loss": 0.2193,
+      "step": 2016
+    },
+    {
+      "epoch": 0.3016300284133393,
+      "grad_norm": 1.7397719431371887,
+      "learning_rate": 8.187269230269458e-06,
+      "loss": 0.4049,
+      "step": 2017
+    },
+    {
+      "epoch": 0.3017795723044714,
+      "grad_norm": 1.626350301455843,
+      "learning_rate": 8.185402867295373e-06,
+      "loss": 0.2327,
+      "step": 2018
+    },
+    {
+      "epoch": 0.3019291161956034,
+      "grad_norm": 1.3932636681559976,
+      "learning_rate": 8.183535756995078e-06,
+      "loss": 0.2575,
+      "step": 2019
+    },
+    {
+      "epoch": 0.30207866008673545,
+      "grad_norm": 1.7626753702911266,
+      "learning_rate": 8.181667899806613e-06,
+      "loss": 0.4597,
+      "step": 2020
+    },
+    {
+      "epoch": 0.3022282039778675,
+      "grad_norm": 1.684539378891143,
+      "learning_rate": 8.179799296168194e-06,
+      "loss": 0.4608,
+      "step": 2021
+    },
+    {
+      "epoch": 0.30237774786899957,
+      "grad_norm": 1.0640465826224796,
+      "learning_rate": 8.177929946518217e-06,
+      "loss": 0.2221,
+      "step": 2022
+    },
+    {
+      "epoch": 0.3025272917601316,
+      "grad_norm": 1.5147026539291037,
+      "learning_rate": 8.176059851295248e-06,
+      "loss": 0.2353,
+      "step": 2023
+    },
+    {
+      "epoch": 0.30267683565126363,
+      "grad_norm": 1.7530304138656525,
+      "learning_rate": 8.174189010938028e-06,
+      "loss": 0.4141,
+      "step": 2024
+    },
+    {
+      "epoch": 0.3028263795423957,
+      "grad_norm": 1.9415201656454513,
+      "learning_rate": 8.172317425885477e-06,
+      "loss": 0.5352,
+      "step": 2025
+    },
+    {
+      "epoch": 0.30297592343352775,
+      "grad_norm": 1.9383678107900466,
+      "learning_rate": 8.170445096576683e-06,
+      "loss": 0.573,
+      "step": 2026
+    },
+    {
+      "epoch": 0.3031254673246598,
+      "grad_norm": 1.803033239816454,
+      "learning_rate": 8.168572023450915e-06,
+      "loss": 0.5225,
+      "step": 2027
+    },
+    {
+      "epoch": 0.3032750112157918,
+      "grad_norm": 2.1395850335411244,
+      "learning_rate": 8.166698206947614e-06,
+      "loss": 0.7007,
+      "step": 2028
+    },
+    {
+      "epoch": 0.3034245551069239,
+      "grad_norm": 1.5110862268685226,
+      "learning_rate": 8.164823647506394e-06,
+      "loss": 0.2269,
+      "step": 2029
+    },
+    {
+      "epoch": 0.30357409899805593,
+      "grad_norm": 1.6974467146899468,
+      "learning_rate": 8.162948345567048e-06,
+      "loss": 0.3921,
+      "step": 2030
+    },
+    {
+      "epoch": 0.30372364288918796,
+      "grad_norm": 1.3936816440378674,
+      "learning_rate": 8.161072301569536e-06,
+      "loss": 0.2049,
+      "step": 2031
+    },
+    {
+      "epoch": 0.30387318678032005,
+      "grad_norm": 1.4848287127408877,
+      "learning_rate": 8.159195515953998e-06,
+      "loss": 0.2062,
+      "step": 2032
+    },
+    {
+      "epoch": 0.3040227306714521,
+      "grad_norm": 1.2105792130825146,
+      "learning_rate": 8.157317989160746e-06,
+      "loss": 0.1788,
+      "step": 2033
+    },
+    {
+      "epoch": 0.3041722745625841,
+      "grad_norm": 1.3820320839391493,
+      "learning_rate": 8.155439721630265e-06,
+      "loss": 0.2331,
+      "step": 2034
+    },
+    {
+      "epoch": 0.30432181845371614,
+      "grad_norm": 2.150262549914666,
+      "learning_rate": 8.153560713803215e-06,
+      "loss": 0.1875,
+      "step": 2035
+    },
+    {
+      "epoch": 0.30447136234484823,
+      "grad_norm": 1.0906698589335555,
+      "learning_rate": 8.15168096612043e-06,
+      "loss": 0.2033,
+      "step": 2036
+    },
+    {
+      "epoch": 0.30462090623598026,
+      "grad_norm": 1.065733777923844,
+      "learning_rate": 8.149800479022917e-06,
+      "loss": 0.2001,
+      "step": 2037
+    },
+    {
+      "epoch": 0.3047704501271123,
+      "grad_norm": 1.6644796466482528,
+      "learning_rate": 8.147919252951855e-06,
+      "loss": 0.5517,
+      "step": 2038
+    },
+    {
+      "epoch": 0.3049199940182444,
+      "grad_norm": 1.5874527096275228,
+      "learning_rate": 8.146037288348598e-06,
+      "loss": 0.3491,
+      "step": 2039
+    },
+    {
+      "epoch": 0.3050695379093764,
+      "grad_norm": 2.1160201341828513,
+      "learning_rate": 8.144154585654675e-06,
+      "loss": 0.4023,
+      "step": 2040
+    },
+    {
+      "epoch": 0.30521908180050844,
+      "grad_norm": 1.449019528908297,
+      "learning_rate": 8.142271145311784e-06,
+      "loss": 0.3529,
+      "step": 2041
+    },
+    {
+      "epoch": 0.30536862569164047,
+      "grad_norm": 1.0741990035582996,
+      "learning_rate": 8.1403869677618e-06,
+      "loss": 0.2042,
+      "step": 2042
+    },
+    {
+      "epoch": 0.30551816958277256,
+      "grad_norm": 1.6081149900134495,
+      "learning_rate": 8.138502053446766e-06,
+      "loss": 0.3908,
+      "step": 2043
+    },
+    {
+      "epoch": 0.3056677134739046,
+      "grad_norm": 2.0690220211720716,
+      "learning_rate": 8.136616402808906e-06,
+      "loss": 0.5508,
+      "step": 2044
+    },
+    {
+      "epoch": 0.3058172573650366,
+      "grad_norm": 1.6409484884961714,
+      "learning_rate": 8.13473001629061e-06,
+      "loss": 0.3161,
+      "step": 2045
+    },
+    {
+      "epoch": 0.3059668012561687,
+      "grad_norm": 1.858300355296472,
+      "learning_rate": 8.132842894334438e-06,
+      "loss": 0.5614,
+      "step": 2046
+    },
+    {
+      "epoch": 0.30611634514730074,
+      "grad_norm": 1.2324767485914987,
+      "learning_rate": 8.130955037383132e-06,
+      "loss": 0.2132,
+      "step": 2047
+    },
+    {
+      "epoch": 0.30626588903843277,
+      "grad_norm": 1.7272643652056465,
+      "learning_rate": 8.1290664458796e-06,
+      "loss": 0.3132,
+      "step": 2048
+    },
+    {
+      "epoch": 0.3064154329295648,
+      "grad_norm": 1.3637333594365966,
+      "learning_rate": 8.127177120266926e-06,
+      "loss": 0.1739,
+      "step": 2049
+    },
+    {
+      "epoch": 0.3065649768206969,
+      "grad_norm": 1.538728380001006,
+      "learning_rate": 8.12528706098836e-06,
+      "loss": 0.4072,
+      "step": 2050
+    },
+    {
+      "epoch": 0.3067145207118289,
+      "grad_norm": 1.4591751944118272,
+      "learning_rate": 8.12339626848733e-06,
+      "loss": 0.3859,
+      "step": 2051
+    },
+    {
+      "epoch": 0.30686406460296095,
+      "grad_norm": 1.9028261579580017,
+      "learning_rate": 8.121504743207436e-06,
+      "loss": 0.4893,
+      "step": 2052
+    },
+    {
+      "epoch": 0.30701360849409304,
+      "grad_norm": 1.7379512730839208,
+      "learning_rate": 8.119612485592442e-06,
+      "loss": 0.3564,
+      "step": 2053
+    },
+    {
+      "epoch": 0.30716315238522507,
+      "grad_norm": 1.388968630495014,
+      "learning_rate": 8.117719496086298e-06,
+      "loss": 0.3179,
+      "step": 2054
+    },
+    {
+      "epoch": 0.3073126962763571,
+      "grad_norm": 1.403056196180425,
+      "learning_rate": 8.115825775133112e-06,
+      "loss": 0.352,
+      "step": 2055
+    },
+    {
+      "epoch": 0.30746224016748913,
+      "grad_norm": 1.508992427712604,
+      "learning_rate": 8.113931323177171e-06,
+      "loss": 0.4021,
+      "step": 2056
+    },
+    {
+      "epoch": 0.3076117840586212,
+      "grad_norm": 2.1741360894355015,
+      "learning_rate": 8.112036140662931e-06,
+      "loss": 0.467,
+      "step": 2057
+    },
+    {
+      "epoch": 0.30776132794975325,
+      "grad_norm": 1.6144568383492353,
+      "learning_rate": 8.110140228035022e-06,
+      "loss": 0.3924,
+      "step": 2058
+    },
+    {
+      "epoch": 0.3079108718408853,
+      "grad_norm": 1.4893071701125404,
+      "learning_rate": 8.10824358573824e-06,
+      "loss": 0.5036,
+      "step": 2059
+    },
+    {
+      "epoch": 0.30806041573201737,
+      "grad_norm": 1.4815595377852415,
+      "learning_rate": 8.10634621421756e-06,
+      "loss": 0.3972,
+      "step": 2060
+    },
+    {
+      "epoch": 0.3082099596231494,
+      "grad_norm": 1.7353454529628423,
+      "learning_rate": 8.104448113918118e-06,
+      "loss": 0.2237,
+      "step": 2061
+    },
+    {
+      "epoch": 0.30835950351428143,
+      "grad_norm": 1.0247854772257274,
+      "learning_rate": 8.102549285285233e-06,
+      "loss": 0.1595,
+      "step": 2062
+    },
+    {
+      "epoch": 0.30850904740541346,
+      "grad_norm": 1.6101502986170517,
+      "learning_rate": 8.100649728764382e-06,
+      "loss": 0.519,
+      "step": 2063
+    },
+    {
+      "epoch": 0.30865859129654555,
+      "grad_norm": 1.4785633507789944,
+      "learning_rate": 8.098749444801226e-06,
+      "loss": 0.3757,
+      "step": 2064
+    },
+    {
+      "epoch": 0.3088081351876776,
+      "grad_norm": 1.1653508922546973,
+      "learning_rate": 8.096848433841585e-06,
+      "loss": 0.1843,
+      "step": 2065
+    },
+    {
+      "epoch": 0.3089576790788096,
+      "grad_norm": 2.23989146500782,
+      "learning_rate": 8.094946696331454e-06,
+      "loss": 0.5717,
+      "step": 2066
+    },
+    {
+      "epoch": 0.3091072229699417,
+      "grad_norm": 1.4086002710437948,
+      "learning_rate": 8.093044232717004e-06,
+      "loss": 0.3305,
+      "step": 2067
+    },
+    {
+      "epoch": 0.30925676686107373,
+      "grad_norm": 1.3532349541812738,
+      "learning_rate": 8.091141043444565e-06,
+      "loss": 0.2031,
+      "step": 2068
+    },
+    {
+      "epoch": 0.30940631075220576,
+      "grad_norm": 1.5659716733424975,
+      "learning_rate": 8.08923712896065e-06,
+      "loss": 0.264,
+      "step": 2069
+    },
+    {
+      "epoch": 0.30955585464333785,
+      "grad_norm": 1.188194674034748,
+      "learning_rate": 8.087332489711931e-06,
+      "loss": 0.2084,
+      "step": 2070
+    },
+    {
+      "epoch": 0.3097053985344699,
+      "grad_norm": 1.3639153806477138,
+      "learning_rate": 8.085427126145255e-06,
+      "loss": 0.2999,
+      "step": 2071
+    },
+    {
+      "epoch": 0.3098549424256019,
+      "grad_norm": 1.743026958811897,
+      "learning_rate": 8.083521038707643e-06,
+      "loss": 0.5313,
+      "step": 2072
+    },
+    {
+      "epoch": 0.31000448631673394,
+      "grad_norm": 1.1747827948023122,
+      "learning_rate": 8.081614227846275e-06,
+      "loss": 0.1753,
+      "step": 2073
+    },
+    {
+      "epoch": 0.310154030207866,
+      "grad_norm": 1.1941064963550336,
+      "learning_rate": 8.079706694008512e-06,
+      "loss": 0.2424,
+      "step": 2074
+    },
+    {
+      "epoch": 0.31030357409899806,
+      "grad_norm": 1.7186546887175254,
+      "learning_rate": 8.077798437641878e-06,
+      "loss": 0.2263,
+      "step": 2075
+    },
+    {
+      "epoch": 0.3104531179901301,
+      "grad_norm": 1.0317324156623773,
+      "learning_rate": 8.075889459194069e-06,
+      "loss": 0.1556,
+      "step": 2076
+    },
+    {
+      "epoch": 0.3106026618812622,
+      "grad_norm": 1.5907870570782063,
+      "learning_rate": 8.073979759112949e-06,
+      "loss": 0.2729,
+      "step": 2077
+    },
+    {
+      "epoch": 0.3107522057723942,
+      "grad_norm": 1.8674735165445506,
+      "learning_rate": 8.072069337846554e-06,
+      "loss": 0.3821,
+      "step": 2078
+    },
+    {
+      "epoch": 0.31090174966352624,
+      "grad_norm": 1.983553521137882,
+      "learning_rate": 8.070158195843084e-06,
+      "loss": 0.6787,
+      "step": 2079
+    },
+    {
+      "epoch": 0.31105129355465827,
+      "grad_norm": 1.0817414697908496,
+      "learning_rate": 8.068246333550913e-06,
+      "loss": 0.1601,
+      "step": 2080
+    },
+    {
+      "epoch": 0.31120083744579036,
+      "grad_norm": 1.4614119919401352,
+      "learning_rate": 8.066333751418582e-06,
+      "loss": 0.2074,
+      "step": 2081
+    },
+    {
+      "epoch": 0.3113503813369224,
+      "grad_norm": 1.2801507618992622,
+      "learning_rate": 8.064420449894802e-06,
+      "loss": 0.2384,
+      "step": 2082
+    },
+    {
+      "epoch": 0.3114999252280544,
+      "grad_norm": 1.856756460170748,
+      "learning_rate": 8.062506429428451e-06,
+      "loss": 0.3569,
+      "step": 2083
+    },
+    {
+      "epoch": 0.3116494691191865,
+      "grad_norm": 1.5574561714489221,
+      "learning_rate": 8.060591690468579e-06,
+      "loss": 0.2001,
+      "step": 2084
+    },
+    {
+      "epoch": 0.31179901301031854,
+      "grad_norm": 0.8139823838456822,
+      "learning_rate": 8.0586762334644e-06,
+      "loss": 0.1493,
+      "step": 2085
+    },
+    {
+      "epoch": 0.31194855690145057,
+      "grad_norm": 1.8147690398413907,
+      "learning_rate": 8.056760058865298e-06,
+      "loss": 0.3726,
+      "step": 2086
+    },
+    {
+      "epoch": 0.3120981007925826,
+      "grad_norm": 1.3551230470716993,
+      "learning_rate": 8.054843167120827e-06,
+      "loss": 0.3161,
+      "step": 2087
+    },
+    {
+      "epoch": 0.3122476446837147,
+      "grad_norm": 1.423454170150374,
+      "learning_rate": 8.052925558680708e-06,
+      "loss": 0.3286,
+      "step": 2088
+    },
+    {
+      "epoch": 0.3123971885748467,
+      "grad_norm": 1.1629738020950542,
+      "learning_rate": 8.051007233994833e-06,
+      "loss": 0.2128,
+      "step": 2089
+    },
+    {
+      "epoch": 0.31254673246597875,
+      "grad_norm": 1.0328139570935773,
+      "learning_rate": 8.049088193513257e-06,
+      "loss": 0.2028,
+      "step": 2090
+    },
+    {
+      "epoch": 0.31269627635711084,
+      "grad_norm": 0.9561546972801208,
+      "learning_rate": 8.047168437686204e-06,
+      "loss": 0.1901,
+      "step": 2091
+    },
+    {
+      "epoch": 0.31284582024824287,
+      "grad_norm": 1.5085687791327238,
+      "learning_rate": 8.045247966964069e-06,
+      "loss": 0.3978,
+      "step": 2092
+    },
+    {
+      "epoch": 0.3129953641393749,
+      "grad_norm": 1.7990952747958247,
+      "learning_rate": 8.043326781797413e-06,
+      "loss": 0.4526,
+      "step": 2093
+    },
+    {
+      "epoch": 0.31314490803050693,
+      "grad_norm": 1.6047231005366211,
+      "learning_rate": 8.041404882636964e-06,
+      "loss": 0.3136,
+      "step": 2094
+    },
+    {
+      "epoch": 0.313294451921639,
+      "grad_norm": 1.0885180525142448,
+      "learning_rate": 8.039482269933619e-06,
+      "loss": 0.1868,
+      "step": 2095
+    },
+    {
+      "epoch": 0.31344399581277105,
+      "grad_norm": 1.6731122777438763,
+      "learning_rate": 8.03755894413844e-06,
+      "loss": 0.253,
+      "step": 2096
+    },
+    {
+      "epoch": 0.3135935397039031,
+      "grad_norm": 1.7388060302674344,
+      "learning_rate": 8.03563490570266e-06,
+      "loss": 0.3781,
+      "step": 2097
+    },
+    {
+      "epoch": 0.31374308359503517,
+      "grad_norm": 1.3948567328495716,
+      "learning_rate": 8.033710155077675e-06,
+      "loss": 0.3065,
+      "step": 2098
+    },
+    {
+      "epoch": 0.3138926274861672,
+      "grad_norm": 1.7899046169473158,
+      "learning_rate": 8.03178469271505e-06,
+      "loss": 0.4497,
+      "step": 2099
+    },
+    {
+      "epoch": 0.31404217137729923,
+      "grad_norm": 1.1376056341100615,
+      "learning_rate": 8.029858519066519e-06,
+      "loss": 0.1793,
+      "step": 2100
+    },
+    {
+      "epoch": 0.31419171526843126,
+      "grad_norm": 1.4405559368087475,
+      "learning_rate": 8.027931634583978e-06,
+      "loss": 0.1984,
+      "step": 2101
+    },
+    {
+      "epoch": 0.31434125915956335,
+      "grad_norm": 1.7393459445529194,
+      "learning_rate": 8.026004039719494e-06,
+      "loss": 0.4943,
+      "step": 2102
+    },
+    {
+      "epoch": 0.3144908030506954,
+      "grad_norm": 1.6507941300141975,
+      "learning_rate": 8.024075734925302e-06,
+      "loss": 0.3702,
+      "step": 2103
+    },
+    {
+      "epoch": 0.3146403469418274,
+      "grad_norm": 1.175812152126952,
+      "learning_rate": 8.022146720653797e-06,
+      "loss": 0.2246,
+      "step": 2104
+    },
+    {
+      "epoch": 0.3147898908329595,
+      "grad_norm": 1.505696425174867,
+      "learning_rate": 8.020216997357547e-06,
+      "loss": 0.3827,
+      "step": 2105
+    },
+    {
+      "epoch": 0.3149394347240915,
+      "grad_norm": 1.5053350968315022,
+      "learning_rate": 8.018286565489281e-06,
+      "loss": 0.3602,
+      "step": 2106
+    },
+    {
+      "epoch": 0.31508897861522356,
+      "grad_norm": 7.395352954217879,
+      "learning_rate": 8.016355425501899e-06,
+      "loss": 0.5292,
+      "step": 2107
+    },
+    {
+      "epoch": 0.3152385225063556,
+      "grad_norm": 1.4124551721650014,
+      "learning_rate": 8.014423577848465e-06,
+      "loss": 0.3058,
+      "step": 2108
+    },
+    {
+      "epoch": 0.3153880663974877,
+      "grad_norm": 1.7291348054164273,
+      "learning_rate": 8.012491022982206e-06,
+      "loss": 0.4039,
+      "step": 2109
+    },
+    {
+      "epoch": 0.3155376102886197,
+      "grad_norm": 1.3762981669774148,
+      "learning_rate": 8.010557761356523e-06,
+      "loss": 0.3214,
+      "step": 2110
+    },
+    {
+      "epoch": 0.31568715417975174,
+      "grad_norm": 1.9151602263688279,
+      "learning_rate": 8.008623793424975e-06,
+      "loss": 0.5339,
+      "step": 2111
+    },
+    {
+      "epoch": 0.3158366980708838,
+      "grad_norm": 1.6710466712643997,
+      "learning_rate": 8.006689119641289e-06,
+      "loss": 0.3143,
+      "step": 2112
+    },
+    {
+      "epoch": 0.31598624196201586,
+      "grad_norm": 1.2763038771322965,
+      "learning_rate": 8.00475374045936e-06,
+      "loss": 0.3106,
+      "step": 2113
+    },
+    {
+      "epoch": 0.3161357858531479,
+      "grad_norm": 0.9519969897990209,
+      "learning_rate": 8.002817656333246e-06,
+      "loss": 0.1709,
+      "step": 2114
+    },
+    {
+      "epoch": 0.3162853297442799,
+      "grad_norm": 1.0941737648338323,
+      "learning_rate": 8.000880867717168e-06,
+      "loss": 0.3558,
+      "step": 2115
+    },
+    {
+      "epoch": 0.316434873635412,
+      "grad_norm": 1.6804061150860001,
+      "learning_rate": 7.99894337506552e-06,
+      "loss": 0.3005,
+      "step": 2116
+    },
+    {
+      "epoch": 0.31658441752654404,
+      "grad_norm": 1.1783654687222251,
+      "learning_rate": 7.997005178832853e-06,
+      "loss": 0.2406,
+      "step": 2117
+    },
+    {
+      "epoch": 0.31673396141767607,
+      "grad_norm": 1.6348023119949275,
+      "learning_rate": 7.99506627947389e-06,
+      "loss": 0.2475,
+      "step": 2118
+    },
+    {
+      "epoch": 0.31688350530880816,
+      "grad_norm": 1.4962501892290498,
+      "learning_rate": 7.993126677443513e-06,
+      "loss": 0.3126,
+      "step": 2119
+    },
+    {
+      "epoch": 0.3170330491999402,
+      "grad_norm": 1.1269074931566945,
+      "learning_rate": 7.991186373196771e-06,
+      "loss": 0.2101,
+      "step": 2120
+    },
+    {
+      "epoch": 0.3171825930910722,
+      "grad_norm": 1.5451433092999418,
+      "learning_rate": 7.989245367188877e-06,
+      "loss": 0.4435,
+      "step": 2121
+    },
+    {
+      "epoch": 0.31733213698220425,
+      "grad_norm": 1.6954504563414547,
+      "learning_rate": 7.987303659875212e-06,
+      "loss": 0.3777,
+      "step": 2122
+    },
+    {
+      "epoch": 0.31748168087333634,
+      "grad_norm": 1.7046868376726214,
+      "learning_rate": 7.98536125171132e-06,
+      "loss": 0.375,
+      "step": 2123
+    },
+    {
+      "epoch": 0.31763122476446837,
+      "grad_norm": 1.6583768126843783,
+      "learning_rate": 7.983418143152906e-06,
+      "loss": 0.1679,
+      "step": 2124
+    },
+    {
+      "epoch": 0.3177807686556004,
+      "grad_norm": 1.652523256475839,
+      "learning_rate": 7.981474334655845e-06,
+      "loss": 0.3821,
+      "step": 2125
+    },
+    {
+      "epoch": 0.3179303125467325,
+      "grad_norm": 1.6865422789510223,
+      "learning_rate": 7.979529826676172e-06,
+      "loss": 0.2312,
+      "step": 2126
+    },
+    {
+      "epoch": 0.3180798564378645,
+      "grad_norm": 1.318155763232294,
+      "learning_rate": 7.977584619670084e-06,
+      "loss": 0.2417,
+      "step": 2127
+    },
+    {
+      "epoch": 0.31822940032899655,
+      "grad_norm": 0.9387354141491452,
+      "learning_rate": 7.97563871409395e-06,
+      "loss": 0.2152,
+      "step": 2128
+    },
+    {
+      "epoch": 0.31837894422012863,
+      "grad_norm": 1.4740638203723162,
+      "learning_rate": 7.973692110404295e-06,
+      "loss": 0.3663,
+      "step": 2129
+    },
+    {
+      "epoch": 0.31852848811126067,
+      "grad_norm": 1.5147303848676803,
+      "learning_rate": 7.971744809057815e-06,
+      "loss": 0.2317,
+      "step": 2130
+    },
+    {
+      "epoch": 0.3186780320023927,
+      "grad_norm": 1.3220893173215598,
+      "learning_rate": 7.96979681051136e-06,
+      "loss": 0.3673,
+      "step": 2131
+    },
+    {
+      "epoch": 0.31882757589352473,
+      "grad_norm": 1.778728683314461,
+      "learning_rate": 7.967848115221953e-06,
+      "loss": 0.3838,
+      "step": 2132
+    },
+    {
+      "epoch": 0.3189771197846568,
+      "grad_norm": 1.79666907747646,
+      "learning_rate": 7.965898723646777e-06,
+      "loss": 0.4668,
+      "step": 2133
+    },
+    {
+      "epoch": 0.31912666367578885,
+      "grad_norm": 1.1363334854959442,
+      "learning_rate": 7.963948636243175e-06,
+      "loss": 0.1881,
+      "step": 2134
+    },
+    {
+      "epoch": 0.3192762075669209,
+      "grad_norm": 1.2727017471208417,
+      "learning_rate": 7.96199785346866e-06,
+      "loss": 0.3583,
+      "step": 2135
+    },
+    {
+      "epoch": 0.31942575145805296,
+      "grad_norm": 1.6879630689693983,
+      "learning_rate": 7.960046375780903e-06,
+      "loss": 0.528,
+      "step": 2136
+    },
+    {
+      "epoch": 0.319575295349185,
+      "grad_norm": 1.441558984867543,
+      "learning_rate": 7.958094203637738e-06,
+      "loss": 0.3621,
+      "step": 2137
+    },
+    {
+      "epoch": 0.319724839240317,
+      "grad_norm": 0.9760649090504697,
+      "learning_rate": 7.956141337497166e-06,
+      "loss": 0.1714,
+      "step": 2138
+    },
+    {
+      "epoch": 0.31987438313144906,
+      "grad_norm": 0.9469336665426322,
+      "learning_rate": 7.954187777817345e-06,
+      "loss": 0.1795,
+      "step": 2139
+    },
+    {
+      "epoch": 0.32002392702258114,
+      "grad_norm": 1.0476640767317869,
+      "learning_rate": 7.952233525056603e-06,
+      "loss": 0.1922,
+      "step": 2140
+    },
+    {
+      "epoch": 0.3201734709137132,
+      "grad_norm": 1.254945459546166,
+      "learning_rate": 7.950278579673422e-06,
+      "loss": 0.2742,
+      "step": 2141
+    },
+    {
+      "epoch": 0.3203230148048452,
+      "grad_norm": 0.7230488496853383,
+      "learning_rate": 7.948322942126456e-06,
+      "loss": 0.1742,
+      "step": 2142
+    },
+    {
+      "epoch": 0.3204725586959773,
+      "grad_norm": 0.960182470952214,
+      "learning_rate": 7.946366612874512e-06,
+      "loss": 0.1851,
+      "step": 2143
+    },
+    {
+      "epoch": 0.3206221025871093,
+      "grad_norm": 1.1966344295913398,
+      "learning_rate": 7.944409592376565e-06,
+      "loss": 0.1877,
+      "step": 2144
+    },
+    {
+      "epoch": 0.32077164647824136,
+      "grad_norm": 1.9057275611187459,
+      "learning_rate": 7.942451881091752e-06,
+      "loss": 0.2618,
+      "step": 2145
+    },
+    {
+      "epoch": 0.3209211903693734,
+      "grad_norm": 1.2274689028407764,
+      "learning_rate": 7.94049347947937e-06,
+      "loss": 0.1885,
+      "step": 2146
+    },
+    {
+      "epoch": 0.3210707342605055,
+      "grad_norm": 1.678924591333489,
+      "learning_rate": 7.93853438799888e-06,
+      "loss": 0.2338,
+      "step": 2147
+    },
+    {
+      "epoch": 0.3212202781516375,
+      "grad_norm": 1.7464181014190676,
+      "learning_rate": 7.936574607109901e-06,
+      "loss": 0.3676,
+      "step": 2148
+    },
+    {
+      "epoch": 0.32136982204276954,
+      "grad_norm": 1.5398927642089624,
+      "learning_rate": 7.934614137272218e-06,
+      "loss": 0.2293,
+      "step": 2149
+    },
+    {
+      "epoch": 0.3215193659339016,
+      "grad_norm": 1.5519843635324424,
+      "learning_rate": 7.932652978945779e-06,
+      "loss": 0.199,
+      "step": 2150
+    },
+    {
+      "epoch": 0.32166890982503366,
+      "grad_norm": 1.1920599004222825,
+      "learning_rate": 7.930691132590686e-06,
+      "loss": 0.1696,
+      "step": 2151
+    },
+    {
+      "epoch": 0.3218184537161657,
+      "grad_norm": 0.8145849226466982,
+      "learning_rate": 7.928728598667211e-06,
+      "loss": 0.1537,
+      "step": 2152
+    },
+    {
+      "epoch": 0.3219679976072977,
+      "grad_norm": 1.3926444820307884,
+      "learning_rate": 7.926765377635781e-06,
+      "loss": 0.1999,
+      "step": 2153
+    },
+    {
+      "epoch": 0.3221175414984298,
+      "grad_norm": 1.6412886866609875,
+      "learning_rate": 7.924801469956986e-06,
+      "loss": 0.2527,
+      "step": 2154
+    },
+    {
+      "epoch": 0.32226708538956184,
+      "grad_norm": 1.5325273883442905,
+      "learning_rate": 7.92283687609158e-06,
+      "loss": 0.2634,
+      "step": 2155
+    },
+    {
+      "epoch": 0.32241662928069387,
+      "grad_norm": 1.0316905287286056,
+      "learning_rate": 7.920871596500473e-06,
+      "loss": 0.174,
+      "step": 2156
+    },
+    {
+      "epoch": 0.32256617317182595,
+      "grad_norm": 1.1349047741297784,
+      "learning_rate": 7.91890563164474e-06,
+      "loss": 0.1721,
+      "step": 2157
+    },
+    {
+      "epoch": 0.322715717062958,
+      "grad_norm": 2.057460755429332,
+      "learning_rate": 7.916938981985619e-06,
+      "loss": 0.6121,
+      "step": 2158
+    },
+    {
+      "epoch": 0.32286526095409,
+      "grad_norm": 1.236449628888766,
+      "learning_rate": 7.914971647984494e-06,
+      "loss": 0.1979,
+      "step": 2159
+    },
+    {
+      "epoch": 0.32301480484522205,
+      "grad_norm": 1.952661192183289,
+      "learning_rate": 7.913003630102934e-06,
+      "loss": 0.4105,
+      "step": 2160
+    },
+    {
+      "epoch": 0.32316434873635413,
+      "grad_norm": 1.6581077143849778,
+      "learning_rate": 7.911034928802647e-06,
+      "loss": 0.4989,
+      "step": 2161
+    },
+    {
+      "epoch": 0.32331389262748617,
+      "grad_norm": 1.0964604376832452,
+      "learning_rate": 7.909065544545511e-06,
+      "loss": 0.2086,
+      "step": 2162
+    },
+    {
+      "epoch": 0.3234634365186182,
+      "grad_norm": 1.4794895335762217,
+      "learning_rate": 7.907095477793563e-06,
+      "loss": 0.2232,
+      "step": 2163
+    },
+    {
+      "epoch": 0.3236129804097503,
+      "grad_norm": 1.7007378332145011,
+      "learning_rate": 7.905124729008996e-06,
+      "loss": 0.3449,
+      "step": 2164
+    },
+    {
+      "epoch": 0.3237625243008823,
+      "grad_norm": 5.5988236758577,
+      "learning_rate": 7.903153298654173e-06,
+      "loss": 0.2263,
+      "step": 2165
+    },
+    {
+      "epoch": 0.32391206819201435,
+      "grad_norm": 1.7620287916310595,
+      "learning_rate": 7.901181187191606e-06,
+      "loss": 0.4116,
+      "step": 2166
+    },
+    {
+      "epoch": 0.3240616120831464,
+      "grad_norm": 1.6633135381694006,
+      "learning_rate": 7.899208395083974e-06,
+      "loss": 0.2297,
+      "step": 2167
+    },
+    {
+      "epoch": 0.32421115597427846,
+      "grad_norm": 0.9621828860582361,
+      "learning_rate": 7.897234922794113e-06,
+      "loss": 0.2098,
+      "step": 2168
+    },
+    {
+      "epoch": 0.3243606998654105,
+      "grad_norm": 1.3513242795516456,
+      "learning_rate": 7.895260770785014e-06,
+      "loss": 0.3103,
+      "step": 2169
+    },
+    {
+      "epoch": 0.3245102437565425,
+      "grad_norm": 0.9735967195147534,
+      "learning_rate": 7.893285939519836e-06,
+      "loss": 0.1699,
+      "step": 2170
+    },
+    {
+      "epoch": 0.3246597876476746,
+      "grad_norm": 2.0453322690120554,
+      "learning_rate": 7.891310429461895e-06,
+      "loss": 0.5192,
+      "step": 2171
+    },
+    {
+      "epoch": 0.32480933153880664,
+      "grad_norm": 1.4629318475963238,
+      "learning_rate": 7.889334241074663e-06,
+      "loss": 0.4312,
+      "step": 2172
+    },
+    {
+      "epoch": 0.3249588754299387,
+      "grad_norm": 1.5333310991077456,
+      "learning_rate": 7.887357374821768e-06,
+      "loss": 0.2691,
+      "step": 2173
+    },
+    {
+      "epoch": 0.3251084193210707,
+      "grad_norm": 1.6289642332812333,
+      "learning_rate": 7.88537983116701e-06,
+      "loss": 0.4631,
+      "step": 2174
+    },
+    {
+      "epoch": 0.3252579632122028,
+      "grad_norm": 1.166966622363864,
+      "learning_rate": 7.883401610574338e-06,
+      "loss": 0.1523,
+      "step": 2175
+    },
+    {
+      "epoch": 0.3254075071033348,
+      "grad_norm": 1.7643050095516288,
+      "learning_rate": 7.881422713507857e-06,
+      "loss": 0.4944,
+      "step": 2176
+    },
+    {
+      "epoch": 0.32555705099446686,
+      "grad_norm": 2.0778286729797943,
+      "learning_rate": 7.879443140431837e-06,
+      "loss": 0.5285,
+      "step": 2177
+    },
+    {
+      "epoch": 0.32570659488559894,
+      "grad_norm": 1.22732284237921,
+      "learning_rate": 7.877462891810708e-06,
+      "loss": 0.2215,
+      "step": 2178
+    },
+    {
+      "epoch": 0.325856138776731,
+      "grad_norm": 1.794981875573838,
+      "learning_rate": 7.875481968109052e-06,
+      "loss": 0.4738,
+      "step": 2179
+    },
+    {
+      "epoch": 0.326005682667863,
+      "grad_norm": 1.8625636048305199,
+      "learning_rate": 7.873500369791615e-06,
+      "loss": 0.5112,
+      "step": 2180
+    },
+    {
+      "epoch": 0.3261552265589951,
+      "grad_norm": 1.5970308413484968,
+      "learning_rate": 7.8715180973233e-06,
+      "loss": 0.3491,
+      "step": 2181
+    },
+    {
+      "epoch": 0.3263047704501271,
+      "grad_norm": 4.445014172618314,
+      "learning_rate": 7.869535151169163e-06,
+      "loss": 0.4128,
+      "step": 2182
+    },
+    {
+      "epoch": 0.32645431434125916,
+      "grad_norm": 2.199504833687433,
+      "learning_rate": 7.867551531794427e-06,
+      "loss": 0.5042,
+      "step": 2183
+    },
+    {
+      "epoch": 0.3266038582323912,
+      "grad_norm": 1.9133519547070805,
+      "learning_rate": 7.865567239664463e-06,
+      "loss": 0.5783,
+      "step": 2184
+    },
+    {
+      "epoch": 0.3267534021235233,
+      "grad_norm": 1.847425837596219,
+      "learning_rate": 7.86358227524481e-06,
+      "loss": 0.5434,
+      "step": 2185
+    },
+    {
+      "epoch": 0.3269029460146553,
+      "grad_norm": 1.7506088747332584,
+      "learning_rate": 7.861596639001157e-06,
+      "loss": 0.3848,
+      "step": 2186
+    },
+    {
+      "epoch": 0.32705248990578734,
+      "grad_norm": 1.6285208330736813,
+      "learning_rate": 7.859610331399354e-06,
+      "loss": 0.2588,
+      "step": 2187
+    },
+    {
+      "epoch": 0.3272020337969194,
+      "grad_norm": 1.0354818947520257,
+      "learning_rate": 7.85762335290541e-06,
+      "loss": 0.1884,
+      "step": 2188
+    },
+    {
+      "epoch": 0.32735157768805145,
+      "grad_norm": 2.061471227348859,
+      "learning_rate": 7.855635703985487e-06,
+      "loss": 0.5502,
+      "step": 2189
+    },
+    {
+      "epoch": 0.3275011215791835,
+      "grad_norm": 1.0589713974920532,
+      "learning_rate": 7.853647385105905e-06,
+      "loss": 0.1894,
+      "step": 2190
+    },
+    {
+      "epoch": 0.3276506654703155,
+      "grad_norm": 1.27660152324297,
+      "learning_rate": 7.851658396733148e-06,
+      "loss": 0.2325,
+      "step": 2191
+    },
+    {
+      "epoch": 0.3278002093614476,
+      "grad_norm": 1.602286282256986,
+      "learning_rate": 7.849668739333846e-06,
+      "loss": 0.1959,
+      "step": 2192
+    },
+    {
+      "epoch": 0.32794975325257963,
+      "grad_norm": 1.3119010750867572,
+      "learning_rate": 7.847678413374795e-06,
+      "loss": 0.1982,
+      "step": 2193
+    },
+    {
+      "epoch": 0.32809929714371167,
+      "grad_norm": 1.2840343160987466,
+      "learning_rate": 7.845687419322945e-06,
+      "loss": 0.2299,
+      "step": 2194
+    },
+    {
+      "epoch": 0.32824884103484375,
+      "grad_norm": 1.5111699549772195,
+      "learning_rate": 7.843695757645402e-06,
+      "loss": 0.2698,
+      "step": 2195
+    },
+    {
+      "epoch": 0.3283983849259758,
+      "grad_norm": 1.1728180317528463,
+      "learning_rate": 7.841703428809426e-06,
+      "loss": 0.1911,
+      "step": 2196
+    },
+    {
+      "epoch": 0.3285479288171078,
+      "grad_norm": 1.623773790762259,
+      "learning_rate": 7.839710433282441e-06,
+      "loss": 0.4539,
+      "step": 2197
+    },
+    {
+      "epoch": 0.32869747270823985,
+      "grad_norm": 1.3523478321165887,
+      "learning_rate": 7.83771677153202e-06,
+      "loss": 0.1695,
+      "step": 2198
+    },
+    {
+      "epoch": 0.32884701659937193,
+      "grad_norm": 1.669502440977397,
+      "learning_rate": 7.835722444025898e-06,
+      "loss": 0.2157,
+      "step": 2199
+    },
+    {
+      "epoch": 0.32899656049050396,
+      "grad_norm": 1.3304613082395382,
+      "learning_rate": 7.83372745123196e-06,
+      "loss": 0.1949,
+      "step": 2200
+    },
+    {
+      "epoch": 0.329146104381636,
+      "grad_norm": 1.857526998824266,
+      "learning_rate": 7.831731793618253e-06,
+      "loss": 0.2801,
+      "step": 2201
+    },
+    {
+      "epoch": 0.3292956482727681,
+      "grad_norm": 1.4165319093489228,
+      "learning_rate": 7.829735471652978e-06,
+      "loss": 0.3415,
+      "step": 2202
+    },
+    {
+      "epoch": 0.3294451921639001,
+      "grad_norm": 1.193763566127306,
+      "learning_rate": 7.827738485804488e-06,
+      "loss": 0.1967,
+      "step": 2203
+    },
+    {
+      "epoch": 0.32959473605503214,
+      "grad_norm": 1.4097654443065837,
+      "learning_rate": 7.825740836541299e-06,
+      "loss": 0.2322,
+      "step": 2204
+    },
+    {
+      "epoch": 0.3297442799461642,
+      "grad_norm": 0.8446160856119209,
+      "learning_rate": 7.823742524332074e-06,
+      "loss": 0.1748,
+      "step": 2205
+    },
+    {
+      "epoch": 0.32989382383729626,
+      "grad_norm": 1.4283812084385221,
+      "learning_rate": 7.821743549645642e-06,
+      "loss": 0.2485,
+      "step": 2206
+    },
+    {
+      "epoch": 0.3300433677284283,
+      "grad_norm": 1.522841097169646,
+      "learning_rate": 7.819743912950979e-06,
+      "loss": 0.3102,
+      "step": 2207
+    },
+    {
+      "epoch": 0.3301929116195603,
+      "grad_norm": 1.07026655557967,
+      "learning_rate": 7.817743614717218e-06,
+      "loss": 0.2192,
+      "step": 2208
+    },
+    {
+      "epoch": 0.3303424555106924,
+      "grad_norm": 1.5961930831288902,
+      "learning_rate": 7.815742655413651e-06,
+      "loss": 0.3224,
+      "step": 2209
+    },
+    {
+      "epoch": 0.33049199940182444,
+      "grad_norm": 1.4925568208469404,
+      "learning_rate": 7.813741035509718e-06,
+      "loss": 0.4544,
+      "step": 2210
+    },
+    {
+      "epoch": 0.3306415432929565,
+      "grad_norm": 1.0278606288492513,
+      "learning_rate": 7.811738755475024e-06,
+      "loss": 0.1586,
+      "step": 2211
+    },
+    {
+      "epoch": 0.3307910871840885,
+      "grad_norm": 1.6801015432287125,
+      "learning_rate": 7.80973581577932e-06,
+      "loss": 0.3744,
+      "step": 2212
+    },
+    {
+      "epoch": 0.3309406310752206,
+      "grad_norm": 1.7419776704612941,
+      "learning_rate": 7.807732216892514e-06,
+      "loss": 0.5292,
+      "step": 2213
+    },
+    {
+      "epoch": 0.3310901749663526,
+      "grad_norm": 1.2090204199623609,
+      "learning_rate": 7.80572795928467e-06,
+      "loss": 0.3566,
+      "step": 2214
+    },
+    {
+      "epoch": 0.33123971885748466,
+      "grad_norm": 1.0659334336267288,
+      "learning_rate": 7.803723043426008e-06,
+      "loss": 0.1775,
+      "step": 2215
+    },
+    {
+      "epoch": 0.33138926274861674,
+      "grad_norm": 1.2471584103742004,
+      "learning_rate": 7.8017174697869e-06,
+      "loss": 0.2591,
+      "step": 2216
+    },
+    {
+      "epoch": 0.3315388066397488,
+      "grad_norm": 1.2868762000450256,
+      "learning_rate": 7.799711238837871e-06,
+      "loss": 0.2203,
+      "step": 2217
+    },
+    {
+      "epoch": 0.3316883505308808,
+      "grad_norm": 1.597988709645616,
+      "learning_rate": 7.797704351049604e-06,
+      "loss": 0.4692,
+      "step": 2218
+    },
+    {
+      "epoch": 0.33183789442201284,
+      "grad_norm": 1.6497915275322537,
+      "learning_rate": 7.795696806892936e-06,
+      "loss": 0.5004,
+      "step": 2219
+    },
+    {
+      "epoch": 0.3319874383131449,
+      "grad_norm": 1.3728703674339666,
+      "learning_rate": 7.793688606838852e-06,
+      "loss": 0.2219,
+      "step": 2220
+    },
+    {
+      "epoch": 0.33213698220427695,
+      "grad_norm": 2.095794714839839,
+      "learning_rate": 7.791679751358497e-06,
+      "loss": 0.3322,
+      "step": 2221
+    },
+    {
+      "epoch": 0.332286526095409,
+      "grad_norm": 1.5144434482126574,
+      "learning_rate": 7.789670240923169e-06,
+      "loss": 0.3003,
+      "step": 2222
+    },
+    {
+      "epoch": 0.33243606998654107,
+      "grad_norm": 1.5220974726280843,
+      "learning_rate": 7.787660076004316e-06,
+      "loss": 0.3285,
+      "step": 2223
+    },
+    {
+      "epoch": 0.3325856138776731,
+      "grad_norm": 1.440338460328666,
+      "learning_rate": 7.785649257073544e-06,
+      "loss": 0.2101,
+      "step": 2224
+    },
+    {
+      "epoch": 0.33273515776880513,
+      "grad_norm": 1.5213659837860871,
+      "learning_rate": 7.783637784602608e-06,
+      "loss": 0.2036,
+      "step": 2225
+    },
+    {
+      "epoch": 0.33288470165993717,
+      "grad_norm": 2.991037147646805,
+      "learning_rate": 7.781625659063423e-06,
+      "loss": 0.6472,
+      "step": 2226
+    },
+    {
+      "epoch": 0.33303424555106925,
+      "grad_norm": 1.511479617581753,
+      "learning_rate": 7.779612880928052e-06,
+      "loss": 0.1741,
+      "step": 2227
+    },
+    {
+      "epoch": 0.3331837894422013,
+      "grad_norm": 1.4094694160926888,
+      "learning_rate": 7.777599450668708e-06,
+      "loss": 0.1864,
+      "step": 2228
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 1.9589341253223518,
+      "learning_rate": 7.775585368757767e-06,
+      "loss": 0.2585,
+      "step": 2229
+    },
+    {
+      "epoch": 0.3334828772244654,
+      "grad_norm": 1.5850528260281727,
+      "learning_rate": 7.773570635667746e-06,
+      "loss": 0.3359,
+      "step": 2230
+    },
+    {
+      "epoch": 0.33363242111559743,
+      "grad_norm": 1.928677128362692,
+      "learning_rate": 7.771555251871326e-06,
+      "loss": 0.5367,
+      "step": 2231
+    },
+    {
+      "epoch": 0.33378196500672946,
+      "grad_norm": 1.8147255946503564,
+      "learning_rate": 7.769539217841333e-06,
+      "loss": 0.2452,
+      "step": 2232
+    },
+    {
+      "epoch": 0.3339315088978615,
+      "grad_norm": 1.533478905239483,
+      "learning_rate": 7.76752253405075e-06,
+      "loss": 0.3273,
+      "step": 2233
+    },
+    {
+      "epoch": 0.3340810527889936,
+      "grad_norm": 1.3049158892916473,
+      "learning_rate": 7.765505200972708e-06,
+      "loss": 0.2014,
+      "step": 2234
+    },
+    {
+      "epoch": 0.3342305966801256,
+      "grad_norm": 1.6006461196766226,
+      "learning_rate": 7.763487219080492e-06,
+      "loss": 0.3297,
+      "step": 2235
+    },
+    {
+      "epoch": 0.33438014057125764,
+      "grad_norm": 1.3959531694066092,
+      "learning_rate": 7.761468588847543e-06,
+      "loss": 0.1987,
+      "step": 2236
+    },
+    {
+      "epoch": 0.33452968446238973,
+      "grad_norm": 1.413036480020334,
+      "learning_rate": 7.75944931074745e-06,
+      "loss": 0.2507,
+      "step": 2237
+    },
+    {
+      "epoch": 0.33467922835352176,
+      "grad_norm": 1.6143147896405532,
+      "learning_rate": 7.757429385253955e-06,
+      "loss": 0.365,
+      "step": 2238
+    },
+    {
+      "epoch": 0.3348287722446538,
+      "grad_norm": 1.2552402165493568,
+      "learning_rate": 7.755408812840952e-06,
+      "loss": 0.2404,
+      "step": 2239
+    },
+    {
+      "epoch": 0.3349783161357859,
+      "grad_norm": 1.884570594935436,
+      "learning_rate": 7.753387593982485e-06,
+      "loss": 0.4998,
+      "step": 2240
+    },
+    {
+      "epoch": 0.3351278600269179,
+      "grad_norm": 1.5525626441469005,
+      "learning_rate": 7.751365729152754e-06,
+      "loss": 0.1856,
+      "step": 2241
+    },
+    {
+      "epoch": 0.33527740391804994,
+      "grad_norm": 1.5710541861971588,
+      "learning_rate": 7.749343218826107e-06,
+      "loss": 0.1989,
+      "step": 2242
+    },
+    {
+      "epoch": 0.335426947809182,
+      "grad_norm": 1.9778222905539145,
+      "learning_rate": 7.747320063477045e-06,
+      "loss": 0.602,
+      "step": 2243
+    },
+    {
+      "epoch": 0.33557649170031406,
+      "grad_norm": 1.6193619290401855,
+      "learning_rate": 7.745296263580218e-06,
+      "loss": 0.2811,
+      "step": 2244
+    },
+    {
+      "epoch": 0.3357260355914461,
+      "grad_norm": 1.9654135361350595,
+      "learning_rate": 7.743271819610432e-06,
+      "loss": 0.5169,
+      "step": 2245
+    },
+    {
+      "epoch": 0.3358755794825781,
+      "grad_norm": 1.5452430622711606,
+      "learning_rate": 7.741246732042638e-06,
+      "loss": 0.345,
+      "step": 2246
+    },
+    {
+      "epoch": 0.3360251233737102,
+      "grad_norm": 1.7587162322576044,
+      "learning_rate": 7.739221001351942e-06,
+      "loss": 0.4502,
+      "step": 2247
+    },
+    {
+      "epoch": 0.33617466726484224,
+      "grad_norm": 1.9373589145144738,
+      "learning_rate": 7.7371946280136e-06,
+      "loss": 0.3852,
+      "step": 2248
+    },
+    {
+      "epoch": 0.3363242111559743,
+      "grad_norm": 1.9597929999948922,
+      "learning_rate": 7.73516761250302e-06,
+      "loss": 0.3904,
+      "step": 2249
+    },
+    {
+      "epoch": 0.3364737550471063,
+      "grad_norm": 1.8798765980010916,
+      "learning_rate": 7.733139955295756e-06,
+      "loss": 0.5182,
+      "step": 2250
+    },
+    {
+      "epoch": 0.3366232989382384,
+      "grad_norm": 1.626082298346251,
+      "learning_rate": 7.73111165686752e-06,
+      "loss": 0.3692,
+      "step": 2251
+    },
+    {
+      "epoch": 0.3367728428293704,
+      "grad_norm": 0.9953658155537086,
+      "learning_rate": 7.72908271769417e-06,
+      "loss": 0.1844,
+      "step": 2252
+    },
+    {
+      "epoch": 0.33692238672050245,
+      "grad_norm": 1.3727196538195634,
+      "learning_rate": 7.727053138251712e-06,
+      "loss": 0.2179,
+      "step": 2253
+    },
+    {
+      "epoch": 0.33707193061163454,
+      "grad_norm": 1.4912804935224222,
+      "learning_rate": 7.725022919016306e-06,
+      "loss": 0.3038,
+      "step": 2254
+    },
+    {
+      "epoch": 0.33722147450276657,
+      "grad_norm": 1.5520829591707979,
+      "learning_rate": 7.722992060464261e-06,
+      "loss": 0.2481,
+      "step": 2255
+    },
+    {
+      "epoch": 0.3373710183938986,
+      "grad_norm": 0.9325160815572334,
+      "learning_rate": 7.720960563072035e-06,
+      "loss": 0.1387,
+      "step": 2256
+    },
+    {
+      "epoch": 0.33752056228503063,
+      "grad_norm": 1.1905265030205905,
+      "learning_rate": 7.718928427316241e-06,
+      "loss": 0.2595,
+      "step": 2257
+    },
+    {
+      "epoch": 0.3376701061761627,
+      "grad_norm": 1.0242402171948481,
+      "learning_rate": 7.716895653673633e-06,
+      "loss": 0.1728,
+      "step": 2258
+    },
+    {
+      "epoch": 0.33781965006729475,
+      "grad_norm": 1.9399325651192516,
+      "learning_rate": 7.714862242621121e-06,
+      "loss": 0.236,
+      "step": 2259
+    },
+    {
+      "epoch": 0.3379691939584268,
+      "grad_norm": 1.2589445942870727,
+      "learning_rate": 7.712828194635762e-06,
+      "loss": 0.1777,
+      "step": 2260
+    },
+    {
+      "epoch": 0.33811873784955887,
+      "grad_norm": 1.0189796376895088,
+      "learning_rate": 7.710793510194765e-06,
+      "loss": 0.2061,
+      "step": 2261
+    },
+    {
+      "epoch": 0.3382682817406909,
+      "grad_norm": 1.3110462323223497,
+      "learning_rate": 7.708758189775485e-06,
+      "loss": 0.1907,
+      "step": 2262
+    },
+    {
+      "epoch": 0.33841782563182293,
+      "grad_norm": 1.8977870965152843,
+      "learning_rate": 7.706722233855428e-06,
+      "loss": 0.4697,
+      "step": 2263
+    },
+    {
+      "epoch": 0.33856736952295496,
+      "grad_norm": 1.676037286436981,
+      "learning_rate": 7.70468564291225e-06,
+      "loss": 0.5385,
+      "step": 2264
+    },
+    {
+      "epoch": 0.33871691341408705,
+      "grad_norm": 1.8926850673510127,
+      "learning_rate": 7.702648417423755e-06,
+      "loss": 0.6545,
+      "step": 2265
+    },
+    {
+      "epoch": 0.3388664573052191,
+      "grad_norm": 1.211799579288037,
+      "learning_rate": 7.700610557867894e-06,
+      "loss": 0.1933,
+      "step": 2266
+    },
+    {
+      "epoch": 0.3390160011963511,
+      "grad_norm": 1.3249080152566644,
+      "learning_rate": 7.69857206472277e-06,
+      "loss": 0.3369,
+      "step": 2267
+    },
+    {
+      "epoch": 0.3391655450874832,
+      "grad_norm": 1.197942464661273,
+      "learning_rate": 7.696532938466631e-06,
+      "loss": 0.2168,
+      "step": 2268
+    },
+    {
+      "epoch": 0.33931508897861523,
+      "grad_norm": 1.5336422309655264,
+      "learning_rate": 7.69449317957788e-06,
+      "loss": 0.1862,
+      "step": 2269
+    },
+    {
+      "epoch": 0.33946463286974726,
+      "grad_norm": 1.4755775481717086,
+      "learning_rate": 7.692452788535058e-06,
+      "loss": 0.3991,
+      "step": 2270
+    },
+    {
+      "epoch": 0.3396141767608793,
+      "grad_norm": 1.2003981851028729,
+      "learning_rate": 7.690411765816864e-06,
+      "loss": 0.1887,
+      "step": 2271
+    },
+    {
+      "epoch": 0.3397637206520114,
+      "grad_norm": 1.5232479955721698,
+      "learning_rate": 7.688370111902141e-06,
+      "loss": 0.3291,
+      "step": 2272
+    },
+    {
+      "epoch": 0.3399132645431434,
+      "grad_norm": 1.1630358360895665,
+      "learning_rate": 7.686327827269883e-06,
+      "loss": 0.3571,
+      "step": 2273
+    },
+    {
+      "epoch": 0.34006280843427544,
+      "grad_norm": 1.2180304152897419,
+      "learning_rate": 7.684284912399227e-06,
+      "loss": 0.2026,
+      "step": 2274
+    },
+    {
+      "epoch": 0.34021235232540753,
+      "grad_norm": 1.2838128091625958,
+      "learning_rate": 7.68224136776946e-06,
+      "loss": 0.1883,
+      "step": 2275
+    },
+    {
+      "epoch": 0.34036189621653956,
+      "grad_norm": 1.9043193483249239,
+      "learning_rate": 7.680197193860019e-06,
+      "loss": 0.5335,
+      "step": 2276
+    },
+    {
+      "epoch": 0.3405114401076716,
+      "grad_norm": 1.6719584667012997,
+      "learning_rate": 7.678152391150488e-06,
+      "loss": 0.4823,
+      "step": 2277
+    },
+    {
+      "epoch": 0.3406609839988036,
+      "grad_norm": 1.614799856208538,
+      "learning_rate": 7.676106960120595e-06,
+      "loss": 0.3589,
+      "step": 2278
+    },
+    {
+      "epoch": 0.3408105278899357,
+      "grad_norm": 1.915461278238496,
+      "learning_rate": 7.674060901250217e-06,
+      "loss": 0.4702,
+      "step": 2279
+    },
+    {
+      "epoch": 0.34096007178106774,
+      "grad_norm": 1.80642832565446,
+      "learning_rate": 7.672014215019382e-06,
+      "loss": 0.5783,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3411096156721998,
+      "grad_norm": 1.8185018530729478,
+      "learning_rate": 7.66996690190826e-06,
+      "loss": 0.2136,
+      "step": 2281
+    },
+    {
+      "epoch": 0.34125915956333186,
+      "grad_norm": 1.5750847935557877,
+      "learning_rate": 7.667918962397172e-06,
+      "loss": 0.3577,
+      "step": 2282
+    },
+    {
+      "epoch": 0.3414087034544639,
+      "grad_norm": 1.4052021520161486,
+      "learning_rate": 7.665870396966582e-06,
+      "loss": 0.3421,
+      "step": 2283
+    },
+    {
+      "epoch": 0.3415582473455959,
+      "grad_norm": 2.4796312435140964,
+      "learning_rate": 7.663821206097106e-06,
+      "loss": 0.5199,
+      "step": 2284
+    },
+    {
+      "epoch": 0.34170779123672795,
+      "grad_norm": 1.9135083427877795,
+      "learning_rate": 7.661771390269506e-06,
+      "loss": 0.5724,
+      "step": 2285
+    },
+    {
+      "epoch": 0.34185733512786004,
+      "grad_norm": 1.47040509500929,
+      "learning_rate": 7.65972094996468e-06,
+      "loss": 0.3446,
+      "step": 2286
+    },
+    {
+      "epoch": 0.34200687901899207,
+      "grad_norm": 1.8826362598607522,
+      "learning_rate": 7.65766988566369e-06,
+      "loss": 0.4584,
+      "step": 2287
+    },
+    {
+      "epoch": 0.3421564229101241,
+      "grad_norm": 1.583414431479772,
+      "learning_rate": 7.65561819784773e-06,
+      "loss": 0.2571,
+      "step": 2288
+    },
+    {
+      "epoch": 0.3423059668012562,
+      "grad_norm": 1.1374097342884393,
+      "learning_rate": 7.653565886998149e-06,
+      "loss": 0.2772,
+      "step": 2289
+    },
+    {
+      "epoch": 0.3424555106923882,
+      "grad_norm": 1.6018927652316786,
+      "learning_rate": 7.651512953596438e-06,
+      "loss": 0.3834,
+      "step": 2290
+    },
+    {
+      "epoch": 0.34260505458352025,
+      "grad_norm": 1.5898030999912747,
+      "learning_rate": 7.649459398124233e-06,
+      "loss": 0.268,
+      "step": 2291
+    },
+    {
+      "epoch": 0.3427545984746523,
+      "grad_norm": 1.788880984497509,
+      "learning_rate": 7.64740522106332e-06,
+      "loss": 0.4031,
+      "step": 2292
+    },
+    {
+      "epoch": 0.34290414236578437,
+      "grad_norm": 1.4103716345007062,
+      "learning_rate": 7.645350422895627e-06,
+      "loss": 0.34,
+      "step": 2293
+    },
+    {
+      "epoch": 0.3430536862569164,
+      "grad_norm": 1.455320179410942,
+      "learning_rate": 7.643295004103232e-06,
+      "loss": 0.393,
+      "step": 2294
+    },
+    {
+      "epoch": 0.34320323014804843,
+      "grad_norm": 1.121798327145041,
+      "learning_rate": 7.641238965168356e-06,
+      "loss": 0.1951,
+      "step": 2295
+    },
+    {
+      "epoch": 0.3433527740391805,
+      "grad_norm": 1.959743352728859,
+      "learning_rate": 7.639182306573362e-06,
+      "loss": 0.5306,
+      "step": 2296
+    },
+    {
+      "epoch": 0.34350231793031255,
+      "grad_norm": 1.0222053736742456,
+      "learning_rate": 7.637125028800765e-06,
+      "loss": 0.1877,
+      "step": 2297
+    },
+    {
+      "epoch": 0.3436518618214446,
+      "grad_norm": 1.2126683631983057,
+      "learning_rate": 7.63506713233322e-06,
+      "loss": 0.2219,
+      "step": 2298
+    },
+    {
+      "epoch": 0.34380140571257667,
+      "grad_norm": 1.182330026578451,
+      "learning_rate": 7.633008617653531e-06,
+      "loss": 0.2427,
+      "step": 2299
+    },
+    {
+      "epoch": 0.3439509496037087,
+      "grad_norm": 1.3284239825142252,
+      "learning_rate": 7.630949485244646e-06,
+      "loss": 0.3342,
+      "step": 2300
+    },
+    {
+      "epoch": 0.34410049349484073,
+      "grad_norm": 1.4991905889604522,
+      "learning_rate": 7.6288897355896565e-06,
+      "loss": 0.3277,
+      "step": 2301
+    },
+    {
+      "epoch": 0.34425003738597276,
+      "grad_norm": 1.2105004198602645,
+      "learning_rate": 7.6268293691718e-06,
+      "loss": 0.2175,
+      "step": 2302
+    },
+    {
+      "epoch": 0.34439958127710485,
+      "grad_norm": 1.137201185526714,
+      "learning_rate": 7.624768386474456e-06,
+      "loss": 0.1951,
+      "step": 2303
+    },
+    {
+      "epoch": 0.3445491251682369,
+      "grad_norm": 1.2921483182674967,
+      "learning_rate": 7.622706787981153e-06,
+      "loss": 0.2069,
+      "step": 2304
+    },
+    {
+      "epoch": 0.3446986690593689,
+      "grad_norm": 1.5385416209630458,
+      "learning_rate": 7.620644574175562e-06,
+      "loss": 0.3467,
+      "step": 2305
+    },
+    {
+      "epoch": 0.344848212950501,
+      "grad_norm": 2.272200349391826,
+      "learning_rate": 7.6185817455414975e-06,
+      "loss": 0.4775,
+      "step": 2306
+    },
+    {
+      "epoch": 0.34499775684163303,
+      "grad_norm": 1.7594806067750843,
+      "learning_rate": 7.61651830256292e-06,
+      "loss": 0.2168,
+      "step": 2307
+    },
+    {
+      "epoch": 0.34514730073276506,
+      "grad_norm": 1.9724043128153095,
+      "learning_rate": 7.614454245723932e-06,
+      "loss": 0.5726,
+      "step": 2308
+    },
+    {
+      "epoch": 0.3452968446238971,
+      "grad_norm": 1.4742051107263883,
+      "learning_rate": 7.612389575508781e-06,
+      "loss": 0.3032,
+      "step": 2309
+    },
+    {
+      "epoch": 0.3454463885150292,
+      "grad_norm": 2.0620637201052325,
+      "learning_rate": 7.610324292401861e-06,
+      "loss": 0.309,
+      "step": 2310
+    },
+    {
+      "epoch": 0.3455959324061612,
+      "grad_norm": 1.0544585597609104,
+      "learning_rate": 7.608258396887702e-06,
+      "loss": 0.2339,
+      "step": 2311
+    },
+    {
+      "epoch": 0.34574547629729324,
+      "grad_norm": 1.2388411799285035,
+      "learning_rate": 7.606191889450989e-06,
+      "loss": 0.1824,
+      "step": 2312
+    },
+    {
+      "epoch": 0.34589502018842533,
+      "grad_norm": 1.9886690672129868,
+      "learning_rate": 7.604124770576539e-06,
+      "loss": 0.3321,
+      "step": 2313
+    },
+    {
+      "epoch": 0.34604456407955736,
+      "grad_norm": 1.3774072344684907,
+      "learning_rate": 7.602057040749325e-06,
+      "loss": 0.3824,
+      "step": 2314
+    },
+    {
+      "epoch": 0.3461941079706894,
+      "grad_norm": 0.842321998785562,
+      "learning_rate": 7.59998870045445e-06,
+      "loss": 0.1966,
+      "step": 2315
+    },
+    {
+      "epoch": 0.3463436518618214,
+      "grad_norm": 1.6133223844738274,
+      "learning_rate": 7.597919750177168e-06,
+      "loss": 0.3439,
+      "step": 2316
+    },
+    {
+      "epoch": 0.3464931957529535,
+      "grad_norm": 2.701793044817531,
+      "learning_rate": 7.595850190402877e-06,
+      "loss": 0.3649,
+      "step": 2317
+    },
+    {
+      "epoch": 0.34664273964408554,
+      "grad_norm": 2.042826281041951,
+      "learning_rate": 7.593780021617115e-06,
+      "loss": 0.5945,
+      "step": 2318
+    },
+    {
+      "epoch": 0.34679228353521757,
+      "grad_norm": 1.8226181825863899,
+      "learning_rate": 7.591709244305561e-06,
+      "loss": 0.4039,
+      "step": 2319
+    },
+    {
+      "epoch": 0.34694182742634966,
+      "grad_norm": 1.7612529231819496,
+      "learning_rate": 7.589637858954041e-06,
+      "loss": 0.2384,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3470913713174817,
+      "grad_norm": 1.2495900823088222,
+      "learning_rate": 7.587565866048523e-06,
+      "loss": 0.3633,
+      "step": 2321
+    },
+    {
+      "epoch": 0.3472409152086137,
+      "grad_norm": 1.2456808766471323,
+      "learning_rate": 7.5854932660751144e-06,
+      "loss": 0.2318,
+      "step": 2322
+    },
+    {
+      "epoch": 0.34739045909974575,
+      "grad_norm": 1.772092412043284,
+      "learning_rate": 7.58342005952007e-06,
+      "loss": 0.2514,
+      "step": 2323
+    },
+    {
+      "epoch": 0.34754000299087784,
+      "grad_norm": 1.0718322965745777,
+      "learning_rate": 7.581346246869781e-06,
+      "loss": 0.2001,
+      "step": 2324
+    },
+    {
+      "epoch": 0.34768954688200987,
+      "grad_norm": 1.9749261015333692,
+      "learning_rate": 7.579271828610786e-06,
+      "loss": 0.642,
+      "step": 2325
+    },
+    {
+      "epoch": 0.3478390907731419,
+      "grad_norm": 3.5434908077443183,
+      "learning_rate": 7.5771968052297605e-06,
+      "loss": 0.2132,
+      "step": 2326
+    },
+    {
+      "epoch": 0.347988634664274,
+      "grad_norm": 1.3138111751860577,
+      "learning_rate": 7.575121177213528e-06,
+      "loss": 0.2074,
+      "step": 2327
+    },
+    {
+      "epoch": 0.348138178555406,
+      "grad_norm": 1.4549216960822273,
+      "learning_rate": 7.573044945049051e-06,
+      "loss": 0.2213,
+      "step": 2328
+    },
+    {
+      "epoch": 0.34828772244653805,
+      "grad_norm": 1.9285857186532291,
+      "learning_rate": 7.5709681092234315e-06,
+      "loss": 0.5573,
+      "step": 2329
+    },
+    {
+      "epoch": 0.3484372663376701,
+      "grad_norm": 1.528743992296175,
+      "learning_rate": 7.568890670223918e-06,
+      "loss": 0.2517,
+      "step": 2330
+    },
+    {
+      "epoch": 0.34858681022880217,
+      "grad_norm": 1.3064713646436477,
+      "learning_rate": 7.566812628537894e-06,
+      "loss": 0.1867,
+      "step": 2331
+    },
+    {
+      "epoch": 0.3487363541199342,
+      "grad_norm": 1.3819819935958808,
+      "learning_rate": 7.56473398465289e-06,
+      "loss": 0.2354,
+      "step": 2332
+    },
+    {
+      "epoch": 0.34888589801106623,
+      "grad_norm": 1.90386469705261,
+      "learning_rate": 7.5626547390565766e-06,
+      "loss": 0.6091,
+      "step": 2333
+    },
+    {
+      "epoch": 0.3490354419021983,
+      "grad_norm": 1.1572160857164981,
+      "learning_rate": 7.5605748922367636e-06,
+      "loss": 0.1634,
+      "step": 2334
+    },
+    {
+      "epoch": 0.34918498579333035,
+      "grad_norm": 1.120922647106074,
+      "learning_rate": 7.558494444681405e-06,
+      "loss": 0.2191,
+      "step": 2335
+    },
+    {
+      "epoch": 0.3493345296844624,
+      "grad_norm": 1.8933761063913324,
+      "learning_rate": 7.556413396878593e-06,
+      "loss": 0.4644,
+      "step": 2336
+    },
+    {
+      "epoch": 0.3494840735755944,
+      "grad_norm": 2.2982217544810526,
+      "learning_rate": 7.554331749316559e-06,
+      "loss": 0.5312,
+      "step": 2337
+    },
+    {
+      "epoch": 0.3496336174667265,
+      "grad_norm": 1.4598497201666212,
+      "learning_rate": 7.552249502483681e-06,
+      "loss": 0.2322,
+      "step": 2338
+    },
+    {
+      "epoch": 0.34978316135785853,
+      "grad_norm": 1.0002482253315497,
+      "learning_rate": 7.550166656868472e-06,
+      "loss": 0.183,
+      "step": 2339
+    },
+    {
+      "epoch": 0.34993270524899056,
+      "grad_norm": 1.7931068633163847,
+      "learning_rate": 7.548083212959588e-06,
+      "loss": 0.3655,
+      "step": 2340
+    },
+    {
+      "epoch": 0.35008224914012265,
+      "grad_norm": 1.9088729651038523,
+      "learning_rate": 7.545999171245826e-06,
+      "loss": 0.5897,
+      "step": 2341
+    },
+    {
+      "epoch": 0.3502317930312547,
+      "grad_norm": 1.6014278355250144,
+      "learning_rate": 7.543914532216121e-06,
+      "loss": 0.3399,
+      "step": 2342
+    },
+    {
+      "epoch": 0.3503813369223867,
+      "grad_norm": 1.589397325565207,
+      "learning_rate": 7.541829296359552e-06,
+      "loss": 0.5276,
+      "step": 2343
+    },
+    {
+      "epoch": 0.35053088081351874,
+      "grad_norm": 1.3666145854556788,
+      "learning_rate": 7.5397434641653325e-06,
+      "loss": 0.297,
+      "step": 2344
+    },
+    {
+      "epoch": 0.35068042470465083,
+      "grad_norm": 1.1296429490766309,
+      "learning_rate": 7.53765703612282e-06,
+      "loss": 0.1881,
+      "step": 2345
+    },
+    {
+      "epoch": 0.35082996859578286,
+      "grad_norm": 1.3370573999376023,
+      "learning_rate": 7.535570012721509e-06,
+      "loss": 0.1978,
+      "step": 2346
+    },
+    {
+      "epoch": 0.3509795124869149,
+      "grad_norm": 1.80559813243436,
+      "learning_rate": 7.533482394451037e-06,
+      "loss": 0.5293,
+      "step": 2347
+    },
+    {
+      "epoch": 0.351129056378047,
+      "grad_norm": 1.2601205754536828,
+      "learning_rate": 7.531394181801182e-06,
+      "loss": 0.3029,
+      "step": 2348
+    },
+    {
+      "epoch": 0.351278600269179,
+      "grad_norm": 1.5151339748208972,
+      "learning_rate": 7.529305375261852e-06,
+      "loss": 0.354,
+      "step": 2349
+    },
+    {
+      "epoch": 0.35142814416031104,
+      "grad_norm": 1.1154348924689401,
+      "learning_rate": 7.52721597532311e-06,
+      "loss": 0.1877,
+      "step": 2350
+    },
+    {
+      "epoch": 0.35157768805144307,
+      "grad_norm": 1.6903766996209675,
+      "learning_rate": 7.525125982475141e-06,
+      "loss": 0.3838,
+      "step": 2351
+    },
+    {
+      "epoch": 0.35172723194257516,
+      "grad_norm": 1.5111917986875527,
+      "learning_rate": 7.523035397208281e-06,
+      "loss": 0.3664,
+      "step": 2352
+    },
+    {
+      "epoch": 0.3518767758337072,
+      "grad_norm": 14.577070159201043,
+      "learning_rate": 7.520944220013002e-06,
+      "loss": 0.2204,
+      "step": 2353
+    },
+    {
+      "epoch": 0.3520263197248392,
+      "grad_norm": 1.2720960634080132,
+      "learning_rate": 7.518852451379914e-06,
+      "loss": 0.2156,
+      "step": 2354
+    },
+    {
+      "epoch": 0.3521758636159713,
+      "grad_norm": 1.4752321481872246,
+      "learning_rate": 7.516760091799766e-06,
+      "loss": 0.2068,
+      "step": 2355
+    },
+    {
+      "epoch": 0.35232540750710334,
+      "grad_norm": 1.2757349060760932,
+      "learning_rate": 7.5146671417634456e-06,
+      "loss": 0.3031,
+      "step": 2356
+    },
+    {
+      "epoch": 0.35247495139823537,
+      "grad_norm": 1.1892854387567553,
+      "learning_rate": 7.512573601761979e-06,
+      "loss": 0.1938,
+      "step": 2357
+    },
+    {
+      "epoch": 0.35262449528936746,
+      "grad_norm": 1.1695321760256765,
+      "learning_rate": 7.5104794722865305e-06,
+      "loss": 0.1994,
+      "step": 2358
+    },
+    {
+      "epoch": 0.3527740391804995,
+      "grad_norm": 1.4157720422061164,
+      "learning_rate": 7.5083847538284025e-06,
+      "loss": 0.2415,
+      "step": 2359
+    },
+    {
+      "epoch": 0.3529235830716315,
+      "grad_norm": 1.642650861184269,
+      "learning_rate": 7.506289446879038e-06,
+      "loss": 0.3864,
+      "step": 2360
+    },
+    {
+      "epoch": 0.35307312696276355,
+      "grad_norm": 1.66074120706464,
+      "learning_rate": 7.504193551930014e-06,
+      "loss": 0.2436,
+      "step": 2361
+    },
+    {
+      "epoch": 0.35322267085389564,
+      "grad_norm": 1.6293105897438254,
+      "learning_rate": 7.502097069473051e-06,
+      "loss": 0.4727,
+      "step": 2362
+    },
+    {
+      "epoch": 0.35337221474502767,
+      "grad_norm": 2.1484736491543326,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.6399,
+      "step": 2363
+    },
+    {
+      "epoch": 0.3535217586361597,
+      "grad_norm": 1.998088679617451,
+      "learning_rate": 7.497902344002858e-06,
+      "loss": 0.4879,
+      "step": 2364
+    },
+    {
+      "epoch": 0.3536713025272918,
+      "grad_norm": 1.0466569132461796,
+      "learning_rate": 7.495804101973751e-06,
+      "loss": 0.175,
+      "step": 2365
+    },
+    {
+      "epoch": 0.3538208464184238,
+      "grad_norm": 1.483423152464052,
+      "learning_rate": 7.49370527440495e-06,
+      "loss": 0.1817,
+      "step": 2366
+    },
+    {
+      "epoch": 0.35397039030955585,
+      "grad_norm": 1.181675839490534,
+      "learning_rate": 7.491605861788856e-06,
+      "loss": 0.2158,
+      "step": 2367
+    },
+    {
+      "epoch": 0.3541199342006879,
+      "grad_norm": 1.272613662606653,
+      "learning_rate": 7.4895058646180165e-06,
+      "loss": 0.2161,
+      "step": 2368
+    },
+    {
+      "epoch": 0.35426947809181997,
+      "grad_norm": 2.249706812141276,
+      "learning_rate": 7.487405283385109e-06,
+      "loss": 0.4302,
+      "step": 2369
+    },
+    {
+      "epoch": 0.354419021982952,
+      "grad_norm": 1.790056947576118,
+      "learning_rate": 7.485304118582949e-06,
+      "loss": 0.4237,
+      "step": 2370
+    },
+    {
+      "epoch": 0.35456856587408403,
+      "grad_norm": 1.728469311368499,
+      "learning_rate": 7.483202370704492e-06,
+      "loss": 0.232,
+      "step": 2371
+    },
+    {
+      "epoch": 0.3547181097652161,
+      "grad_norm": 1.7204599933992115,
+      "learning_rate": 7.481100040242827e-06,
+      "loss": 0.2945,
+      "step": 2372
+    },
+    {
+      "epoch": 0.35486765365634815,
+      "grad_norm": 1.7017419956837787,
+      "learning_rate": 7.478997127691181e-06,
+      "loss": 0.324,
+      "step": 2373
+    },
+    {
+      "epoch": 0.3550171975474802,
+      "grad_norm": 1.7242409041398354,
+      "learning_rate": 7.476893633542917e-06,
+      "loss": 0.379,
+      "step": 2374
+    },
+    {
+      "epoch": 0.3551667414386122,
+      "grad_norm": 1.7352420641389674,
+      "learning_rate": 7.474789558291537e-06,
+      "loss": 0.3666,
+      "step": 2375
+    },
+    {
+      "epoch": 0.3553162853297443,
+      "grad_norm": 2.417484284214745,
+      "learning_rate": 7.472684902430678e-06,
+      "loss": 0.5313,
+      "step": 2376
+    },
+    {
+      "epoch": 0.35546582922087633,
+      "grad_norm": 1.9863314858317191,
+      "learning_rate": 7.470579666454108e-06,
+      "loss": 0.4731,
+      "step": 2377
+    },
+    {
+      "epoch": 0.35561537311200836,
+      "grad_norm": 1.3767673556499953,
+      "learning_rate": 7.46847385085574e-06,
+      "loss": 0.2309,
+      "step": 2378
+    },
+    {
+      "epoch": 0.35576491700314045,
+      "grad_norm": 2.1027176103910423,
+      "learning_rate": 7.466367456129616e-06,
+      "loss": 0.2419,
+      "step": 2379
+    },
+    {
+      "epoch": 0.3559144608942725,
+      "grad_norm": 1.3223085016147877,
+      "learning_rate": 7.464260482769917e-06,
+      "loss": 0.2214,
+      "step": 2380
+    },
+    {
+      "epoch": 0.3560640047854045,
+      "grad_norm": 3.308046838444924,
+      "learning_rate": 7.462152931270961e-06,
+      "loss": 0.1637,
+      "step": 2381
+    },
+    {
+      "epoch": 0.35621354867653654,
+      "grad_norm": 1.2529639271363728,
+      "learning_rate": 7.4600448021271975e-06,
+      "loss": 0.209,
+      "step": 2382
+    },
+    {
+      "epoch": 0.3563630925676686,
+      "grad_norm": 1.4685959227889136,
+      "learning_rate": 7.457936095833216e-06,
+      "loss": 0.3916,
+      "step": 2383
+    },
+    {
+      "epoch": 0.35651263645880066,
+      "grad_norm": 1.762730610248373,
+      "learning_rate": 7.455826812883738e-06,
+      "loss": 0.2468,
+      "step": 2384
+    },
+    {
+      "epoch": 0.3566621803499327,
+      "grad_norm": 1.3974489415451616,
+      "learning_rate": 7.453716953773622e-06,
+      "loss": 0.2198,
+      "step": 2385
+    },
+    {
+      "epoch": 0.3568117242410648,
+      "grad_norm": 1.6020912414110884,
+      "learning_rate": 7.4516065189978625e-06,
+      "loss": 0.3663,
+      "step": 2386
+    },
+    {
+      "epoch": 0.3569612681321968,
+      "grad_norm": 1.1698502771825479,
+      "learning_rate": 7.449495509051584e-06,
+      "loss": 0.2301,
+      "step": 2387
+    },
+    {
+      "epoch": 0.35711081202332884,
+      "grad_norm": 1.6443713783244578,
+      "learning_rate": 7.447383924430055e-06,
+      "loss": 0.2023,
+      "step": 2388
+    },
+    {
+      "epoch": 0.35726035591446087,
+      "grad_norm": 1.320656374724735,
+      "learning_rate": 7.44527176562867e-06,
+      "loss": 0.1615,
+      "step": 2389
+    },
+    {
+      "epoch": 0.35740989980559296,
+      "grad_norm": 2.147900342981391,
+      "learning_rate": 7.4431590331429615e-06,
+      "loss": 0.2496,
+      "step": 2390
+    },
+    {
+      "epoch": 0.357559443696725,
+      "grad_norm": 1.0493224753110673,
+      "learning_rate": 7.441045727468601e-06,
+      "loss": 0.174,
+      "step": 2391
+    },
+    {
+      "epoch": 0.357708987587857,
+      "grad_norm": 1.4365731625537776,
+      "learning_rate": 7.4389318491013855e-06,
+      "loss": 0.2509,
+      "step": 2392
+    },
+    {
+      "epoch": 0.3578585314789891,
+      "grad_norm": 1.3570392728095118,
+      "learning_rate": 7.436817398537253e-06,
+      "loss": 0.3505,
+      "step": 2393
+    },
+    {
+      "epoch": 0.35800807537012114,
+      "grad_norm": 0.9836140143478714,
+      "learning_rate": 7.434702376272275e-06,
+      "loss": 0.1676,
+      "step": 2394
+    },
+    {
+      "epoch": 0.35815761926125317,
+      "grad_norm": 1.6386886795873667,
+      "learning_rate": 7.4325867828026555e-06,
+      "loss": 0.3311,
+      "step": 2395
+    },
+    {
+      "epoch": 0.3583071631523852,
+      "grad_norm": 1.7664542023376915,
+      "learning_rate": 7.4304706186247344e-06,
+      "loss": 0.4952,
+      "step": 2396
+    },
+    {
+      "epoch": 0.3584567070435173,
+      "grad_norm": 1.970974898714591,
+      "learning_rate": 7.42835388423498e-06,
+      "loss": 0.5721,
+      "step": 2397
+    },
+    {
+      "epoch": 0.3586062509346493,
+      "grad_norm": 1.234101790360312,
+      "learning_rate": 7.426236580130004e-06,
+      "loss": 0.1842,
+      "step": 2398
+    },
+    {
+      "epoch": 0.35875579482578135,
+      "grad_norm": 0.8359881523865066,
+      "learning_rate": 7.424118706806543e-06,
+      "loss": 0.1574,
+      "step": 2399
+    },
+    {
+      "epoch": 0.35890533871691344,
+      "grad_norm": 1.2167515845771457,
+      "learning_rate": 7.422000264761471e-06,
+      "loss": 0.3393,
+      "step": 2400
+    },
+    {
+      "epoch": 0.35905488260804547,
+      "grad_norm": 1.1652590311365825,
+      "learning_rate": 7.419881254491794e-06,
+      "loss": 0.3154,
+      "step": 2401
+    },
+    {
+      "epoch": 0.3592044264991775,
+      "grad_norm": 1.7451208164751588,
+      "learning_rate": 7.417761676494654e-06,
+      "loss": 0.3306,
+      "step": 2402
+    },
+    {
+      "epoch": 0.35935397039030953,
+      "grad_norm": 1.50819933578857,
+      "learning_rate": 7.415641531267325e-06,
+      "loss": 0.3414,
+      "step": 2403
+    },
+    {
+      "epoch": 0.3595035142814416,
+      "grad_norm": 1.159271590101521,
+      "learning_rate": 7.4135208193072126e-06,
+      "loss": 0.1953,
+      "step": 2404
+    },
+    {
+      "epoch": 0.35965305817257365,
+      "grad_norm": 1.6763078000161467,
+      "learning_rate": 7.411399541111855e-06,
+      "loss": 0.3306,
+      "step": 2405
+    },
+    {
+      "epoch": 0.3598026020637057,
+      "grad_norm": 1.6105048047144768,
+      "learning_rate": 7.409277697178926e-06,
+      "loss": 0.2899,
+      "step": 2406
+    },
+    {
+      "epoch": 0.35995214595483777,
+      "grad_norm": 1.6423579180660441,
+      "learning_rate": 7.4071552880062295e-06,
+      "loss": 0.1941,
+      "step": 2407
+    },
+    {
+      "epoch": 0.3601016898459698,
+      "grad_norm": 1.984850391003558,
+      "learning_rate": 7.4050323140917035e-06,
+      "loss": 0.509,
+      "step": 2408
+    },
+    {
+      "epoch": 0.36025123373710183,
+      "grad_norm": 1.6162723010996198,
+      "learning_rate": 7.402908775933419e-06,
+      "loss": 0.3583,
+      "step": 2409
+    },
+    {
+      "epoch": 0.36040077762823386,
+      "grad_norm": 1.3526627301579437,
+      "learning_rate": 7.400784674029579e-06,
+      "loss": 0.2228,
+      "step": 2410
+    },
+    {
+      "epoch": 0.36055032151936595,
+      "grad_norm": 1.0613997053031348,
+      "learning_rate": 7.398660008878517e-06,
+      "loss": 0.1782,
+      "step": 2411
+    },
+    {
+      "epoch": 0.360699865410498,
+      "grad_norm": 1.4156014777328207,
+      "learning_rate": 7.396534780978699e-06,
+      "loss": 0.3675,
+      "step": 2412
+    },
+    {
+      "epoch": 0.36084940930163,
+      "grad_norm": 2.0634027441697778,
+      "learning_rate": 7.394408990828726e-06,
+      "loss": 0.3529,
+      "step": 2413
+    },
+    {
+      "epoch": 0.3609989531927621,
+      "grad_norm": 1.6015524340522593,
+      "learning_rate": 7.392282638927326e-06,
+      "loss": 0.233,
+      "step": 2414
+    },
+    {
+      "epoch": 0.3611484970838941,
+      "grad_norm": 1.047148380596162,
+      "learning_rate": 7.390155725773365e-06,
+      "loss": 0.1861,
+      "step": 2415
+    },
+    {
+      "epoch": 0.36129804097502616,
+      "grad_norm": 1.6393955983196027,
+      "learning_rate": 7.388028251865837e-06,
+      "loss": 0.1914,
+      "step": 2416
+    },
+    {
+      "epoch": 0.36144758486615824,
+      "grad_norm": 1.3926718614765299,
+      "learning_rate": 7.385900217703865e-06,
+      "loss": 0.186,
+      "step": 2417
+    },
+    {
+      "epoch": 0.3615971287572903,
+      "grad_norm": 1.5958216662832228,
+      "learning_rate": 7.383771623786709e-06,
+      "loss": 0.2127,
+      "step": 2418
+    },
+    {
+      "epoch": 0.3617466726484223,
+      "grad_norm": 1.956943418621964,
+      "learning_rate": 7.381642470613758e-06,
+      "loss": 0.5055,
+      "step": 2419
+    },
+    {
+      "epoch": 0.36189621653955434,
+      "grad_norm": 2.2836322625873855,
+      "learning_rate": 7.37951275868453e-06,
+      "loss": 0.6337,
+      "step": 2420
+    },
+    {
+      "epoch": 0.3620457604306864,
+      "grad_norm": 1.922712275476603,
+      "learning_rate": 7.3773824884986744e-06,
+      "loss": 0.4718,
+      "step": 2421
+    },
+    {
+      "epoch": 0.36219530432181846,
+      "grad_norm": 1.01255124960959,
+      "learning_rate": 7.375251660555978e-06,
+      "loss": 0.1985,
+      "step": 2422
+    },
+    {
+      "epoch": 0.3623448482129505,
+      "grad_norm": 1.6681331401639201,
+      "learning_rate": 7.373120275356349e-06,
+      "loss": 0.4116,
+      "step": 2423
+    },
+    {
+      "epoch": 0.3624943921040826,
+      "grad_norm": 1.3837406537641457,
+      "learning_rate": 7.370988333399834e-06,
+      "loss": 0.2207,
+      "step": 2424
+    },
+    {
+      "epoch": 0.3626439359952146,
+      "grad_norm": 1.559290768311182,
+      "learning_rate": 7.3688558351866055e-06,
+      "loss": 0.2715,
+      "step": 2425
+    },
+    {
+      "epoch": 0.36279347988634664,
+      "grad_norm": 1.305139066591499,
+      "learning_rate": 7.366722781216968e-06,
+      "loss": 0.3011,
+      "step": 2426
+    },
+    {
+      "epoch": 0.36294302377747867,
+      "grad_norm": 1.6209491159652774,
+      "learning_rate": 7.3645891719913584e-06,
+      "loss": 0.3261,
+      "step": 2427
+    },
+    {
+      "epoch": 0.36309256766861076,
+      "grad_norm": 1.4951914962551227,
+      "learning_rate": 7.3624550080103385e-06,
+      "loss": 0.3393,
+      "step": 2428
+    },
+    {
+      "epoch": 0.3632421115597428,
+      "grad_norm": 1.8477503569108042,
+      "learning_rate": 7.360320289774607e-06,
+      "loss": 0.3535,
+      "step": 2429
+    },
+    {
+      "epoch": 0.3633916554508748,
+      "grad_norm": 1.8411787417724799,
+      "learning_rate": 7.358185017784989e-06,
+      "loss": 0.3847,
+      "step": 2430
+    },
+    {
+      "epoch": 0.3635411993420069,
+      "grad_norm": 1.3297346850852803,
+      "learning_rate": 7.356049192542439e-06,
+      "loss": 0.1938,
+      "step": 2431
+    },
+    {
+      "epoch": 0.36369074323313894,
+      "grad_norm": 1.746289398019253,
+      "learning_rate": 7.353912814548042e-06,
+      "loss": 0.5313,
+      "step": 2432
+    },
+    {
+      "epoch": 0.36384028712427097,
+      "grad_norm": 1.4432992935580808,
+      "learning_rate": 7.351775884303013e-06,
+      "loss": 0.3349,
+      "step": 2433
+    },
+    {
+      "epoch": 0.363989831015403,
+      "grad_norm": 0.9649898248590643,
+      "learning_rate": 7.349638402308696e-06,
+      "loss": 0.2208,
+      "step": 2434
+    },
+    {
+      "epoch": 0.3641393749065351,
+      "grad_norm": 1.8449998111001102,
+      "learning_rate": 7.347500369066567e-06,
+      "loss": 0.3445,
+      "step": 2435
+    },
+    {
+      "epoch": 0.3642889187976671,
+      "grad_norm": 1.7052431627521538,
+      "learning_rate": 7.345361785078227e-06,
+      "loss": 0.2205,
+      "step": 2436
+    },
+    {
+      "epoch": 0.36443846268879915,
+      "grad_norm": 1.172361472405274,
+      "learning_rate": 7.343222650845408e-06,
+      "loss": 0.217,
+      "step": 2437
+    },
+    {
+      "epoch": 0.36458800657993123,
+      "grad_norm": 1.6796925766446777,
+      "learning_rate": 7.341082966869975e-06,
+      "loss": 0.3589,
+      "step": 2438
+    },
+    {
+      "epoch": 0.36473755047106327,
+      "grad_norm": 1.7021974172455587,
+      "learning_rate": 7.3389427336539146e-06,
+      "loss": 0.3936,
+      "step": 2439
+    },
+    {
+      "epoch": 0.3648870943621953,
+      "grad_norm": 1.5350766828655338,
+      "learning_rate": 7.336801951699348e-06,
+      "loss": 0.4299,
+      "step": 2440
+    },
+    {
+      "epoch": 0.36503663825332733,
+      "grad_norm": 2.0471270711764524,
+      "learning_rate": 7.334660621508523e-06,
+      "loss": 0.6307,
+      "step": 2441
+    },
+    {
+      "epoch": 0.3651861821444594,
+      "grad_norm": 1.3866667939244877,
+      "learning_rate": 7.3325187435838145e-06,
+      "loss": 0.3435,
+      "step": 2442
+    },
+    {
+      "epoch": 0.36533572603559145,
+      "grad_norm": 1.9710057444965332,
+      "learning_rate": 7.330376318427731e-06,
+      "loss": 0.3715,
+      "step": 2443
+    },
+    {
+      "epoch": 0.3654852699267235,
+      "grad_norm": 1.9432365570364065,
+      "learning_rate": 7.328233346542906e-06,
+      "loss": 0.5569,
+      "step": 2444
+    },
+    {
+      "epoch": 0.36563481381785556,
+      "grad_norm": 1.0580374357411853,
+      "learning_rate": 7.326089828432097e-06,
+      "loss": 0.2205,
+      "step": 2445
+    },
+    {
+      "epoch": 0.3657843577089876,
+      "grad_norm": 1.9944374991832907,
+      "learning_rate": 7.323945764598198e-06,
+      "loss": 0.5225,
+      "step": 2446
+    },
+    {
+      "epoch": 0.3659339016001196,
+      "grad_norm": 1.3286346243167817,
+      "learning_rate": 7.321801155544227e-06,
+      "loss": 0.3126,
+      "step": 2447
+    },
+    {
+      "epoch": 0.36608344549125166,
+      "grad_norm": 1.797949701766733,
+      "learning_rate": 7.319656001773326e-06,
+      "loss": 0.5528,
+      "step": 2448
+    },
+    {
+      "epoch": 0.36623298938238374,
+      "grad_norm": 1.3871718042688361,
+      "learning_rate": 7.317510303788775e-06,
+      "loss": 0.1986,
+      "step": 2449
+    },
+    {
+      "epoch": 0.3663825332735158,
+      "grad_norm": 1.696323576703982,
+      "learning_rate": 7.31536406209397e-06,
+      "loss": 0.4646,
+      "step": 2450
+    },
+    {
+      "epoch": 0.3665320771646478,
+      "grad_norm": 1.112116781879356,
+      "learning_rate": 7.313217277192441e-06,
+      "loss": 0.1698,
+      "step": 2451
+    },
+    {
+      "epoch": 0.3666816210557799,
+      "grad_norm": 1.6253111202173256,
+      "learning_rate": 7.311069949587849e-06,
+      "loss": 0.368,
+      "step": 2452
+    },
+    {
+      "epoch": 0.3668311649469119,
+      "grad_norm": 1.8012549383451812,
+      "learning_rate": 7.308922079783972e-06,
+      "loss": 0.4943,
+      "step": 2453
+    },
+    {
+      "epoch": 0.36698070883804396,
+      "grad_norm": 1.73000381975083,
+      "learning_rate": 7.306773668284723e-06,
+      "loss": 0.543,
+      "step": 2454
+    },
+    {
+      "epoch": 0.367130252729176,
+      "grad_norm": 1.7982349971106675,
+      "learning_rate": 7.30462471559414e-06,
+      "loss": 0.3192,
+      "step": 2455
+    },
+    {
+      "epoch": 0.3672797966203081,
+      "grad_norm": 1.3452770378011227,
+      "learning_rate": 7.302475222216388e-06,
+      "loss": 0.3435,
+      "step": 2456
+    },
+    {
+      "epoch": 0.3674293405114401,
+      "grad_norm": 1.673472364447151,
+      "learning_rate": 7.300325188655762e-06,
+      "loss": 0.2812,
+      "step": 2457
+    },
+    {
+      "epoch": 0.36757888440257214,
+      "grad_norm": 1.5298434482917413,
+      "learning_rate": 7.298174615416676e-06,
+      "loss": 0.2351,
+      "step": 2458
+    },
+    {
+      "epoch": 0.3677284282937042,
+      "grad_norm": 0.9700694807997261,
+      "learning_rate": 7.2960235030036765e-06,
+      "loss": 0.1807,
+      "step": 2459
+    },
+    {
+      "epoch": 0.36787797218483625,
+      "grad_norm": 1.3628120017832577,
+      "learning_rate": 7.293871851921435e-06,
+      "loss": 0.1775,
+      "step": 2460
+    },
+    {
+      "epoch": 0.3680275160759683,
+      "grad_norm": 1.331669281066242,
+      "learning_rate": 7.29171966267475e-06,
+      "loss": 0.1953,
+      "step": 2461
+    },
+    {
+      "epoch": 0.3681770599671003,
+      "grad_norm": 1.7478823536432608,
+      "learning_rate": 7.2895669357685465e-06,
+      "loss": 0.402,
+      "step": 2462
+    },
+    {
+      "epoch": 0.3683266038582324,
+      "grad_norm": 1.0780884237235537,
+      "learning_rate": 7.287413671707875e-06,
+      "loss": 0.2042,
+      "step": 2463
+    },
+    {
+      "epoch": 0.36847614774936444,
+      "grad_norm": 1.39413350641363,
+      "learning_rate": 7.285259870997911e-06,
+      "loss": 0.2161,
+      "step": 2464
+    },
+    {
+      "epoch": 0.36862569164049647,
+      "grad_norm": 1.781457240613985,
+      "learning_rate": 7.283105534143957e-06,
+      "loss": 0.4025,
+      "step": 2465
+    },
+    {
+      "epoch": 0.36877523553162855,
+      "grad_norm": 1.298349295097918,
+      "learning_rate": 7.280950661651443e-06,
+      "loss": 0.2216,
+      "step": 2466
+    },
+    {
+      "epoch": 0.3689247794227606,
+      "grad_norm": 1.6025246428329694,
+      "learning_rate": 7.278795254025921e-06,
+      "loss": 0.2132,
+      "step": 2467
+    },
+    {
+      "epoch": 0.3690743233138926,
+      "grad_norm": 1.9856391732187633,
+      "learning_rate": 7.276639311773068e-06,
+      "loss": 0.5661,
+      "step": 2468
+    },
+    {
+      "epoch": 0.36922386720502465,
+      "grad_norm": 1.901744160559575,
+      "learning_rate": 7.274482835398695e-06,
+      "loss": 0.5376,
+      "step": 2469
+    },
+    {
+      "epoch": 0.36937341109615673,
+      "grad_norm": 1.3567504841626332,
+      "learning_rate": 7.272325825408728e-06,
+      "loss": 0.2408,
+      "step": 2470
+    },
+    {
+      "epoch": 0.36952295498728877,
+      "grad_norm": 2.2451530510094253,
+      "learning_rate": 7.270168282309222e-06,
+      "loss": 0.2229,
+      "step": 2471
+    },
+    {
+      "epoch": 0.3696724988784208,
+      "grad_norm": 1.3887976362795125,
+      "learning_rate": 7.268010206606361e-06,
+      "loss": 0.3222,
+      "step": 2472
+    },
+    {
+      "epoch": 0.3698220427695529,
+      "grad_norm": 1.2797777276840512,
+      "learning_rate": 7.265851598806446e-06,
+      "loss": 0.2033,
+      "step": 2473
+    },
+    {
+      "epoch": 0.3699715866606849,
+      "grad_norm": 1.191177092854937,
+      "learning_rate": 7.263692459415909e-06,
+      "loss": 0.2131,
+      "step": 2474
+    },
+    {
+      "epoch": 0.37012113055181695,
+      "grad_norm": 1.5776822158264978,
+      "learning_rate": 7.261532788941306e-06,
+      "loss": 0.3744,
+      "step": 2475
+    },
+    {
+      "epoch": 0.37027067444294903,
+      "grad_norm": 1.0854274920898752,
+      "learning_rate": 7.259372587889314e-06,
+      "loss": 0.233,
+      "step": 2476
+    },
+    {
+      "epoch": 0.37042021833408106,
+      "grad_norm": 1.2018243893675977,
+      "learning_rate": 7.25721185676674e-06,
+      "loss": 0.2363,
+      "step": 2477
+    },
+    {
+      "epoch": 0.3705697622252131,
+      "grad_norm": 0.9967469112175203,
+      "learning_rate": 7.25505059608051e-06,
+      "loss": 0.1983,
+      "step": 2478
+    },
+    {
+      "epoch": 0.3707193061163451,
+      "grad_norm": 1.2775894695719086,
+      "learning_rate": 7.252888806337678e-06,
+      "loss": 0.208,
+      "step": 2479
+    },
+    {
+      "epoch": 0.3708688500074772,
+      "grad_norm": 1.5952345161170771,
+      "learning_rate": 7.25072648804542e-06,
+      "loss": 0.2703,
+      "step": 2480
+    },
+    {
+      "epoch": 0.37101839389860924,
+      "grad_norm": 1.1681377853119423,
+      "learning_rate": 7.248563641711036e-06,
+      "loss": 0.2119,
+      "step": 2481
+    },
+    {
+      "epoch": 0.3711679377897413,
+      "grad_norm": 1.9271882432441645,
+      "learning_rate": 7.2464002678419524e-06,
+      "loss": 0.3501,
+      "step": 2482
+    },
+    {
+      "epoch": 0.37131748168087336,
+      "grad_norm": 1.4801233084882979,
+      "learning_rate": 7.244236366945715e-06,
+      "loss": 0.2263,
+      "step": 2483
+    },
+    {
+      "epoch": 0.3714670255720054,
+      "grad_norm": 1.464705478500087,
+      "learning_rate": 7.242071939529999e-06,
+      "loss": 0.2709,
+      "step": 2484
+    },
+    {
+      "epoch": 0.3716165694631374,
+      "grad_norm": 1.5487734242577513,
+      "learning_rate": 7.239906986102598e-06,
+      "loss": 0.1828,
+      "step": 2485
+    },
+    {
+      "epoch": 0.37176611335426946,
+      "grad_norm": 1.1511925750155394,
+      "learning_rate": 7.237741507171432e-06,
+      "loss": 0.1896,
+      "step": 2486
+    },
+    {
+      "epoch": 0.37191565724540154,
+      "grad_norm": 1.6178218431333782,
+      "learning_rate": 7.235575503244542e-06,
+      "loss": 0.3265,
+      "step": 2487
+    },
+    {
+      "epoch": 0.3720652011365336,
+      "grad_norm": 1.005983926413353,
+      "learning_rate": 7.233408974830093e-06,
+      "loss": 0.173,
+      "step": 2488
+    },
+    {
+      "epoch": 0.3722147450276656,
+      "grad_norm": 1.0227776759122873,
+      "learning_rate": 7.231241922436374e-06,
+      "loss": 0.1948,
+      "step": 2489
+    },
+    {
+      "epoch": 0.3723642889187977,
+      "grad_norm": 1.6953767597307763,
+      "learning_rate": 7.229074346571798e-06,
+      "loss": 0.5005,
+      "step": 2490
+    },
+    {
+      "epoch": 0.3725138328099297,
+      "grad_norm": 1.5875223115043986,
+      "learning_rate": 7.226906247744897e-06,
+      "loss": 0.3688,
+      "step": 2491
+    },
+    {
+      "epoch": 0.37266337670106175,
+      "grad_norm": 2.27185613487943,
+      "learning_rate": 7.2247376264643294e-06,
+      "loss": 0.4033,
+      "step": 2492
+    },
+    {
+      "epoch": 0.3728129205921938,
+      "grad_norm": 1.6290529057007825,
+      "learning_rate": 7.2225684832388745e-06,
+      "loss": 0.3383,
+      "step": 2493
+    },
+    {
+      "epoch": 0.3729624644833259,
+      "grad_norm": 1.0562741707118548,
+      "learning_rate": 7.220398818577432e-06,
+      "loss": 0.182,
+      "step": 2494
+    },
+    {
+      "epoch": 0.3731120083744579,
+      "grad_norm": 1.3801253740702135,
+      "learning_rate": 7.21822863298903e-06,
+      "loss": 0.4305,
+      "step": 2495
+    },
+    {
+      "epoch": 0.37326155226558994,
+      "grad_norm": 1.3558340102680362,
+      "learning_rate": 7.216057926982811e-06,
+      "loss": 0.2011,
+      "step": 2496
+    },
+    {
+      "epoch": 0.373411096156722,
+      "grad_norm": 1.6070744311623764,
+      "learning_rate": 7.213886701068047e-06,
+      "loss": 0.3777,
+      "step": 2497
+    },
+    {
+      "epoch": 0.37356064004785405,
+      "grad_norm": 1.0945948539403993,
+      "learning_rate": 7.211714955754125e-06,
+      "loss": 0.154,
+      "step": 2498
+    },
+    {
+      "epoch": 0.3737101839389861,
+      "grad_norm": 1.1736153236745563,
+      "learning_rate": 7.2095426915505605e-06,
+      "loss": 0.1973,
+      "step": 2499
+    },
+    {
+      "epoch": 0.3738597278301181,
+      "grad_norm": 1.5381442491891666,
+      "learning_rate": 7.207369908966987e-06,
+      "loss": 0.2319,
+      "step": 2500
+    },
+    {
+      "epoch": 0.3740092717212502,
+      "grad_norm": 1.143218245070331,
+      "learning_rate": 7.2051966085131584e-06,
+      "loss": 0.1947,
+      "step": 2501
+    },
+    {
+      "epoch": 0.37415881561238223,
+      "grad_norm": 1.756151915421153,
+      "learning_rate": 7.203022790698954e-06,
+      "loss": 0.4682,
+      "step": 2502
+    },
+    {
+      "epoch": 0.37430835950351427,
+      "grad_norm": 1.956298123550812,
+      "learning_rate": 7.20084845603437e-06,
+      "loss": 0.5118,
+      "step": 2503
+    },
+    {
+      "epoch": 0.37445790339464635,
+      "grad_norm": 1.9158174915897688,
+      "learning_rate": 7.198673605029529e-06,
+      "loss": 0.3117,
+      "step": 2504
+    },
+    {
+      "epoch": 0.3746074472857784,
+      "grad_norm": 4.295053824375029,
+      "learning_rate": 7.196498238194672e-06,
+      "loss": 0.2234,
+      "step": 2505
+    },
+    {
+      "epoch": 0.3747569911769104,
+      "grad_norm": 1.6941035724235585,
+      "learning_rate": 7.194322356040159e-06,
+      "loss": 0.2312,
+      "step": 2506
+    },
+    {
+      "epoch": 0.37490653506804245,
+      "grad_norm": 1.5640016182855432,
+      "learning_rate": 7.192145959076474e-06,
+      "loss": 0.3622,
+      "step": 2507
+    },
+    {
+      "epoch": 0.37505607895917453,
+      "grad_norm": 1.8466707952598986,
+      "learning_rate": 7.1899690478142196e-06,
+      "loss": 0.4891,
+      "step": 2508
+    },
+    {
+      "epoch": 0.37520562285030656,
+      "grad_norm": 1.4306072393075668,
+      "learning_rate": 7.18779162276412e-06,
+      "loss": 0.2988,
+      "step": 2509
+    },
+    {
+      "epoch": 0.3753551667414386,
+      "grad_norm": 1.4306403617067467,
+      "learning_rate": 7.185613684437024e-06,
+      "loss": 0.3184,
+      "step": 2510
+    },
+    {
+      "epoch": 0.3755047106325707,
+      "grad_norm": 1.6012766699315721,
+      "learning_rate": 7.183435233343892e-06,
+      "loss": 0.3621,
+      "step": 2511
+    },
+    {
+      "epoch": 0.3756542545237027,
+      "grad_norm": 1.2931435832729816,
+      "learning_rate": 7.181256269995813e-06,
+      "loss": 0.3318,
+      "step": 2512
+    },
+    {
+      "epoch": 0.37580379841483474,
+      "grad_norm": 1.422093979772559,
+      "learning_rate": 7.179076794903991e-06,
+      "loss": 0.2357,
+      "step": 2513
+    },
+    {
+      "epoch": 0.3759533423059668,
+      "grad_norm": 0.9257904335966806,
+      "learning_rate": 7.176896808579752e-06,
+      "loss": 0.1813,
+      "step": 2514
+    },
+    {
+      "epoch": 0.37610288619709886,
+      "grad_norm": 2.037815545812073,
+      "learning_rate": 7.174716311534542e-06,
+      "loss": 0.2783,
+      "step": 2515
+    },
+    {
+      "epoch": 0.3762524300882309,
+      "grad_norm": 1.71379595660131,
+      "learning_rate": 7.172535304279926e-06,
+      "loss": 0.4839,
+      "step": 2516
+    },
+    {
+      "epoch": 0.3764019739793629,
+      "grad_norm": 2.00916868636801,
+      "learning_rate": 7.170353787327593e-06,
+      "loss": 0.6153,
+      "step": 2517
+    },
+    {
+      "epoch": 0.376551517870495,
+      "grad_norm": 1.5789476931879063,
+      "learning_rate": 7.168171761189343e-06,
+      "loss": 0.2031,
+      "step": 2518
+    },
+    {
+      "epoch": 0.37670106176162704,
+      "grad_norm": 1.684853217653668,
+      "learning_rate": 7.165989226377103e-06,
+      "loss": 0.3816,
+      "step": 2519
+    },
+    {
+      "epoch": 0.3768506056527591,
+      "grad_norm": 1.9200462619229837,
+      "learning_rate": 7.163806183402916e-06,
+      "loss": 0.4962,
+      "step": 2520
+    },
+    {
+      "epoch": 0.3770001495438911,
+      "grad_norm": 1.8855548807104212,
+      "learning_rate": 7.161622632778944e-06,
+      "loss": 0.2185,
+      "step": 2521
+    },
+    {
+      "epoch": 0.3771496934350232,
+      "grad_norm": 1.6740717226550226,
+      "learning_rate": 7.159438575017471e-06,
+      "loss": 0.3777,
+      "step": 2522
+    },
+    {
+      "epoch": 0.3772992373261552,
+      "grad_norm": 1.397870156480417,
+      "learning_rate": 7.157254010630896e-06,
+      "loss": 0.2426,
+      "step": 2523
+    },
+    {
+      "epoch": 0.37744878121728725,
+      "grad_norm": 1.6145415521158457,
+      "learning_rate": 7.155068940131741e-06,
+      "loss": 0.3926,
+      "step": 2524
+    },
+    {
+      "epoch": 0.37759832510841934,
+      "grad_norm": 1.7042594169803778,
+      "learning_rate": 7.152883364032644e-06,
+      "loss": 0.1917,
+      "step": 2525
+    },
+    {
+      "epoch": 0.3777478689995514,
+      "grad_norm": 1.5289841742678918,
+      "learning_rate": 7.15069728284636e-06,
+      "loss": 0.351,
+      "step": 2526
+    },
+    {
+      "epoch": 0.3778974128906834,
+      "grad_norm": 2.395918097586016,
+      "learning_rate": 7.148510697085767e-06,
+      "loss": 0.3126,
+      "step": 2527
+    },
+    {
+      "epoch": 0.37804695678181544,
+      "grad_norm": 2.703467639529761,
+      "learning_rate": 7.146323607263859e-06,
+      "loss": 0.2623,
+      "step": 2528
+    },
+    {
+      "epoch": 0.3781965006729475,
+      "grad_norm": 1.7671642324701808,
+      "learning_rate": 7.144136013893745e-06,
+      "loss": 0.2137,
+      "step": 2529
+    },
+    {
+      "epoch": 0.37834604456407955,
+      "grad_norm": 1.932948758218732,
+      "learning_rate": 7.141947917488663e-06,
+      "loss": 0.3711,
+      "step": 2530
+    },
+    {
+      "epoch": 0.3784955884552116,
+      "grad_norm": 1.2026669943340598,
+      "learning_rate": 7.139759318561954e-06,
+      "loss": 0.171,
+      "step": 2531
+    },
+    {
+      "epoch": 0.37864513234634367,
+      "grad_norm": 1.7885589140732676,
+      "learning_rate": 7.137570217627088e-06,
+      "loss": 0.4937,
+      "step": 2532
+    },
+    {
+      "epoch": 0.3787946762374757,
+      "grad_norm": 1.6235037343132346,
+      "learning_rate": 7.135380615197649e-06,
+      "loss": 0.4338,
+      "step": 2533
+    },
+    {
+      "epoch": 0.37894422012860773,
+      "grad_norm": 1.4352072567044432,
+      "learning_rate": 7.133190511787337e-06,
+      "loss": 0.203,
+      "step": 2534
+    },
+    {
+      "epoch": 0.3790937640197398,
+      "grad_norm": 2.021523435648341,
+      "learning_rate": 7.130999907909972e-06,
+      "loss": 0.4868,
+      "step": 2535
+    },
+    {
+      "epoch": 0.37924330791087185,
+      "grad_norm": 1.575689088228456,
+      "learning_rate": 7.128808804079492e-06,
+      "loss": 0.3587,
+      "step": 2536
+    },
+    {
+      "epoch": 0.3793928518020039,
+      "grad_norm": 1.538572538393702,
+      "learning_rate": 7.126617200809951e-06,
+      "loss": 0.3164,
+      "step": 2537
+    },
+    {
+      "epoch": 0.3795423956931359,
+      "grad_norm": 2.1498396130276727,
+      "learning_rate": 7.12442509861552e-06,
+      "loss": 0.6093,
+      "step": 2538
+    },
+    {
+      "epoch": 0.379691939584268,
+      "grad_norm": 1.5246107523171812,
+      "learning_rate": 7.122232498010486e-06,
+      "loss": 0.2357,
+      "step": 2539
+    },
+    {
+      "epoch": 0.37984148347540003,
+      "grad_norm": 1.5356177541463252,
+      "learning_rate": 7.120039399509257e-06,
+      "loss": 0.2945,
+      "step": 2540
+    },
+    {
+      "epoch": 0.37999102736653206,
+      "grad_norm": 1.2696297372616192,
+      "learning_rate": 7.117845803626352e-06,
+      "loss": 0.2568,
+      "step": 2541
+    },
+    {
+      "epoch": 0.38014057125766415,
+      "grad_norm": 1.2940629840823208,
+      "learning_rate": 7.115651710876411e-06,
+      "loss": 0.2167,
+      "step": 2542
+    },
+    {
+      "epoch": 0.3802901151487962,
+      "grad_norm": 1.6760687859259409,
+      "learning_rate": 7.11345712177419e-06,
+      "loss": 0.4574,
+      "step": 2543
+    },
+    {
+      "epoch": 0.3804396590399282,
+      "grad_norm": 1.8274306217532401,
+      "learning_rate": 7.111262036834559e-06,
+      "loss": 0.1839,
+      "step": 2544
+    },
+    {
+      "epoch": 0.38058920293106024,
+      "grad_norm": 1.7239605405645464,
+      "learning_rate": 7.109066456572508e-06,
+      "loss": 0.3814,
+      "step": 2545
+    },
+    {
+      "epoch": 0.38073874682219233,
+      "grad_norm": 2.0774382697068146,
+      "learning_rate": 7.106870381503139e-06,
+      "loss": 0.3502,
+      "step": 2546
+    },
+    {
+      "epoch": 0.38088829071332436,
+      "grad_norm": 1.6407126721162997,
+      "learning_rate": 7.104673812141676e-06,
+      "loss": 0.3768,
+      "step": 2547
+    },
+    {
+      "epoch": 0.3810378346044564,
+      "grad_norm": 1.4264932056782313,
+      "learning_rate": 7.102476749003453e-06,
+      "loss": 0.2587,
+      "step": 2548
+    },
+    {
+      "epoch": 0.3811873784955885,
+      "grad_norm": 1.4315314856298393,
+      "learning_rate": 7.1002791926039204e-06,
+      "loss": 0.1918,
+      "step": 2549
+    },
+    {
+      "epoch": 0.3813369223867205,
+      "grad_norm": 2.268263838831847,
+      "learning_rate": 7.098081143458649e-06,
+      "loss": 0.2129,
+      "step": 2550
+    },
+    {
+      "epoch": 0.38148646627785254,
+      "grad_norm": 1.4599724810600025,
+      "learning_rate": 7.095882602083321e-06,
+      "loss": 0.3225,
+      "step": 2551
+    },
+    {
+      "epoch": 0.3816360101689846,
+      "grad_norm": 1.4198436584338292,
+      "learning_rate": 7.0936835689937366e-06,
+      "loss": 0.1947,
+      "step": 2552
+    },
+    {
+      "epoch": 0.38178555406011666,
+      "grad_norm": 1.3581488206510595,
+      "learning_rate": 7.09148404470581e-06,
+      "loss": 0.2492,
+      "step": 2553
+    },
+    {
+      "epoch": 0.3819350979512487,
+      "grad_norm": 1.8329873345541536,
+      "learning_rate": 7.089284029735568e-06,
+      "loss": 0.1999,
+      "step": 2554
+    },
+    {
+      "epoch": 0.3820846418423807,
+      "grad_norm": 1.879696219955982,
+      "learning_rate": 7.087083524599158e-06,
+      "loss": 0.2107,
+      "step": 2555
+    },
+    {
+      "epoch": 0.3822341857335128,
+      "grad_norm": 1.2252237248239084,
+      "learning_rate": 7.08488252981284e-06,
+      "loss": 0.1991,
+      "step": 2556
+    },
+    {
+      "epoch": 0.38238372962464484,
+      "grad_norm": 1.952002278934872,
+      "learning_rate": 7.082681045892988e-06,
+      "loss": 0.3916,
+      "step": 2557
+    },
+    {
+      "epoch": 0.3825332735157769,
+      "grad_norm": 1.7256855312550468,
+      "learning_rate": 7.08047907335609e-06,
+      "loss": 0.3116,
+      "step": 2558
+    },
+    {
+      "epoch": 0.3826828174069089,
+      "grad_norm": 1.587038951947857,
+      "learning_rate": 7.078276612718752e-06,
+      "loss": 0.2224,
+      "step": 2559
+    },
+    {
+      "epoch": 0.382832361298041,
+      "grad_norm": 1.8262578444211246,
+      "learning_rate": 7.076073664497691e-06,
+      "loss": 0.3618,
+      "step": 2560
+    },
+    {
+      "epoch": 0.382981905189173,
+      "grad_norm": 2.0217394092028167,
+      "learning_rate": 7.07387022920974e-06,
+      "loss": 0.2429,
+      "step": 2561
+    },
+    {
+      "epoch": 0.38313144908030505,
+      "grad_norm": 1.9141258813134443,
+      "learning_rate": 7.071666307371847e-06,
+      "loss": 0.4741,
+      "step": 2562
+    },
+    {
+      "epoch": 0.38328099297143714,
+      "grad_norm": 1.432198404107898,
+      "learning_rate": 7.069461899501073e-06,
+      "loss": 0.1918,
+      "step": 2563
+    },
+    {
+      "epoch": 0.38343053686256917,
+      "grad_norm": 1.5928510931135798,
+      "learning_rate": 7.067257006114593e-06,
+      "loss": 0.3226,
+      "step": 2564
+    },
+    {
+      "epoch": 0.3835800807537012,
+      "grad_norm": 1.6585372345576896,
+      "learning_rate": 7.065051627729698e-06,
+      "loss": 0.3276,
+      "step": 2565
+    },
+    {
+      "epoch": 0.38372962464483323,
+      "grad_norm": 1.2734067275418923,
+      "learning_rate": 7.062845764863787e-06,
+      "loss": 0.3313,
+      "step": 2566
+    },
+    {
+      "epoch": 0.3838791685359653,
+      "grad_norm": 1.1374789118122002,
+      "learning_rate": 7.0606394180343805e-06,
+      "loss": 0.1907,
+      "step": 2567
+    },
+    {
+      "epoch": 0.38402871242709735,
+      "grad_norm": 1.549559123650284,
+      "learning_rate": 7.058432587759107e-06,
+      "loss": 0.3915,
+      "step": 2568
+    },
+    {
+      "epoch": 0.3841782563182294,
+      "grad_norm": 1.4708906613930632,
+      "learning_rate": 7.0562252745557115e-06,
+      "loss": 0.1992,
+      "step": 2569
+    },
+    {
+      "epoch": 0.38432780020936147,
+      "grad_norm": 1.6323056255608734,
+      "learning_rate": 7.054017478942048e-06,
+      "loss": 0.2221,
+      "step": 2570
+    },
+    {
+      "epoch": 0.3844773441004935,
+      "grad_norm": 1.5203359282551667,
+      "learning_rate": 7.0518092014360905e-06,
+      "loss": 0.3545,
+      "step": 2571
+    },
+    {
+      "epoch": 0.38462688799162553,
+      "grad_norm": 1.159124982100583,
+      "learning_rate": 7.0496004425559195e-06,
+      "loss": 0.1549,
+      "step": 2572
+    },
+    {
+      "epoch": 0.38477643188275756,
+      "grad_norm": 2.1232683966349586,
+      "learning_rate": 7.047391202819734e-06,
+      "loss": 0.6264,
+      "step": 2573
+    },
+    {
+      "epoch": 0.38492597577388965,
+      "grad_norm": 1.0622764450389663,
+      "learning_rate": 7.045181482745837e-06,
+      "loss": 0.2163,
+      "step": 2574
+    },
+    {
+      "epoch": 0.3850755196650217,
+      "grad_norm": 1.077605292843362,
+      "learning_rate": 7.042971282852656e-06,
+      "loss": 0.2079,
+      "step": 2575
+    },
+    {
+      "epoch": 0.3852250635561537,
+      "grad_norm": 1.0767369768412656,
+      "learning_rate": 7.040760603658723e-06,
+      "loss": 0.2035,
+      "step": 2576
+    },
+    {
+      "epoch": 0.3853746074472858,
+      "grad_norm": 1.5110599625002836,
+      "learning_rate": 7.038549445682685e-06,
+      "loss": 0.1561,
+      "step": 2577
+    },
+    {
+      "epoch": 0.38552415133841783,
+      "grad_norm": 1.6236972538616068,
+      "learning_rate": 7.036337809443301e-06,
+      "loss": 0.4143,
+      "step": 2578
+    },
+    {
+      "epoch": 0.38567369522954986,
+      "grad_norm": 1.6146980603793564,
+      "learning_rate": 7.0341256954594415e-06,
+      "loss": 0.332,
+      "step": 2579
+    },
+    {
+      "epoch": 0.3858232391206819,
+      "grad_norm": 2.047834790133371,
+      "learning_rate": 7.031913104250091e-06,
+      "loss": 0.4432,
+      "step": 2580
+    },
+    {
+      "epoch": 0.385972783011814,
+      "grad_norm": 1.5013503959890904,
+      "learning_rate": 7.029700036334344e-06,
+      "loss": 0.344,
+      "step": 2581
+    },
+    {
+      "epoch": 0.386122326902946,
+      "grad_norm": 1.0672948108215132,
+      "learning_rate": 7.027486492231407e-06,
+      "loss": 0.2127,
+      "step": 2582
+    },
+    {
+      "epoch": 0.38627187079407804,
+      "grad_norm": 1.6877080780242262,
+      "learning_rate": 7.025272472460598e-06,
+      "loss": 0.4702,
+      "step": 2583
+    },
+    {
+      "epoch": 0.38642141468521013,
+      "grad_norm": 1.5687512279976832,
+      "learning_rate": 7.02305797754135e-06,
+      "loss": 0.345,
+      "step": 2584
+    },
+    {
+      "epoch": 0.38657095857634216,
+      "grad_norm": 1.268209431704893,
+      "learning_rate": 7.020843007993203e-06,
+      "loss": 0.1817,
+      "step": 2585
+    },
+    {
+      "epoch": 0.3867205024674742,
+      "grad_norm": 2.1026474658538508,
+      "learning_rate": 7.018627564335813e-06,
+      "loss": 0.3595,
+      "step": 2586
+    },
+    {
+      "epoch": 0.3868700463586062,
+      "grad_norm": 1.283399204038543,
+      "learning_rate": 7.01641164708894e-06,
+      "loss": 0.3985,
+      "step": 2587
+    },
+    {
+      "epoch": 0.3870195902497383,
+      "grad_norm": 1.5628231644223662,
+      "learning_rate": 7.014195256772462e-06,
+      "loss": 0.243,
+      "step": 2588
+    },
+    {
+      "epoch": 0.38716913414087034,
+      "grad_norm": 1.4666711014261535,
+      "learning_rate": 7.011978393906366e-06,
+      "loss": 0.3757,
+      "step": 2589
+    },
+    {
+      "epoch": 0.3873186780320024,
+      "grad_norm": 1.7424029593387464,
+      "learning_rate": 7.009761059010746e-06,
+      "loss": 0.3297,
+      "step": 2590
+    },
+    {
+      "epoch": 0.38746822192313446,
+      "grad_norm": 1.793037388304316,
+      "learning_rate": 7.007543252605815e-06,
+      "loss": 0.1902,
+      "step": 2591
+    },
+    {
+      "epoch": 0.3876177658142665,
+      "grad_norm": 1.3063151548645278,
+      "learning_rate": 7.005324975211889e-06,
+      "loss": 0.2154,
+      "step": 2592
+    },
+    {
+      "epoch": 0.3877673097053985,
+      "grad_norm": 1.5791521412257716,
+      "learning_rate": 7.003106227349399e-06,
+      "loss": 0.353,
+      "step": 2593
+    },
+    {
+      "epoch": 0.3879168535965306,
+      "grad_norm": 1.1004684710498887,
+      "learning_rate": 7.0008870095388815e-06,
+      "loss": 0.1875,
+      "step": 2594
+    },
+    {
+      "epoch": 0.38806639748766264,
+      "grad_norm": 1.5070698124572854,
+      "learning_rate": 6.998667322300989e-06,
+      "loss": 0.3211,
+      "step": 2595
+    },
+    {
+      "epoch": 0.38821594137879467,
+      "grad_norm": 1.5281886376155964,
+      "learning_rate": 6.9964471661564815e-06,
+      "loss": 0.2161,
+      "step": 2596
+    },
+    {
+      "epoch": 0.3883654852699267,
+      "grad_norm": 1.5252156175553138,
+      "learning_rate": 6.994226541626227e-06,
+      "loss": 0.4639,
+      "step": 2597
+    },
+    {
+      "epoch": 0.3885150291610588,
+      "grad_norm": 1.0809213992053146,
+      "learning_rate": 6.9920054492312086e-06,
+      "loss": 0.1773,
+      "step": 2598
+    },
+    {
+      "epoch": 0.3886645730521908,
+      "grad_norm": 1.0167620236333288,
+      "learning_rate": 6.989783889492512e-06,
+      "loss": 0.197,
+      "step": 2599
+    },
+    {
+      "epoch": 0.38881411694332285,
+      "grad_norm": 1.9918245714247895,
+      "learning_rate": 6.98756186293134e-06,
+      "loss": 0.2173,
+      "step": 2600
+    },
+    {
+      "epoch": 0.38896366083445494,
+      "grad_norm": 1.134571833649737,
+      "learning_rate": 6.9853393700689995e-06,
+      "loss": 0.2177,
+      "step": 2601
+    },
+    {
+      "epoch": 0.38911320472558697,
+      "grad_norm": 1.2111914722156627,
+      "learning_rate": 6.98311641142691e-06,
+      "loss": 0.189,
+      "step": 2602
+    },
+    {
+      "epoch": 0.389262748616719,
+      "grad_norm": 1.5006726664035277,
+      "learning_rate": 6.9808929875265974e-06,
+      "loss": 0.3254,
+      "step": 2603
+    },
+    {
+      "epoch": 0.38941229250785103,
+      "grad_norm": 1.6055883887420528,
+      "learning_rate": 6.9786690988897e-06,
+      "loss": 0.3344,
+      "step": 2604
+    },
+    {
+      "epoch": 0.3895618363989831,
+      "grad_norm": 0.9973538599901819,
+      "learning_rate": 6.9764447460379625e-06,
+      "loss": 0.1393,
+      "step": 2605
+    },
+    {
+      "epoch": 0.38971138029011515,
+      "grad_norm": 1.6233042755021416,
+      "learning_rate": 6.9742199294932415e-06,
+      "loss": 0.204,
+      "step": 2606
+    },
+    {
+      "epoch": 0.3898609241812472,
+      "grad_norm": 1.3344019910552745,
+      "learning_rate": 6.971994649777497e-06,
+      "loss": 0.1945,
+      "step": 2607
+    },
+    {
+      "epoch": 0.39001046807237927,
+      "grad_norm": 1.3137908772185043,
+      "learning_rate": 6.969768907412804e-06,
+      "loss": 0.1984,
+      "step": 2608
+    },
+    {
+      "epoch": 0.3901600119635113,
+      "grad_norm": 1.777140683321665,
+      "learning_rate": 6.9675427029213405e-06,
+      "loss": 0.3008,
+      "step": 2609
+    },
+    {
+      "epoch": 0.39030955585464333,
+      "grad_norm": 1.0681843692748616,
+      "learning_rate": 6.965316036825398e-06,
+      "loss": 0.2006,
+      "step": 2610
+    },
+    {
+      "epoch": 0.39045909974577536,
+      "grad_norm": 1.8543912871172343,
+      "learning_rate": 6.963088909647372e-06,
+      "loss": 0.5414,
+      "step": 2611
+    },
+    {
+      "epoch": 0.39060864363690745,
+      "grad_norm": 1.9579644690684397,
+      "learning_rate": 6.960861321909769e-06,
+      "loss": 0.2134,
+      "step": 2612
+    },
+    {
+      "epoch": 0.3907581875280395,
+      "grad_norm": 1.5162323884971123,
+      "learning_rate": 6.9586332741352025e-06,
+      "loss": 0.4165,
+      "step": 2613
+    },
+    {
+      "epoch": 0.3909077314191715,
+      "grad_norm": 1.6022574800937832,
+      "learning_rate": 6.956404766846394e-06,
+      "loss": 0.2272,
+      "step": 2614
+    },
+    {
+      "epoch": 0.3910572753103036,
+      "grad_norm": 1.6770508498447996,
+      "learning_rate": 6.954175800566172e-06,
+      "loss": 0.4977,
+      "step": 2615
+    },
+    {
+      "epoch": 0.39120681920143563,
+      "grad_norm": 1.8009556625869532,
+      "learning_rate": 6.9519463758174745e-06,
+      "loss": 0.5158,
+      "step": 2616
+    },
+    {
+      "epoch": 0.39135636309256766,
+      "grad_norm": 1.3758659710920118,
+      "learning_rate": 6.949716493123345e-06,
+      "loss": 0.1789,
+      "step": 2617
+    },
+    {
+      "epoch": 0.3915059069836997,
+      "grad_norm": 1.3547231618901736,
+      "learning_rate": 6.947486153006937e-06,
+      "loss": 0.3534,
+      "step": 2618
+    },
+    {
+      "epoch": 0.3916554508748318,
+      "grad_norm": 1.5074942912698235,
+      "learning_rate": 6.945255355991509e-06,
+      "loss": 0.2497,
+      "step": 2619
+    },
+    {
+      "epoch": 0.3918049947659638,
+      "grad_norm": 1.2442646683398755,
+      "learning_rate": 6.943024102600428e-06,
+      "loss": 0.1861,
+      "step": 2620
+    },
+    {
+      "epoch": 0.39195453865709584,
+      "grad_norm": 1.5075321060331324,
+      "learning_rate": 6.940792393357165e-06,
+      "loss": 0.2194,
+      "step": 2621
+    },
+    {
+      "epoch": 0.39210408254822793,
+      "grad_norm": 1.683076961752896,
+      "learning_rate": 6.938560228785304e-06,
+      "loss": 0.2156,
+      "step": 2622
+    },
+    {
+      "epoch": 0.39225362643935996,
+      "grad_norm": 1.7112132072394493,
+      "learning_rate": 6.9363276094085296e-06,
+      "loss": 0.3958,
+      "step": 2623
+    },
+    {
+      "epoch": 0.392403170330492,
+      "grad_norm": 0.7048488756753248,
+      "learning_rate": 6.934094535750638e-06,
+      "loss": 0.2183,
+      "step": 2624
+    },
+    {
+      "epoch": 0.392552714221624,
+      "grad_norm": 1.7586652848799478,
+      "learning_rate": 6.931861008335527e-06,
+      "loss": 0.469,
+      "step": 2625
+    },
+    {
+      "epoch": 0.3927022581127561,
+      "grad_norm": 1.1588740664998758,
+      "learning_rate": 6.929627027687207e-06,
+      "loss": 0.2168,
+      "step": 2626
+    },
+    {
+      "epoch": 0.39285180200388814,
+      "grad_norm": 1.548868104243127,
+      "learning_rate": 6.927392594329789e-06,
+      "loss": 0.2312,
+      "step": 2627
+    },
+    {
+      "epoch": 0.39300134589502017,
+      "grad_norm": 1.595689192189307,
+      "learning_rate": 6.925157708787493e-06,
+      "loss": 0.4765,
+      "step": 2628
+    },
+    {
+      "epoch": 0.39315088978615226,
+      "grad_norm": 1.0277723372062315,
+      "learning_rate": 6.922922371584647e-06,
+      "loss": 0.1517,
+      "step": 2629
+    },
+    {
+      "epoch": 0.3933004336772843,
+      "grad_norm": 1.7591622574617085,
+      "learning_rate": 6.920686583245679e-06,
+      "loss": 0.3411,
+      "step": 2630
+    },
+    {
+      "epoch": 0.3934499775684163,
+      "grad_norm": 1.6744764890594985,
+      "learning_rate": 6.918450344295129e-06,
+      "loss": 0.2044,
+      "step": 2631
+    },
+    {
+      "epoch": 0.39359952145954835,
+      "grad_norm": 1.4007607932440016,
+      "learning_rate": 6.916213655257639e-06,
+      "loss": 0.3232,
+      "step": 2632
+    },
+    {
+      "epoch": 0.39374906535068044,
+      "grad_norm": 1.4204862226075206,
+      "learning_rate": 6.9139765166579576e-06,
+      "loss": 0.2314,
+      "step": 2633
+    },
+    {
+      "epoch": 0.39389860924181247,
+      "grad_norm": 1.3405577478235213,
+      "learning_rate": 6.9117389290209415e-06,
+      "loss": 0.211,
+      "step": 2634
+    },
+    {
+      "epoch": 0.3940481531329445,
+      "grad_norm": 1.454026439973042,
+      "learning_rate": 6.909500892871547e-06,
+      "loss": 0.33,
+      "step": 2635
+    },
+    {
+      "epoch": 0.3941976970240766,
+      "grad_norm": 1.3439856339271192,
+      "learning_rate": 6.907262408734842e-06,
+      "loss": 0.1815,
+      "step": 2636
+    },
+    {
+      "epoch": 0.3943472409152086,
+      "grad_norm": 1.6261472549425744,
+      "learning_rate": 6.905023477135992e-06,
+      "loss": 0.3739,
+      "step": 2637
+    },
+    {
+      "epoch": 0.39449678480634065,
+      "grad_norm": 1.576356761161988,
+      "learning_rate": 6.902784098600277e-06,
+      "loss": 0.3767,
+      "step": 2638
+    },
+    {
+      "epoch": 0.3946463286974727,
+      "grad_norm": 1.339777085672714,
+      "learning_rate": 6.9005442736530745e-06,
+      "loss": 0.1876,
+      "step": 2639
+    },
+    {
+      "epoch": 0.39479587258860477,
+      "grad_norm": 1.4248530907151669,
+      "learning_rate": 6.898304002819869e-06,
+      "loss": 0.2061,
+      "step": 2640
+    },
+    {
+      "epoch": 0.3949454164797368,
+      "grad_norm": 1.8979002423907008,
+      "learning_rate": 6.896063286626251e-06,
+      "loss": 0.3686,
+      "step": 2641
+    },
+    {
+      "epoch": 0.39509496037086883,
+      "grad_norm": 1.4615671922332303,
+      "learning_rate": 6.893822125597911e-06,
+      "loss": 0.2747,
+      "step": 2642
+    },
+    {
+      "epoch": 0.3952445042620009,
+      "grad_norm": 1.1050718617792725,
+      "learning_rate": 6.891580520260649e-06,
+      "loss": 0.1886,
+      "step": 2643
+    },
+    {
+      "epoch": 0.39539404815313295,
+      "grad_norm": 1.1930459776952358,
+      "learning_rate": 6.8893384711403675e-06,
+      "loss": 0.1788,
+      "step": 2644
+    },
+    {
+      "epoch": 0.395543592044265,
+      "grad_norm": 1.0036636367374232,
+      "learning_rate": 6.887095978763072e-06,
+      "loss": 0.1815,
+      "step": 2645
+    },
+    {
+      "epoch": 0.39569313593539707,
+      "grad_norm": 1.751476920569849,
+      "learning_rate": 6.884853043654876e-06,
+      "loss": 0.3909,
+      "step": 2646
+    },
+    {
+      "epoch": 0.3958426798265291,
+      "grad_norm": 1.5414130565030935,
+      "learning_rate": 6.882609666341988e-06,
+      "loss": 0.383,
+      "step": 2647
+    },
+    {
+      "epoch": 0.39599222371766113,
+      "grad_norm": 1.7873679635826936,
+      "learning_rate": 6.88036584735073e-06,
+      "loss": 0.5987,
+      "step": 2648
+    },
+    {
+      "epoch": 0.39614176760879316,
+      "grad_norm": 1.4503283763937924,
+      "learning_rate": 6.878121587207522e-06,
+      "loss": 0.1782,
+      "step": 2649
+    },
+    {
+      "epoch": 0.39629131149992525,
+      "grad_norm": 1.1935592615073942,
+      "learning_rate": 6.875876886438889e-06,
+      "loss": 0.3002,
+      "step": 2650
+    },
+    {
+      "epoch": 0.3964408553910573,
+      "grad_norm": 1.471497098484429,
+      "learning_rate": 6.873631745571461e-06,
+      "loss": 0.2105,
+      "step": 2651
+    },
+    {
+      "epoch": 0.3965903992821893,
+      "grad_norm": 1.7129740678391,
+      "learning_rate": 6.871386165131968e-06,
+      "loss": 0.2233,
+      "step": 2652
+    },
+    {
+      "epoch": 0.3967399431733214,
+      "grad_norm": 1.5756431198477898,
+      "learning_rate": 6.869140145647245e-06,
+      "loss": 0.4151,
+      "step": 2653
+    },
+    {
+      "epoch": 0.39688948706445343,
+      "grad_norm": 1.238820133220739,
+      "learning_rate": 6.866893687644232e-06,
+      "loss": 0.2,
+      "step": 2654
+    },
+    {
+      "epoch": 0.39703903095558546,
+      "grad_norm": 1.9264407972582287,
+      "learning_rate": 6.864646791649966e-06,
+      "loss": 0.4677,
+      "step": 2655
+    },
+    {
+      "epoch": 0.3971885748467175,
+      "grad_norm": 1.550112419127813,
+      "learning_rate": 6.862399458191593e-06,
+      "loss": 0.3658,
+      "step": 2656
+    },
+    {
+      "epoch": 0.3973381187378496,
+      "grad_norm": 1.5119124813290858,
+      "learning_rate": 6.860151687796359e-06,
+      "loss": 0.2068,
+      "step": 2657
+    },
+    {
+      "epoch": 0.3974876626289816,
+      "grad_norm": 1.4272874845074974,
+      "learning_rate": 6.857903480991611e-06,
+      "loss": 0.2713,
+      "step": 2658
+    },
+    {
+      "epoch": 0.39763720652011364,
+      "grad_norm": 1.6594720979073507,
+      "learning_rate": 6.855654838304802e-06,
+      "loss": 0.3205,
+      "step": 2659
+    },
+    {
+      "epoch": 0.3977867504112457,
+      "grad_norm": 2.407352823346937,
+      "learning_rate": 6.853405760263485e-06,
+      "loss": 0.2088,
+      "step": 2660
+    },
+    {
+      "epoch": 0.39793629430237776,
+      "grad_norm": 1.7959792599685889,
+      "learning_rate": 6.851156247395313e-06,
+      "loss": 0.393,
+      "step": 2661
+    },
+    {
+      "epoch": 0.3980858381935098,
+      "grad_norm": 1.4851282768283998,
+      "learning_rate": 6.848906300228047e-06,
+      "loss": 0.2354,
+      "step": 2662
+    },
+    {
+      "epoch": 0.3982353820846418,
+      "grad_norm": 1.4691880516424127,
+      "learning_rate": 6.846655919289543e-06,
+      "loss": 0.231,
+      "step": 2663
+    },
+    {
+      "epoch": 0.3983849259757739,
+      "grad_norm": 1.5828677803244833,
+      "learning_rate": 6.844405105107763e-06,
+      "loss": 0.1887,
+      "step": 2664
+    },
+    {
+      "epoch": 0.39853446986690594,
+      "grad_norm": 1.5541799378105448,
+      "learning_rate": 6.842153858210772e-06,
+      "loss": 0.3784,
+      "step": 2665
+    },
+    {
+      "epoch": 0.39868401375803797,
+      "grad_norm": 1.9024949023533122,
+      "learning_rate": 6.83990217912673e-06,
+      "loss": 0.4826,
+      "step": 2666
+    },
+    {
+      "epoch": 0.39883355764917006,
+      "grad_norm": 1.7789769644036557,
+      "learning_rate": 6.837650068383908e-06,
+      "loss": 0.4536,
+      "step": 2667
+    },
+    {
+      "epoch": 0.3989831015403021,
+      "grad_norm": 1.6091866012593252,
+      "learning_rate": 6.835397526510667e-06,
+      "loss": 0.4931,
+      "step": 2668
+    },
+    {
+      "epoch": 0.3991326454314341,
+      "grad_norm": 1.4778294602532251,
+      "learning_rate": 6.83314455403548e-06,
+      "loss": 0.3916,
+      "step": 2669
+    },
+    {
+      "epoch": 0.39928218932256615,
+      "grad_norm": 1.4396363740839195,
+      "learning_rate": 6.8308911514869125e-06,
+      "loss": 0.2875,
+      "step": 2670
+    },
+    {
+      "epoch": 0.39943173321369824,
+      "grad_norm": 1.5237349879970323,
+      "learning_rate": 6.828637319393636e-06,
+      "loss": 0.2372,
+      "step": 2671
+    },
+    {
+      "epoch": 0.39958127710483027,
+      "grad_norm": 1.4821885218044735,
+      "learning_rate": 6.826383058284421e-06,
+      "loss": 0.3351,
+      "step": 2672
+    },
+    {
+      "epoch": 0.3997308209959623,
+      "grad_norm": 1.6760124989546532,
+      "learning_rate": 6.824128368688139e-06,
+      "loss": 0.3599,
+      "step": 2673
+    },
+    {
+      "epoch": 0.3998803648870944,
+      "grad_norm": 1.0076444138222016,
+      "learning_rate": 6.821873251133764e-06,
+      "loss": 0.1925,
+      "step": 2674
+    },
+    {
+      "epoch": 0.4000299087782264,
+      "grad_norm": 1.0371695095164526,
+      "learning_rate": 6.819617706150363e-06,
+      "loss": 0.177,
+      "step": 2675
+    },
+    {
+      "epoch": 0.40017945266935845,
+      "grad_norm": 1.2181986301270449,
+      "learning_rate": 6.817361734267114e-06,
+      "loss": 0.2938,
+      "step": 2676
+    },
+    {
+      "epoch": 0.4003289965604905,
+      "grad_norm": 1.4070959430963377,
+      "learning_rate": 6.815105336013287e-06,
+      "loss": 0.1739,
+      "step": 2677
+    },
+    {
+      "epoch": 0.40047854045162257,
+      "grad_norm": 2.0098290688275737,
+      "learning_rate": 6.812848511918254e-06,
+      "loss": 0.3152,
+      "step": 2678
+    },
+    {
+      "epoch": 0.4006280843427546,
+      "grad_norm": 1.543663585518377,
+      "learning_rate": 6.810591262511492e-06,
+      "loss": 0.3398,
+      "step": 2679
+    },
+    {
+      "epoch": 0.40077762823388663,
+      "grad_norm": 1.231130815440152,
+      "learning_rate": 6.808333588322569e-06,
+      "loss": 0.2017,
+      "step": 2680
+    },
+    {
+      "epoch": 0.4009271721250187,
+      "grad_norm": 1.8122501869118024,
+      "learning_rate": 6.806075489881158e-06,
+      "loss": 0.1972,
+      "step": 2681
+    },
+    {
+      "epoch": 0.40107671601615075,
+      "grad_norm": 1.7484305402942781,
+      "learning_rate": 6.803816967717031e-06,
+      "loss": 0.4729,
+      "step": 2682
+    },
+    {
+      "epoch": 0.4012262599072828,
+      "grad_norm": 1.3000952584298922,
+      "learning_rate": 6.801558022360061e-06,
+      "loss": 0.2636,
+      "step": 2683
+    },
+    {
+      "epoch": 0.4013758037984148,
+      "grad_norm": 1.8438350811417317,
+      "learning_rate": 6.799298654340215e-06,
+      "loss": 0.5354,
+      "step": 2684
+    },
+    {
+      "epoch": 0.4015253476895469,
+      "grad_norm": 2.2848085970399303,
+      "learning_rate": 6.797038864187564e-06,
+      "loss": 0.8736,
+      "step": 2685
+    },
+    {
+      "epoch": 0.4016748915806789,
+      "grad_norm": 1.4866529052233595,
+      "learning_rate": 6.794778652432278e-06,
+      "loss": 0.2244,
+      "step": 2686
+    },
+    {
+      "epoch": 0.40182443547181096,
+      "grad_norm": 1.2521236529684348,
+      "learning_rate": 6.792518019604624e-06,
+      "loss": 0.2239,
+      "step": 2687
+    },
+    {
+      "epoch": 0.40197397936294305,
+      "grad_norm": 1.576212342415195,
+      "learning_rate": 6.790256966234966e-06,
+      "loss": 0.2802,
+      "step": 2688
+    },
+    {
+      "epoch": 0.4021235232540751,
+      "grad_norm": 1.2579155325895004,
+      "learning_rate": 6.787995492853771e-06,
+      "loss": 0.2094,
+      "step": 2689
+    },
+    {
+      "epoch": 0.4022730671452071,
+      "grad_norm": 1.4213546092771336,
+      "learning_rate": 6.785733599991602e-06,
+      "loss": 0.1914,
+      "step": 2690
+    },
+    {
+      "epoch": 0.40242261103633914,
+      "grad_norm": 0.8485344894569877,
+      "learning_rate": 6.78347128817912e-06,
+      "loss": 0.224,
+      "step": 2691
+    },
+    {
+      "epoch": 0.4025721549274712,
+      "grad_norm": 1.6595476897588828,
+      "learning_rate": 6.781208557947085e-06,
+      "loss": 0.3873,
+      "step": 2692
+    },
+    {
+      "epoch": 0.40272169881860326,
+      "grad_norm": 1.0569857811012024,
+      "learning_rate": 6.778945409826356e-06,
+      "loss": 0.2516,
+      "step": 2693
+    },
+    {
+      "epoch": 0.4028712427097353,
+      "grad_norm": 1.2314194821001476,
+      "learning_rate": 6.776681844347892e-06,
+      "loss": 0.2237,
+      "step": 2694
+    },
+    {
+      "epoch": 0.4030207866008674,
+      "grad_norm": 0.9731617862379426,
+      "learning_rate": 6.7744178620427435e-06,
+      "loss": 0.2348,
+      "step": 2695
+    },
+    {
+      "epoch": 0.4031703304919994,
+      "grad_norm": 1.5995980706007595,
+      "learning_rate": 6.7721534634420635e-06,
+      "loss": 0.2542,
+      "step": 2696
+    },
+    {
+      "epoch": 0.40331987438313144,
+      "grad_norm": 1.788090656384711,
+      "learning_rate": 6.769888649077103e-06,
+      "loss": 0.2238,
+      "step": 2697
+    },
+    {
+      "epoch": 0.40346941827426347,
+      "grad_norm": 1.6873214415283535,
+      "learning_rate": 6.767623419479207e-06,
+      "loss": 0.4441,
+      "step": 2698
+    },
+    {
+      "epoch": 0.40361896216539556,
+      "grad_norm": 1.7639180599847581,
+      "learning_rate": 6.765357775179822e-06,
+      "loss": 0.4911,
+      "step": 2699
+    },
+    {
+      "epoch": 0.4037685060565276,
+      "grad_norm": 1.3115613907742758,
+      "learning_rate": 6.76309171671049e-06,
+      "loss": 0.3307,
+      "step": 2700
+    },
+    {
+      "epoch": 0.4039180499476596,
+      "grad_norm": 1.383566483800298,
+      "learning_rate": 6.760825244602849e-06,
+      "loss": 0.2378,
+      "step": 2701
+    },
+    {
+      "epoch": 0.4040675938387917,
+      "grad_norm": 1.0532366239269708,
+      "learning_rate": 6.7585583593886376e-06,
+      "loss": 0.3177,
+      "step": 2702
+    },
+    {
+      "epoch": 0.40421713772992374,
+      "grad_norm": 1.7117686930778293,
+      "learning_rate": 6.7562910615996854e-06,
+      "loss": 0.3372,
+      "step": 2703
+    },
+    {
+      "epoch": 0.40436668162105577,
+      "grad_norm": 1.59208336340781,
+      "learning_rate": 6.754023351767924e-06,
+      "loss": 0.2335,
+      "step": 2704
+    },
+    {
+      "epoch": 0.40451622551218785,
+      "grad_norm": 2.307120871555123,
+      "learning_rate": 6.7517552304253815e-06,
+      "loss": 0.4416,
+      "step": 2705
+    },
+    {
+      "epoch": 0.4046657694033199,
+      "grad_norm": 2.4127825210857075,
+      "learning_rate": 6.7494866981041784e-06,
+      "loss": 0.7162,
+      "step": 2706
+    },
+    {
+      "epoch": 0.4048153132944519,
+      "grad_norm": 1.768470281885931,
+      "learning_rate": 6.747217755336537e-06,
+      "loss": 0.3867,
+      "step": 2707
+    },
+    {
+      "epoch": 0.40496485718558395,
+      "grad_norm": 0.9575195254141551,
+      "learning_rate": 6.7449484026547705e-06,
+      "loss": 0.2125,
+      "step": 2708
+    },
+    {
+      "epoch": 0.40511440107671604,
+      "grad_norm": 1.2201809246818296,
+      "learning_rate": 6.742678640591293e-06,
+      "loss": 0.1848,
+      "step": 2709
+    },
+    {
+      "epoch": 0.40526394496784807,
+      "grad_norm": 1.022034673727343,
+      "learning_rate": 6.740408469678611e-06,
+      "loss": 0.1824,
+      "step": 2710
+    },
+    {
+      "epoch": 0.4054134888589801,
+      "grad_norm": 1.484686409132478,
+      "learning_rate": 6.738137890449329e-06,
+      "loss": 0.2005,
+      "step": 2711
+    },
+    {
+      "epoch": 0.4055630327501122,
+      "grad_norm": 1.5990317632752664,
+      "learning_rate": 6.735866903436146e-06,
+      "loss": 0.2366,
+      "step": 2712
+    },
+    {
+      "epoch": 0.4057125766412442,
+      "grad_norm": 0.840556680462754,
+      "learning_rate": 6.7335955091718595e-06,
+      "loss": 0.1896,
+      "step": 2713
+    },
+    {
+      "epoch": 0.40586212053237625,
+      "grad_norm": 1.9574024358578521,
+      "learning_rate": 6.731323708189358e-06,
+      "loss": 0.5026,
+      "step": 2714
+    },
+    {
+      "epoch": 0.4060116644235083,
+      "grad_norm": 1.0372558946433956,
+      "learning_rate": 6.7290515010216305e-06,
+      "loss": 0.2117,
+      "step": 2715
+    },
+    {
+      "epoch": 0.40616120831464037,
+      "grad_norm": 1.782559024684888,
+      "learning_rate": 6.726778888201757e-06,
+      "loss": 0.401,
+      "step": 2716
+    },
+    {
+      "epoch": 0.4063107522057724,
+      "grad_norm": 1.8957139836111772,
+      "learning_rate": 6.724505870262915e-06,
+      "loss": 0.454,
+      "step": 2717
+    },
+    {
+      "epoch": 0.4064602960969044,
+      "grad_norm": 0.9411673381180247,
+      "learning_rate": 6.722232447738375e-06,
+      "loss": 0.1434,
+      "step": 2718
+    },
+    {
+      "epoch": 0.4066098399880365,
+      "grad_norm": 1.528883853888666,
+      "learning_rate": 6.719958621161505e-06,
+      "loss": 0.2069,
+      "step": 2719
+    },
+    {
+      "epoch": 0.40675938387916855,
+      "grad_norm": 1.0474505826049523,
+      "learning_rate": 6.717684391065769e-06,
+      "loss": 0.2319,
+      "step": 2720
+    },
+    {
+      "epoch": 0.4069089277703006,
+      "grad_norm": 1.608085985877073,
+      "learning_rate": 6.715409757984718e-06,
+      "loss": 0.2077,
+      "step": 2721
+    },
+    {
+      "epoch": 0.4070584716614326,
+      "grad_norm": 1.213620862112801,
+      "learning_rate": 6.713134722452009e-06,
+      "loss": 0.2165,
+      "step": 2722
+    },
+    {
+      "epoch": 0.4072080155525647,
+      "grad_norm": 1.4959997727780725,
+      "learning_rate": 6.710859285001381e-06,
+      "loss": 0.2631,
+      "step": 2723
+    },
+    {
+      "epoch": 0.4073575594436967,
+      "grad_norm": 1.4898775205900512,
+      "learning_rate": 6.7085834461666765e-06,
+      "loss": 0.3716,
+      "step": 2724
+    },
+    {
+      "epoch": 0.40750710333482876,
+      "grad_norm": 2.940723192965302,
+      "learning_rate": 6.70630720648183e-06,
+      "loss": 0.2594,
+      "step": 2725
+    },
+    {
+      "epoch": 0.40765664722596084,
+      "grad_norm": 1.4132307468761955,
+      "learning_rate": 6.704030566480869e-06,
+      "loss": 0.174,
+      "step": 2726
+    },
+    {
+      "epoch": 0.4078061911170929,
+      "grad_norm": 1.1415584187313277,
+      "learning_rate": 6.701753526697915e-06,
+      "loss": 0.2006,
+      "step": 2727
+    },
+    {
+      "epoch": 0.4079557350082249,
+      "grad_norm": 1.4316392497827604,
+      "learning_rate": 6.699476087667183e-06,
+      "loss": 0.1804,
+      "step": 2728
+    },
+    {
+      "epoch": 0.40810527889935694,
+      "grad_norm": 1.828481566061616,
+      "learning_rate": 6.697198249922981e-06,
+      "loss": 0.3837,
+      "step": 2729
+    },
+    {
+      "epoch": 0.408254822790489,
+      "grad_norm": 1.5937472090977185,
+      "learning_rate": 6.694920013999715e-06,
+      "loss": 0.4825,
+      "step": 2730
+    },
+    {
+      "epoch": 0.40840436668162106,
+      "grad_norm": 1.1850332784758328,
+      "learning_rate": 6.692641380431879e-06,
+      "loss": 0.2363,
+      "step": 2731
+    },
+    {
+      "epoch": 0.4085539105727531,
+      "grad_norm": 1.3868683747713086,
+      "learning_rate": 6.690362349754062e-06,
+      "loss": 0.2932,
+      "step": 2732
+    },
+    {
+      "epoch": 0.4087034544638852,
+      "grad_norm": 2.021537411496191,
+      "learning_rate": 6.6880829225009455e-06,
+      "loss": 0.6287,
+      "step": 2733
+    },
+    {
+      "epoch": 0.4088529983550172,
+      "grad_norm": 1.1157967473978323,
+      "learning_rate": 6.685803099207309e-06,
+      "loss": 0.2041,
+      "step": 2734
+    },
+    {
+      "epoch": 0.40900254224614924,
+      "grad_norm": 1.5873384207624706,
+      "learning_rate": 6.683522880408019e-06,
+      "loss": 0.3283,
+      "step": 2735
+    },
+    {
+      "epoch": 0.40915208613728127,
+      "grad_norm": 1.7562023141944056,
+      "learning_rate": 6.681242266638035e-06,
+      "loss": 0.2274,
+      "step": 2736
+    },
+    {
+      "epoch": 0.40930163002841335,
+      "grad_norm": 2.768807471305314,
+      "learning_rate": 6.6789612584324135e-06,
+      "loss": 0.5855,
+      "step": 2737
+    },
+    {
+      "epoch": 0.4094511739195454,
+      "grad_norm": 1.5373476587783403,
+      "learning_rate": 6.676679856326298e-06,
+      "loss": 0.3774,
+      "step": 2738
+    },
+    {
+      "epoch": 0.4096007178106774,
+      "grad_norm": 1.3129213939580835,
+      "learning_rate": 6.674398060854931e-06,
+      "loss": 0.3354,
+      "step": 2739
+    },
+    {
+      "epoch": 0.4097502617018095,
+      "grad_norm": 1.9542337513011827,
+      "learning_rate": 6.672115872553643e-06,
+      "loss": 0.6155,
+      "step": 2740
+    },
+    {
+      "epoch": 0.40989980559294154,
+      "grad_norm": 1.4597570780931846,
+      "learning_rate": 6.669833291957854e-06,
+      "loss": 0.1852,
+      "step": 2741
+    },
+    {
+      "epoch": 0.41004934948407357,
+      "grad_norm": 1.4488787882100997,
+      "learning_rate": 6.667550319603084e-06,
+      "loss": 0.2032,
+      "step": 2742
+    },
+    {
+      "epoch": 0.4101988933752056,
+      "grad_norm": 1.2741107506652722,
+      "learning_rate": 6.665266956024939e-06,
+      "loss": 0.1551,
+      "step": 2743
+    },
+    {
+      "epoch": 0.4103484372663377,
+      "grad_norm": 1.4577006420793812,
+      "learning_rate": 6.662983201759116e-06,
+      "loss": 0.2233,
+      "step": 2744
+    },
+    {
+      "epoch": 0.4104979811574697,
+      "grad_norm": 1.654731508737811,
+      "learning_rate": 6.660699057341407e-06,
+      "loss": 0.163,
+      "step": 2745
+    },
+    {
+      "epoch": 0.41064752504860175,
+      "grad_norm": 1.6694252301810233,
+      "learning_rate": 6.658414523307694e-06,
+      "loss": 0.361,
+      "step": 2746
+    },
+    {
+      "epoch": 0.41079706893973383,
+      "grad_norm": 1.7611241278128056,
+      "learning_rate": 6.656129600193951e-06,
+      "loss": 0.3767,
+      "step": 2747
+    },
+    {
+      "epoch": 0.41094661283086587,
+      "grad_norm": 1.5219042887796945,
+      "learning_rate": 6.653844288536244e-06,
+      "loss": 0.2821,
+      "step": 2748
+    },
+    {
+      "epoch": 0.4110961567219979,
+      "grad_norm": 1.432463624052981,
+      "learning_rate": 6.651558588870728e-06,
+      "loss": 0.278,
+      "step": 2749
+    },
+    {
+      "epoch": 0.4112457006131299,
+      "grad_norm": 1.6511294342102478,
+      "learning_rate": 6.649272501733651e-06,
+      "loss": 0.3371,
+      "step": 2750
+    },
+    {
+      "epoch": 0.411395244504262,
+      "grad_norm": 2.222481645319511,
+      "learning_rate": 6.646986027661351e-06,
+      "loss": 0.8056,
+      "step": 2751
+    },
+    {
+      "epoch": 0.41154478839539405,
+      "grad_norm": 1.7419918403749177,
+      "learning_rate": 6.644699167190254e-06,
+      "loss": 0.375,
+      "step": 2752
+    },
+    {
+      "epoch": 0.4116943322865261,
+      "grad_norm": 1.9346738438100417,
+      "learning_rate": 6.642411920856882e-06,
+      "loss": 0.5479,
+      "step": 2753
+    },
+    {
+      "epoch": 0.41184387617765816,
+      "grad_norm": 0.9808333961270187,
+      "learning_rate": 6.640124289197845e-06,
+      "loss": 0.1877,
+      "step": 2754
+    },
+    {
+      "epoch": 0.4119934200687902,
+      "grad_norm": 2.0713908859757018,
+      "learning_rate": 6.637836272749846e-06,
+      "loss": 0.5199,
+      "step": 2755
+    },
+    {
+      "epoch": 0.4121429639599222,
+      "grad_norm": 1.4679970421838424,
+      "learning_rate": 6.63554787204967e-06,
+      "loss": 0.1927,
+      "step": 2756
+    },
+    {
+      "epoch": 0.41229250785105426,
+      "grad_norm": 1.1272956717609233,
+      "learning_rate": 6.6332590876342015e-06,
+      "loss": 0.2071,
+      "step": 2757
+    },
+    {
+      "epoch": 0.41244205174218634,
+      "grad_norm": 2.53543503085162,
+      "learning_rate": 6.630969920040411e-06,
+      "loss": 0.3872,
+      "step": 2758
+    },
+    {
+      "epoch": 0.4125915956333184,
+      "grad_norm": 1.7338236929187498,
+      "learning_rate": 6.628680369805359e-06,
+      "loss": 0.4355,
+      "step": 2759
+    },
+    {
+      "epoch": 0.4127411395244504,
+      "grad_norm": 1.3212066026939129,
+      "learning_rate": 6.626390437466197e-06,
+      "loss": 0.1943,
+      "step": 2760
+    },
+    {
+      "epoch": 0.4128906834155825,
+      "grad_norm": 1.7339371416327845,
+      "learning_rate": 6.624100123560164e-06,
+      "loss": 0.2408,
+      "step": 2761
+    },
+    {
+      "epoch": 0.4130402273067145,
+      "grad_norm": 1.4122438327200852,
+      "learning_rate": 6.621809428624588e-06,
+      "loss": 0.3316,
+      "step": 2762
+    },
+    {
+      "epoch": 0.41318977119784656,
+      "grad_norm": 1.378447575239376,
+      "learning_rate": 6.619518353196892e-06,
+      "loss": 0.1618,
+      "step": 2763
+    },
+    {
+      "epoch": 0.41333931508897864,
+      "grad_norm": 1.66925577966763,
+      "learning_rate": 6.617226897814582e-06,
+      "loss": 0.24,
+      "step": 2764
+    },
+    {
+      "epoch": 0.4134888589801107,
+      "grad_norm": 1.688913023066291,
+      "learning_rate": 6.614935063015256e-06,
+      "loss": 0.1803,
+      "step": 2765
+    },
+    {
+      "epoch": 0.4136384028712427,
+      "grad_norm": 1.5610098337094958,
+      "learning_rate": 6.612642849336599e-06,
+      "loss": 0.2558,
+      "step": 2766
+    },
+    {
+      "epoch": 0.41378794676237474,
+      "grad_norm": 1.727757709094625,
+      "learning_rate": 6.610350257316389e-06,
+      "loss": 0.3811,
+      "step": 2767
+    },
+    {
+      "epoch": 0.4139374906535068,
+      "grad_norm": 1.6489611874055852,
+      "learning_rate": 6.608057287492491e-06,
+      "loss": 0.4092,
+      "step": 2768
+    },
+    {
+      "epoch": 0.41408703454463885,
+      "grad_norm": 2.3320480486649755,
+      "learning_rate": 6.605763940402854e-06,
+      "loss": 0.6742,
+      "step": 2769
+    },
+    {
+      "epoch": 0.4142365784357709,
+      "grad_norm": 1.4200969362500393,
+      "learning_rate": 6.6034702165855225e-06,
+      "loss": 0.2639,
+      "step": 2770
+    },
+    {
+      "epoch": 0.414386122326903,
+      "grad_norm": 0.9909541939147573,
+      "learning_rate": 6.601176116578624e-06,
+      "loss": 0.1167,
+      "step": 2771
+    },
+    {
+      "epoch": 0.414535666218035,
+      "grad_norm": 1.673456689402228,
+      "learning_rate": 6.598881640920376e-06,
+      "loss": 0.3768,
+      "step": 2772
+    },
+    {
+      "epoch": 0.41468521010916704,
+      "grad_norm": 1.0737793994976543,
+      "learning_rate": 6.596586790149087e-06,
+      "loss": 0.2302,
+      "step": 2773
+    },
+    {
+      "epoch": 0.41483475400029907,
+      "grad_norm": 1.7858247736400699,
+      "learning_rate": 6.5942915648031516e-06,
+      "loss": 0.5081,
+      "step": 2774
+    },
+    {
+      "epoch": 0.41498429789143115,
+      "grad_norm": 1.7377180708768736,
+      "learning_rate": 6.591995965421049e-06,
+      "loss": 0.3567,
+      "step": 2775
+    },
+    {
+      "epoch": 0.4151338417825632,
+      "grad_norm": 1.729058819350988,
+      "learning_rate": 6.58969999254135e-06,
+      "loss": 0.3563,
+      "step": 2776
+    },
+    {
+      "epoch": 0.4152833856736952,
+      "grad_norm": 2.521334430186825,
+      "learning_rate": 6.5874036467027135e-06,
+      "loss": 0.8416,
+      "step": 2777
+    },
+    {
+      "epoch": 0.4154329295648273,
+      "grad_norm": 1.5115105389474632,
+      "learning_rate": 6.585106928443883e-06,
+      "loss": 0.3629,
+      "step": 2778
+    },
+    {
+      "epoch": 0.41558247345595933,
+      "grad_norm": 0.9562243758146778,
+      "learning_rate": 6.5828098383036895e-06,
+      "loss": 0.134,
+      "step": 2779
+    },
+    {
+      "epoch": 0.41573201734709136,
+      "grad_norm": 1.50030075040931,
+      "learning_rate": 6.580512376821055e-06,
+      "loss": 0.2136,
+      "step": 2780
+    },
+    {
+      "epoch": 0.4158815612382234,
+      "grad_norm": 1.3200898857471022,
+      "learning_rate": 6.5782145445349845e-06,
+      "loss": 0.3365,
+      "step": 2781
+    },
+    {
+      "epoch": 0.4160311051293555,
+      "grad_norm": 1.827699208191328,
+      "learning_rate": 6.5759163419845715e-06,
+      "loss": 0.4035,
+      "step": 2782
+    },
+    {
+      "epoch": 0.4161806490204875,
+      "grad_norm": 1.713119928633606,
+      "learning_rate": 6.573617769708998e-06,
+      "loss": 0.2414,
+      "step": 2783
+    },
+    {
+      "epoch": 0.41633019291161955,
+      "grad_norm": 1.46958562232206,
+      "learning_rate": 6.57131882824753e-06,
+      "loss": 0.2587,
+      "step": 2784
+    },
+    {
+      "epoch": 0.41647973680275163,
+      "grad_norm": 1.5831830835473946,
+      "learning_rate": 6.569019518139521e-06,
+      "loss": 0.3713,
+      "step": 2785
+    },
+    {
+      "epoch": 0.41662928069388366,
+      "grad_norm": 1.6336655738252617,
+      "learning_rate": 6.566719839924412e-06,
+      "loss": 0.2954,
+      "step": 2786
+    },
+    {
+      "epoch": 0.4167788245850157,
+      "grad_norm": 1.7291309360657896,
+      "learning_rate": 6.564419794141729e-06,
+      "loss": 0.4808,
+      "step": 2787
+    },
+    {
+      "epoch": 0.4169283684761477,
+      "grad_norm": 1.1926204169727774,
+      "learning_rate": 6.562119381331087e-06,
+      "loss": 0.1963,
+      "step": 2788
+    },
+    {
+      "epoch": 0.4170779123672798,
+      "grad_norm": 1.0629560827548779,
+      "learning_rate": 6.559818602032182e-06,
+      "loss": 0.1881,
+      "step": 2789
+    },
+    {
+      "epoch": 0.41722745625841184,
+      "grad_norm": 1.7790857371942344,
+      "learning_rate": 6.557517456784801e-06,
+      "loss": 0.1946,
+      "step": 2790
+    },
+    {
+      "epoch": 0.4173770001495439,
+      "grad_norm": 1.226190910118689,
+      "learning_rate": 6.555215946128815e-06,
+      "loss": 0.2162,
+      "step": 2791
+    },
+    {
+      "epoch": 0.41752654404067596,
+      "grad_norm": 2.1109476161237204,
+      "learning_rate": 6.552914070604178e-06,
+      "loss": 0.5731,
+      "step": 2792
+    },
+    {
+      "epoch": 0.417676087931808,
+      "grad_norm": 1.7790077510552844,
+      "learning_rate": 6.550611830750935e-06,
+      "loss": 0.3273,
+      "step": 2793
+    },
+    {
+      "epoch": 0.41782563182294,
+      "grad_norm": 1.7552104298119706,
+      "learning_rate": 6.5483092271092115e-06,
+      "loss": 0.4436,
+      "step": 2794
+    },
+    {
+      "epoch": 0.41797517571407206,
+      "grad_norm": 1.7002774108662209,
+      "learning_rate": 6.546006260219223e-06,
+      "loss": 0.3526,
+      "step": 2795
+    },
+    {
+      "epoch": 0.41812471960520414,
+      "grad_norm": 1.4233636661060545,
+      "learning_rate": 6.543702930621267e-06,
+      "loss": 0.3524,
+      "step": 2796
+    },
+    {
+      "epoch": 0.4182742634963362,
+      "grad_norm": 1.8156591047803354,
+      "learning_rate": 6.541399238855726e-06,
+      "loss": 0.3947,
+      "step": 2797
+    },
+    {
+      "epoch": 0.4184238073874682,
+      "grad_norm": 1.2300331934296698,
+      "learning_rate": 6.539095185463069e-06,
+      "loss": 0.2044,
+      "step": 2798
+    },
+    {
+      "epoch": 0.4185733512786003,
+      "grad_norm": 1.2209879703063857,
+      "learning_rate": 6.5367907709838466e-06,
+      "loss": 0.1635,
+      "step": 2799
+    },
+    {
+      "epoch": 0.4187228951697323,
+      "grad_norm": 2.3744121752643044,
+      "learning_rate": 6.534485995958699e-06,
+      "loss": 0.7958,
+      "step": 2800
+    },
+    {
+      "epoch": 0.41887243906086435,
+      "grad_norm": 1.1638106723385093,
+      "learning_rate": 6.53218086092835e-06,
+      "loss": 0.3233,
+      "step": 2801
+    },
+    {
+      "epoch": 0.4190219829519964,
+      "grad_norm": 2.2950286840379124,
+      "learning_rate": 6.529875366433604e-06,
+      "loss": 0.7808,
+      "step": 2802
+    },
+    {
+      "epoch": 0.4191715268431285,
+      "grad_norm": 1.402864714048257,
+      "learning_rate": 6.527569513015354e-06,
+      "loss": 0.2022,
+      "step": 2803
+    },
+    {
+      "epoch": 0.4193210707342605,
+      "grad_norm": 1.4371707658442519,
+      "learning_rate": 6.5252633012145735e-06,
+      "loss": 0.187,
+      "step": 2804
+    },
+    {
+      "epoch": 0.41947061462539253,
+      "grad_norm": 1.2367409752013236,
+      "learning_rate": 6.522956731572323e-06,
+      "loss": 0.1904,
+      "step": 2805
+    },
+    {
+      "epoch": 0.4196201585165246,
+      "grad_norm": 1.5896821279643083,
+      "learning_rate": 6.520649804629746e-06,
+      "loss": 0.3862,
+      "step": 2806
+    },
+    {
+      "epoch": 0.41976970240765665,
+      "grad_norm": 1.8899618435974137,
+      "learning_rate": 6.5183425209280694e-06,
+      "loss": 0.3903,
+      "step": 2807
+    },
+    {
+      "epoch": 0.4199192462987887,
+      "grad_norm": 2.0562104756171364,
+      "learning_rate": 6.5160348810086065e-06,
+      "loss": 0.5683,
+      "step": 2808
+    },
+    {
+      "epoch": 0.4200687901899207,
+      "grad_norm": 0.9348853804908052,
+      "learning_rate": 6.513726885412748e-06,
+      "loss": 0.2017,
+      "step": 2809
+    },
+    {
+      "epoch": 0.4202183340810528,
+      "grad_norm": 1.8140665018134454,
+      "learning_rate": 6.511418534681974e-06,
+      "loss": 0.3551,
+      "step": 2810
+    },
+    {
+      "epoch": 0.42036787797218483,
+      "grad_norm": 1.8148951879922504,
+      "learning_rate": 6.509109829357847e-06,
+      "loss": 0.2311,
+      "step": 2811
+    },
+    {
+      "epoch": 0.42051742186331686,
+      "grad_norm": 1.2263992513762383,
+      "learning_rate": 6.506800769982009e-06,
+      "loss": 0.1839,
+      "step": 2812
+    },
+    {
+      "epoch": 0.42066696575444895,
+      "grad_norm": 1.405409781791798,
+      "learning_rate": 6.504491357096189e-06,
+      "loss": 0.2004,
+      "step": 2813
+    },
+    {
+      "epoch": 0.420816509645581,
+      "grad_norm": 1.280743321393158,
+      "learning_rate": 6.5021815912421955e-06,
+      "loss": 0.3081,
+      "step": 2814
+    },
+    {
+      "epoch": 0.420966053536713,
+      "grad_norm": 1.2542595826019205,
+      "learning_rate": 6.499871472961926e-06,
+      "loss": 0.1564,
+      "step": 2815
+    },
+    {
+      "epoch": 0.42111559742784505,
+      "grad_norm": 2.473171445619144,
+      "learning_rate": 6.4975610027973545e-06,
+      "loss": 0.4957,
+      "step": 2816
+    },
+    {
+      "epoch": 0.42126514131897713,
+      "grad_norm": 1.6907310283363046,
+      "learning_rate": 6.495250181290538e-06,
+      "loss": 0.5282,
+      "step": 2817
+    },
+    {
+      "epoch": 0.42141468521010916,
+      "grad_norm": 1.6101555551733806,
+      "learning_rate": 6.49293900898362e-06,
+      "loss": 0.4522,
+      "step": 2818
+    },
+    {
+      "epoch": 0.4215642291012412,
+      "grad_norm": 1.9979956071229465,
+      "learning_rate": 6.49062748641882e-06,
+      "loss": 0.3861,
+      "step": 2819
+    },
+    {
+      "epoch": 0.4217137729923733,
+      "grad_norm": 1.1590188886249708,
+      "learning_rate": 6.488315614138447e-06,
+      "loss": 0.2171,
+      "step": 2820
+    },
+    {
+      "epoch": 0.4218633168835053,
+      "grad_norm": 1.9821881424242638,
+      "learning_rate": 6.486003392684889e-06,
+      "loss": 0.2409,
+      "step": 2821
+    },
+    {
+      "epoch": 0.42201286077463734,
+      "grad_norm": 1.876013420334667,
+      "learning_rate": 6.483690822600613e-06,
+      "loss": 0.3573,
+      "step": 2822
+    },
+    {
+      "epoch": 0.42216240466576943,
+      "grad_norm": 1.813581563264969,
+      "learning_rate": 6.481377904428171e-06,
+      "loss": 0.4781,
+      "step": 2823
+    },
+    {
+      "epoch": 0.42231194855690146,
+      "grad_norm": 2.0083488534654808,
+      "learning_rate": 6.479064638710197e-06,
+      "loss": 0.2441,
+      "step": 2824
+    },
+    {
+      "epoch": 0.4224614924480335,
+      "grad_norm": 1.3677082786526151,
+      "learning_rate": 6.476751025989403e-06,
+      "loss": 0.3431,
+      "step": 2825
+    },
+    {
+      "epoch": 0.4226110363391655,
+      "grad_norm": 1.0145275734224242,
+      "learning_rate": 6.474437066808585e-06,
+      "loss": 0.1819,
+      "step": 2826
+    },
+    {
+      "epoch": 0.4227605802302976,
+      "grad_norm": 1.490785561817132,
+      "learning_rate": 6.472122761710624e-06,
+      "loss": 0.3551,
+      "step": 2827
+    },
+    {
+      "epoch": 0.42291012412142964,
+      "grad_norm": 1.4108601765445785,
+      "learning_rate": 6.469808111238475e-06,
+      "loss": 0.3113,
+      "step": 2828
+    },
+    {
+      "epoch": 0.4230596680125617,
+      "grad_norm": 1.8058059826416644,
+      "learning_rate": 6.467493115935179e-06,
+      "loss": 0.5298,
+      "step": 2829
+    },
+    {
+      "epoch": 0.42320921190369376,
+      "grad_norm": 1.3575907869698502,
+      "learning_rate": 6.465177776343854e-06,
+      "loss": 0.1908,
+      "step": 2830
+    },
+    {
+      "epoch": 0.4233587557948258,
+      "grad_norm": 1.8817636884787206,
+      "learning_rate": 6.462862093007702e-06,
+      "loss": 0.5272,
+      "step": 2831
+    },
+    {
+      "epoch": 0.4235082996859578,
+      "grad_norm": 1.4050270193232313,
+      "learning_rate": 6.460546066470006e-06,
+      "loss": 0.3227,
+      "step": 2832
+    },
+    {
+      "epoch": 0.42365784357708985,
+      "grad_norm": 1.8465442174063562,
+      "learning_rate": 6.458229697274125e-06,
+      "loss": 0.3563,
+      "step": 2833
+    },
+    {
+      "epoch": 0.42380738746822194,
+      "grad_norm": 1.5798477268932938,
+      "learning_rate": 6.455912985963505e-06,
+      "loss": 0.3733,
+      "step": 2834
+    },
+    {
+      "epoch": 0.423956931359354,
+      "grad_norm": 1.7347800924431869,
+      "learning_rate": 6.4535959330816666e-06,
+      "loss": 0.383,
+      "step": 2835
+    },
+    {
+      "epoch": 0.424106475250486,
+      "grad_norm": 1.9104772036774023,
+      "learning_rate": 6.451278539172215e-06,
+      "loss": 0.5334,
+      "step": 2836
+    },
+    {
+      "epoch": 0.4242560191416181,
+      "grad_norm": 1.4126707638949678,
+      "learning_rate": 6.448960804778831e-06,
+      "loss": 0.3081,
+      "step": 2837
+    },
+    {
+      "epoch": 0.4244055630327501,
+      "grad_norm": 1.594463001904542,
+      "learning_rate": 6.446642730445278e-06,
+      "loss": 0.3175,
+      "step": 2838
+    },
+    {
+      "epoch": 0.42455510692388215,
+      "grad_norm": 1.6332664292262578,
+      "learning_rate": 6.444324316715398e-06,
+      "loss": 0.393,
+      "step": 2839
+    },
+    {
+      "epoch": 0.4247046508150142,
+      "grad_norm": 1.0759394822281065,
+      "learning_rate": 6.4420055641331135e-06,
+      "loss": 0.1674,
+      "step": 2840
+    },
+    {
+      "epoch": 0.42485419470614627,
+      "grad_norm": 1.3272599039432167,
+      "learning_rate": 6.439686473242428e-06,
+      "loss": 0.1989,
+      "step": 2841
+    },
+    {
+      "epoch": 0.4250037385972783,
+      "grad_norm": 0.9431044903567088,
+      "learning_rate": 6.437367044587419e-06,
+      "loss": 0.1963,
+      "step": 2842
+    },
+    {
+      "epoch": 0.42515328248841033,
+      "grad_norm": 1.5321185412479696,
+      "learning_rate": 6.43504727871225e-06,
+      "loss": 0.3409,
+      "step": 2843
+    },
+    {
+      "epoch": 0.4253028263795424,
+      "grad_norm": 1.3126437322786884,
+      "learning_rate": 6.43272717616116e-06,
+      "loss": 0.2271,
+      "step": 2844
+    },
+    {
+      "epoch": 0.42545237027067445,
+      "grad_norm": 1.555272067935412,
+      "learning_rate": 6.4304067374784665e-06,
+      "loss": 0.206,
+      "step": 2845
+    },
+    {
+      "epoch": 0.4256019141618065,
+      "grad_norm": 1.517354875146092,
+      "learning_rate": 6.428085963208567e-06,
+      "loss": 0.2726,
+      "step": 2846
+    },
+    {
+      "epoch": 0.4257514580529385,
+      "grad_norm": 1.468125115279286,
+      "learning_rate": 6.425764853895937e-06,
+      "loss": 0.195,
+      "step": 2847
+    },
+    {
+      "epoch": 0.4259010019440706,
+      "grad_norm": 1.8509310840841346,
+      "learning_rate": 6.423443410085131e-06,
+      "loss": 0.5487,
+      "step": 2848
+    },
+    {
+      "epoch": 0.42605054583520263,
+      "grad_norm": 1.2399783546042413,
+      "learning_rate": 6.421121632320785e-06,
+      "loss": 0.2196,
+      "step": 2849
+    },
+    {
+      "epoch": 0.42620008972633466,
+      "grad_norm": 1.5030129764671514,
+      "learning_rate": 6.418799521147606e-06,
+      "loss": 0.22,
+      "step": 2850
+    },
+    {
+      "epoch": 0.42634963361746675,
+      "grad_norm": 1.0883578935514098,
+      "learning_rate": 6.416477077110388e-06,
+      "loss": 0.1915,
+      "step": 2851
+    },
+    {
+      "epoch": 0.4264991775085988,
+      "grad_norm": 1.162033835127749,
+      "learning_rate": 6.414154300753995e-06,
+      "loss": 0.1868,
+      "step": 2852
+    },
+    {
+      "epoch": 0.4266487213997308,
+      "grad_norm": 0.9782803837081274,
+      "learning_rate": 6.4118311926233746e-06,
+      "loss": 0.22,
+      "step": 2853
+    },
+    {
+      "epoch": 0.42679826529086284,
+      "grad_norm": 1.7110003618772132,
+      "learning_rate": 6.40950775326355e-06,
+      "loss": 0.1888,
+      "step": 2854
+    },
+    {
+      "epoch": 0.42694780918199493,
+      "grad_norm": 1.427911525647864,
+      "learning_rate": 6.407183983219622e-06,
+      "loss": 0.2766,
+      "step": 2855
+    },
+    {
+      "epoch": 0.42709735307312696,
+      "grad_norm": 1.5840868087742603,
+      "learning_rate": 6.404859883036771e-06,
+      "loss": 0.1674,
+      "step": 2856
+    },
+    {
+      "epoch": 0.427246896964259,
+      "grad_norm": 1.8829291374832988,
+      "learning_rate": 6.4025354532602524e-06,
+      "loss": 0.4503,
+      "step": 2857
+    },
+    {
+      "epoch": 0.4273964408553911,
+      "grad_norm": 1.446898713751755,
+      "learning_rate": 6.400210694435399e-06,
+      "loss": 0.2135,
+      "step": 2858
+    },
+    {
+      "epoch": 0.4275459847465231,
+      "grad_norm": 1.3998039538345768,
+      "learning_rate": 6.397885607107623e-06,
+      "loss": 0.2949,
+      "step": 2859
+    },
+    {
+      "epoch": 0.42769552863765514,
+      "grad_norm": 1.4746825648453787,
+      "learning_rate": 6.395560191822411e-06,
+      "loss": 0.2925,
+      "step": 2860
+    },
+    {
+      "epoch": 0.4278450725287872,
+      "grad_norm": 1.4266836184950478,
+      "learning_rate": 6.3932344491253295e-06,
+      "loss": 0.2122,
+      "step": 2861
+    },
+    {
+      "epoch": 0.42799461641991926,
+      "grad_norm": 1.442493331177269,
+      "learning_rate": 6.3909083795620175e-06,
+      "loss": 0.2134,
+      "step": 2862
+    },
+    {
+      "epoch": 0.4281441603110513,
+      "grad_norm": 2.1184696277850064,
+      "learning_rate": 6.388581983678196e-06,
+      "loss": 0.4546,
+      "step": 2863
+    },
+    {
+      "epoch": 0.4282937042021833,
+      "grad_norm": 1.782817853272648,
+      "learning_rate": 6.3862552620196595e-06,
+      "loss": 0.4109,
+      "step": 2864
+    },
+    {
+      "epoch": 0.4284432480933154,
+      "grad_norm": 1.5667820096885934,
+      "learning_rate": 6.383928215132278e-06,
+      "loss": 0.2185,
+      "step": 2865
+    },
+    {
+      "epoch": 0.42859279198444744,
+      "grad_norm": 1.6221826412250884,
+      "learning_rate": 6.381600843561999e-06,
+      "loss": 0.3755,
+      "step": 2866
+    },
+    {
+      "epoch": 0.42874233587557947,
+      "grad_norm": 1.7294106531552245,
+      "learning_rate": 6.379273147854848e-06,
+      "loss": 0.2607,
+      "step": 2867
+    },
+    {
+      "epoch": 0.4288918797667115,
+      "grad_norm": 2.0912433892879214,
+      "learning_rate": 6.3769451285569226e-06,
+      "loss": 0.6155,
+      "step": 2868
+    },
+    {
+      "epoch": 0.4290414236578436,
+      "grad_norm": 1.5578560481589532,
+      "learning_rate": 6.374616786214402e-06,
+      "loss": 0.3255,
+      "step": 2869
+    },
+    {
+      "epoch": 0.4291909675489756,
+      "grad_norm": 2.308644366798669,
+      "learning_rate": 6.372288121373535e-06,
+      "loss": 0.4744,
+      "step": 2870
+    },
+    {
+      "epoch": 0.42934051144010765,
+      "grad_norm": 1.8825047823223349,
+      "learning_rate": 6.369959134580649e-06,
+      "loss": 0.4608,
+      "step": 2871
+    },
+    {
+      "epoch": 0.42949005533123974,
+      "grad_norm": 1.674818620540395,
+      "learning_rate": 6.367629826382148e-06,
+      "loss": 0.1994,
+      "step": 2872
+    },
+    {
+      "epoch": 0.42963959922237177,
+      "grad_norm": 1.608858497846789,
+      "learning_rate": 6.365300197324509e-06,
+      "loss": 0.1895,
+      "step": 2873
+    },
+    {
+      "epoch": 0.4297891431135038,
+      "grad_norm": 1.6863352566219876,
+      "learning_rate": 6.362970247954285e-06,
+      "loss": 0.2895,
+      "step": 2874
+    },
+    {
+      "epoch": 0.42993868700463583,
+      "grad_norm": 1.603267729504892,
+      "learning_rate": 6.3606399788181065e-06,
+      "loss": 0.4429,
+      "step": 2875
+    },
+    {
+      "epoch": 0.4300882308957679,
+      "grad_norm": 1.1309057334924653,
+      "learning_rate": 6.358309390462675e-06,
+      "loss": 0.1826,
+      "step": 2876
+    },
+    {
+      "epoch": 0.43023777478689995,
+      "grad_norm": 1.6293307512813584,
+      "learning_rate": 6.355978483434772e-06,
+      "loss": 0.3851,
+      "step": 2877
+    },
+    {
+      "epoch": 0.430387318678032,
+      "grad_norm": 1.6095937214526117,
+      "learning_rate": 6.353647258281246e-06,
+      "loss": 0.1676,
+      "step": 2878
+    },
+    {
+      "epoch": 0.43053686256916407,
+      "grad_norm": 1.7332943995241852,
+      "learning_rate": 6.35131571554903e-06,
+      "loss": 0.3921,
+      "step": 2879
+    },
+    {
+      "epoch": 0.4306864064602961,
+      "grad_norm": 3.0740209416462285,
+      "learning_rate": 6.348983855785122e-06,
+      "loss": 0.2256,
+      "step": 2880
+    },
+    {
+      "epoch": 0.43083595035142813,
+      "grad_norm": 1.7176886445978465,
+      "learning_rate": 6.346651679536601e-06,
+      "loss": 0.2568,
+      "step": 2881
+    },
+    {
+      "epoch": 0.4309854942425602,
+      "grad_norm": 1.3651054305058985,
+      "learning_rate": 6.34431918735062e-06,
+      "loss": 0.302,
+      "step": 2882
+    },
+    {
+      "epoch": 0.43113503813369225,
+      "grad_norm": 1.9397360315884353,
+      "learning_rate": 6.341986379774399e-06,
+      "loss": 0.2014,
+      "step": 2883
+    },
+    {
+      "epoch": 0.4312845820248243,
+      "grad_norm": 1.930136802042361,
+      "learning_rate": 6.339653257355243e-06,
+      "loss": 0.4523,
+      "step": 2884
+    },
+    {
+      "epoch": 0.4314341259159563,
+      "grad_norm": 1.5662558809247622,
+      "learning_rate": 6.337319820640519e-06,
+      "loss": 0.2364,
+      "step": 2885
+    },
+    {
+      "epoch": 0.4315836698070884,
+      "grad_norm": 1.5689148044077532,
+      "learning_rate": 6.3349860701776795e-06,
+      "loss": 0.3208,
+      "step": 2886
+    },
+    {
+      "epoch": 0.43173321369822043,
+      "grad_norm": 1.0915420487336667,
+      "learning_rate": 6.332652006514241e-06,
+      "loss": 0.1883,
+      "step": 2887
+    },
+    {
+      "epoch": 0.43188275758935246,
+      "grad_norm": 1.4829825194297046,
+      "learning_rate": 6.330317630197797e-06,
+      "loss": 0.2005,
+      "step": 2888
+    },
+    {
+      "epoch": 0.43203230148048455,
+      "grad_norm": 1.681005936318263,
+      "learning_rate": 6.3279829417760195e-06,
+      "loss": 0.1915,
+      "step": 2889
+    },
+    {
+      "epoch": 0.4321818453716166,
+      "grad_norm": 2.026257210929884,
+      "learning_rate": 6.325647941796644e-06,
+      "loss": 0.5229,
+      "step": 2890
+    },
+    {
+      "epoch": 0.4323313892627486,
+      "grad_norm": 2.02662420274482,
+      "learning_rate": 6.323312630807485e-06,
+      "loss": 0.5623,
+      "step": 2891
+    },
+    {
+      "epoch": 0.43248093315388064,
+      "grad_norm": 0.917059657221862,
+      "learning_rate": 6.3209770093564315e-06,
+      "loss": 0.166,
+      "step": 2892
+    },
+    {
+      "epoch": 0.43263047704501273,
+      "grad_norm": 1.7948196473341813,
+      "learning_rate": 6.318641077991438e-06,
+      "loss": 0.2167,
+      "step": 2893
+    },
+    {
+      "epoch": 0.43278002093614476,
+      "grad_norm": 1.5933182870290437,
+      "learning_rate": 6.316304837260542e-06,
+      "loss": 0.348,
+      "step": 2894
+    },
+    {
+      "epoch": 0.4329295648272768,
+      "grad_norm": 1.5675005503133328,
+      "learning_rate": 6.313968287711844e-06,
+      "loss": 0.2499,
+      "step": 2895
+    },
+    {
+      "epoch": 0.4330791087184089,
+      "grad_norm": 2.176887802685468,
+      "learning_rate": 6.311631429893523e-06,
+      "loss": 0.2432,
+      "step": 2896
+    },
+    {
+      "epoch": 0.4332286526095409,
+      "grad_norm": 2.2841410313564903,
+      "learning_rate": 6.309294264353828e-06,
+      "loss": 0.366,
+      "step": 2897
+    },
+    {
+      "epoch": 0.43337819650067294,
+      "grad_norm": 1.584162799649717,
+      "learning_rate": 6.30695679164108e-06,
+      "loss": 0.239,
+      "step": 2898
+    },
+    {
+      "epoch": 0.43352774039180497,
+      "grad_norm": 1.5007265983531852,
+      "learning_rate": 6.3046190123036745e-06,
+      "loss": 0.1616,
+      "step": 2899
+    },
+    {
+      "epoch": 0.43367728428293706,
+      "grad_norm": 2.030068564709013,
+      "learning_rate": 6.302280926890074e-06,
+      "loss": 0.1811,
+      "step": 2900
+    },
+    {
+      "epoch": 0.4338268281740691,
+      "grad_norm": 1.71620144462109,
+      "learning_rate": 6.299942535948816e-06,
+      "loss": 0.4858,
+      "step": 2901
+    },
+    {
+      "epoch": 0.4339763720652011,
+      "grad_norm": 1.6130080927032686,
+      "learning_rate": 6.2976038400285145e-06,
+      "loss": 0.4302,
+      "step": 2902
+    },
+    {
+      "epoch": 0.4341259159563332,
+      "grad_norm": 1.4256957511872124,
+      "learning_rate": 6.295264839677844e-06,
+      "loss": 0.2762,
+      "step": 2903
+    },
+    {
+      "epoch": 0.43427545984746524,
+      "grad_norm": 1.6311831215781756,
+      "learning_rate": 6.292925535445561e-06,
+      "loss": 0.2407,
+      "step": 2904
+    },
+    {
+      "epoch": 0.43442500373859727,
+      "grad_norm": 1.0501034014457853,
+      "learning_rate": 6.290585927880486e-06,
+      "loss": 0.1544,
+      "step": 2905
+    },
+    {
+      "epoch": 0.4345745476297293,
+      "grad_norm": 1.4075211051922059,
+      "learning_rate": 6.288246017531514e-06,
+      "loss": 0.1784,
+      "step": 2906
+    },
+    {
+      "epoch": 0.4347240915208614,
+      "grad_norm": 1.5099442439253756,
+      "learning_rate": 6.285905804947612e-06,
+      "loss": 0.3258,
+      "step": 2907
+    },
+    {
+      "epoch": 0.4348736354119934,
+      "grad_norm": 1.5566923331523632,
+      "learning_rate": 6.283565290677816e-06,
+      "loss": 0.2072,
+      "step": 2908
+    },
+    {
+      "epoch": 0.43502317930312545,
+      "grad_norm": 1.934151158762352,
+      "learning_rate": 6.281224475271232e-06,
+      "loss": 0.3316,
+      "step": 2909
+    },
+    {
+      "epoch": 0.43517272319425754,
+      "grad_norm": 2.0035355877639445,
+      "learning_rate": 6.278883359277037e-06,
+      "loss": 0.5767,
+      "step": 2910
+    },
+    {
+      "epoch": 0.43532226708538957,
+      "grad_norm": 1.3925787643495353,
+      "learning_rate": 6.276541943244482e-06,
+      "loss": 0.3198,
+      "step": 2911
+    },
+    {
+      "epoch": 0.4354718109765216,
+      "grad_norm": 0.9494373556456771,
+      "learning_rate": 6.274200227722887e-06,
+      "loss": 0.1884,
+      "step": 2912
+    },
+    {
+      "epoch": 0.43562135486765363,
+      "grad_norm": 1.4369294236650456,
+      "learning_rate": 6.271858213261635e-06,
+      "loss": 0.1916,
+      "step": 2913
+    },
+    {
+      "epoch": 0.4357708987587857,
+      "grad_norm": 1.248323456170905,
+      "learning_rate": 6.2695159004101905e-06,
+      "loss": 0.1616,
+      "step": 2914
+    },
+    {
+      "epoch": 0.43592044264991775,
+      "grad_norm": 1.1469262780775566,
+      "learning_rate": 6.267173289718079e-06,
+      "loss": 0.1893,
+      "step": 2915
+    },
+    {
+      "epoch": 0.4360699865410498,
+      "grad_norm": 1.5964423980309195,
+      "learning_rate": 6.264830381734903e-06,
+      "loss": 0.3121,
+      "step": 2916
+    },
+    {
+      "epoch": 0.43621953043218187,
+      "grad_norm": 2.019996585766343,
+      "learning_rate": 6.262487177010331e-06,
+      "loss": 0.6759,
+      "step": 2917
+    },
+    {
+      "epoch": 0.4363690743233139,
+      "grad_norm": 1.7460698802434411,
+      "learning_rate": 6.2601436760940995e-06,
+      "loss": 0.3921,
+      "step": 2918
+    },
+    {
+      "epoch": 0.43651861821444593,
+      "grad_norm": 1.6254679162080976,
+      "learning_rate": 6.2577998795360166e-06,
+      "loss": 0.2063,
+      "step": 2919
+    },
+    {
+      "epoch": 0.43666816210557796,
+      "grad_norm": 1.1145785344063608,
+      "learning_rate": 6.25545578788596e-06,
+      "loss": 0.1633,
+      "step": 2920
+    },
+    {
+      "epoch": 0.43681770599671005,
+      "grad_norm": 1.53505082775515,
+      "learning_rate": 6.2531114016938755e-06,
+      "loss": 0.3125,
+      "step": 2921
+    },
+    {
+      "epoch": 0.4369672498878421,
+      "grad_norm": 2.1297043590480085,
+      "learning_rate": 6.250766721509781e-06,
+      "loss": 0.3356,
+      "step": 2922
+    },
+    {
+      "epoch": 0.4371167937789741,
+      "grad_norm": 1.9039440865057902,
+      "learning_rate": 6.24842174788376e-06,
+      "loss": 0.2797,
+      "step": 2923
+    },
+    {
+      "epoch": 0.4372663376701062,
+      "grad_norm": 1.673786518910741,
+      "learning_rate": 6.246076481365964e-06,
+      "loss": 0.3678,
+      "step": 2924
+    },
+    {
+      "epoch": 0.43741588156123823,
+      "grad_norm": 1.1399913425597843,
+      "learning_rate": 6.243730922506618e-06,
+      "loss": 0.1963,
+      "step": 2925
+    },
+    {
+      "epoch": 0.43756542545237026,
+      "grad_norm": 1.9402190325909736,
+      "learning_rate": 6.24138507185601e-06,
+      "loss": 0.5599,
+      "step": 2926
+    },
+    {
+      "epoch": 0.4377149693435023,
+      "grad_norm": 1.7289659635468384,
+      "learning_rate": 6.2390389299645e-06,
+      "loss": 0.3456,
+      "step": 2927
+    },
+    {
+      "epoch": 0.4378645132346344,
+      "grad_norm": 1.6212378839968007,
+      "learning_rate": 6.236692497382517e-06,
+      "loss": 0.2003,
+      "step": 2928
+    },
+    {
+      "epoch": 0.4380140571257664,
+      "grad_norm": 1.3141303268327558,
+      "learning_rate": 6.234345774660555e-06,
+      "loss": 0.2093,
+      "step": 2929
+    },
+    {
+      "epoch": 0.43816360101689844,
+      "grad_norm": 1.4193693342645899,
+      "learning_rate": 6.231998762349179e-06,
+      "loss": 0.3721,
+      "step": 2930
+    },
+    {
+      "epoch": 0.4383131449080305,
+      "grad_norm": 1.5886256685611988,
+      "learning_rate": 6.22965146099902e-06,
+      "loss": 0.1997,
+      "step": 2931
+    },
+    {
+      "epoch": 0.43846268879916256,
+      "grad_norm": 1.8678471944100403,
+      "learning_rate": 6.227303871160779e-06,
+      "loss": 0.3806,
+      "step": 2932
+    },
+    {
+      "epoch": 0.4386122326902946,
+      "grad_norm": 1.4587758386328973,
+      "learning_rate": 6.22495599338522e-06,
+      "loss": 0.3052,
+      "step": 2933
+    },
+    {
+      "epoch": 0.4387617765814266,
+      "grad_norm": 1.4800252577279251,
+      "learning_rate": 6.222607828223179e-06,
+      "loss": 0.3585,
+      "step": 2934
+    },
+    {
+      "epoch": 0.4389113204725587,
+      "grad_norm": 1.8496390423918079,
+      "learning_rate": 6.22025937622556e-06,
+      "loss": 0.5058,
+      "step": 2935
+    },
+    {
+      "epoch": 0.43906086436369074,
+      "grad_norm": 1.5660477130406603,
+      "learning_rate": 6.217910637943332e-06,
+      "loss": 0.202,
+      "step": 2936
+    },
+    {
+      "epoch": 0.43921040825482277,
+      "grad_norm": 2.051942892319578,
+      "learning_rate": 6.21556161392753e-06,
+      "loss": 0.346,
+      "step": 2937
+    },
+    {
+      "epoch": 0.43935995214595486,
+      "grad_norm": 1.1875003670390716,
+      "learning_rate": 6.213212304729259e-06,
+      "loss": 0.2106,
+      "step": 2938
+    },
+    {
+      "epoch": 0.4395094960370869,
+      "grad_norm": 1.5508505327713367,
+      "learning_rate": 6.210862710899688e-06,
+      "loss": 0.3414,
+      "step": 2939
+    },
+    {
+      "epoch": 0.4396590399282189,
+      "grad_norm": 1.4620940174760757,
+      "learning_rate": 6.208512832990057e-06,
+      "loss": 0.2326,
+      "step": 2940
+    },
+    {
+      "epoch": 0.439808583819351,
+      "grad_norm": 1.5695236333564446,
+      "learning_rate": 6.206162671551666e-06,
+      "loss": 0.2074,
+      "step": 2941
+    },
+    {
+      "epoch": 0.43995812771048304,
+      "grad_norm": 1.7058773024979474,
+      "learning_rate": 6.203812227135891e-06,
+      "loss": 0.3473,
+      "step": 2942
+    },
+    {
+      "epoch": 0.44010767160161507,
+      "grad_norm": 1.4295973540897797,
+      "learning_rate": 6.201461500294165e-06,
+      "loss": 0.2694,
+      "step": 2943
+    },
+    {
+      "epoch": 0.4402572154927471,
+      "grad_norm": 1.9125637921122356,
+      "learning_rate": 6.199110491577991e-06,
+      "loss": 0.5139,
+      "step": 2944
+    },
+    {
+      "epoch": 0.4404067593838792,
+      "grad_norm": 2.2653905286807667,
+      "learning_rate": 6.196759201538941e-06,
+      "loss": 0.2828,
+      "step": 2945
+    },
+    {
+      "epoch": 0.4405563032750112,
+      "grad_norm": 1.1078096531641743,
+      "learning_rate": 6.1944076307286485e-06,
+      "loss": 0.222,
+      "step": 2946
+    },
+    {
+      "epoch": 0.44070584716614325,
+      "grad_norm": 1.1756960372209257,
+      "learning_rate": 6.192055779698813e-06,
+      "loss": 0.2178,
+      "step": 2947
+    },
+    {
+      "epoch": 0.44085539105727534,
+      "grad_norm": 1.3681152887792614,
+      "learning_rate": 6.1897036490012055e-06,
+      "loss": 0.1979,
+      "step": 2948
+    },
+    {
+      "epoch": 0.44100493494840737,
+      "grad_norm": 1.7105899375030225,
+      "learning_rate": 6.187351239187654e-06,
+      "loss": 0.4343,
+      "step": 2949
+    },
+    {
+      "epoch": 0.4411544788395394,
+      "grad_norm": 1.3110170942088808,
+      "learning_rate": 6.184998550810062e-06,
+      "loss": 0.2351,
+      "step": 2950
+    },
+    {
+      "epoch": 0.44130402273067143,
+      "grad_norm": 1.6579637186148253,
+      "learning_rate": 6.1826455844203885e-06,
+      "loss": 0.3927,
+      "step": 2951
+    },
+    {
+      "epoch": 0.4414535666218035,
+      "grad_norm": 1.1835524580145198,
+      "learning_rate": 6.180292340570663e-06,
+      "loss": 0.1775,
+      "step": 2952
+    },
+    {
+      "epoch": 0.44160311051293555,
+      "grad_norm": 1.4534094029698128,
+      "learning_rate": 6.177938819812979e-06,
+      "loss": 0.3563,
+      "step": 2953
+    },
+    {
+      "epoch": 0.4417526544040676,
+      "grad_norm": 1.0596251354889077,
+      "learning_rate": 6.175585022699495e-06,
+      "loss": 0.1772,
+      "step": 2954
+    },
+    {
+      "epoch": 0.44190219829519967,
+      "grad_norm": 1.949849754744267,
+      "learning_rate": 6.1732309497824365e-06,
+      "loss": 0.5175,
+      "step": 2955
+    },
+    {
+      "epoch": 0.4420517421863317,
+      "grad_norm": 1.0125156765624894,
+      "learning_rate": 6.170876601614089e-06,
+      "loss": 0.1813,
+      "step": 2956
+    },
+    {
+      "epoch": 0.44220128607746373,
+      "grad_norm": 1.7733454407312104,
+      "learning_rate": 6.168521978746808e-06,
+      "loss": 0.4234,
+      "step": 2957
+    },
+    {
+      "epoch": 0.44235082996859576,
+      "grad_norm": 1.2228581167428354,
+      "learning_rate": 6.16616708173301e-06,
+      "loss": 0.3289,
+      "step": 2958
+    },
+    {
+      "epoch": 0.44250037385972785,
+      "grad_norm": 1.0361601046635946,
+      "learning_rate": 6.1638119111251735e-06,
+      "loss": 0.2069,
+      "step": 2959
+    },
+    {
+      "epoch": 0.4426499177508599,
+      "grad_norm": 1.203904626957853,
+      "learning_rate": 6.1614564674758495e-06,
+      "loss": 0.2516,
+      "step": 2960
+    },
+    {
+      "epoch": 0.4427994616419919,
+      "grad_norm": 1.023639589842975,
+      "learning_rate": 6.1591007513376425e-06,
+      "loss": 0.2318,
+      "step": 2961
+    },
+    {
+      "epoch": 0.442949005533124,
+      "grad_norm": 1.4382415975816416,
+      "learning_rate": 6.156744763263228e-06,
+      "loss": 0.2059,
+      "step": 2962
+    },
+    {
+      "epoch": 0.443098549424256,
+      "grad_norm": 2.5253874416582027,
+      "learning_rate": 6.1543885038053464e-06,
+      "loss": 0.5416,
+      "step": 2963
+    },
+    {
+      "epoch": 0.44324809331538806,
+      "grad_norm": 1.058256199728368,
+      "learning_rate": 6.152031973516795e-06,
+      "loss": 0.1221,
+      "step": 2964
+    },
+    {
+      "epoch": 0.4433976372065201,
+      "grad_norm": 1.4164780696917707,
+      "learning_rate": 6.149675172950441e-06,
+      "loss": 0.2538,
+      "step": 2965
+    },
+    {
+      "epoch": 0.4435471810976522,
+      "grad_norm": 2.030437299566282,
+      "learning_rate": 6.147318102659211e-06,
+      "loss": 0.2116,
+      "step": 2966
+    },
+    {
+      "epoch": 0.4436967249887842,
+      "grad_norm": 0.9394849585455308,
+      "learning_rate": 6.144960763196097e-06,
+      "loss": 0.1816,
+      "step": 2967
+    },
+    {
+      "epoch": 0.44384626887991624,
+      "grad_norm": 1.8329855420116805,
+      "learning_rate": 6.142603155114153e-06,
+      "loss": 0.3582,
+      "step": 2968
+    },
+    {
+      "epoch": 0.4439958127710483,
+      "grad_norm": 1.8735271669425757,
+      "learning_rate": 6.140245278966498e-06,
+      "loss": 0.4366,
+      "step": 2969
+    },
+    {
+      "epoch": 0.44414535666218036,
+      "grad_norm": 1.7892303304786348,
+      "learning_rate": 6.137887135306312e-06,
+      "loss": 0.4139,
+      "step": 2970
+    },
+    {
+      "epoch": 0.4442949005533124,
+      "grad_norm": 1.6664621642434678,
+      "learning_rate": 6.1355287246868365e-06,
+      "loss": 0.3187,
+      "step": 2971
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 1.5224427839350583,
+      "learning_rate": 6.133170047661379e-06,
+      "loss": 0.2085,
+      "step": 2972
+    },
+    {
+      "epoch": 0.4445939883355765,
+      "grad_norm": 1.6171598547263366,
+      "learning_rate": 6.130811104783308e-06,
+      "loss": 0.3763,
+      "step": 2973
+    },
+    {
+      "epoch": 0.44474353222670854,
+      "grad_norm": 1.662874989058589,
+      "learning_rate": 6.128451896606054e-06,
+      "loss": 0.3489,
+      "step": 2974
+    },
+    {
+      "epoch": 0.44489307611784057,
+      "grad_norm": 1.0363220335087222,
+      "learning_rate": 6.12609242368311e-06,
+      "loss": 0.1397,
+      "step": 2975
+    },
+    {
+      "epoch": 0.44504262000897266,
+      "grad_norm": 1.6660010816827036,
+      "learning_rate": 6.123732686568029e-06,
+      "loss": 0.333,
+      "step": 2976
+    },
+    {
+      "epoch": 0.4451921639001047,
+      "grad_norm": 1.1891794312637884,
+      "learning_rate": 6.121372685814431e-06,
+      "loss": 0.2382,
+      "step": 2977
+    },
+    {
+      "epoch": 0.4453417077912367,
+      "grad_norm": 1.576086623428251,
+      "learning_rate": 6.119012421975995e-06,
+      "loss": 0.3202,
+      "step": 2978
+    },
+    {
+      "epoch": 0.44549125168236875,
+      "grad_norm": 1.4353721174926903,
+      "learning_rate": 6.116651895606459e-06,
+      "loss": 0.22,
+      "step": 2979
+    },
+    {
+      "epoch": 0.44564079557350084,
+      "grad_norm": 1.3832487075650706,
+      "learning_rate": 6.1142911072596275e-06,
+      "loss": 0.2093,
+      "step": 2980
+    },
+    {
+      "epoch": 0.44579033946463287,
+      "grad_norm": 1.777859539603896,
+      "learning_rate": 6.111930057489364e-06,
+      "loss": 0.4684,
+      "step": 2981
+    },
+    {
+      "epoch": 0.4459398833557649,
+      "grad_norm": 1.2630447234683253,
+      "learning_rate": 6.10956874684959e-06,
+      "loss": 0.2158,
+      "step": 2982
+    },
+    {
+      "epoch": 0.446089427246897,
+      "grad_norm": 1.6976973376368785,
+      "learning_rate": 6.107207175894297e-06,
+      "loss": 0.3629,
+      "step": 2983
+    },
+    {
+      "epoch": 0.446238971138029,
+      "grad_norm": 1.3370435246604835,
+      "learning_rate": 6.1048453451775305e-06,
+      "loss": 0.178,
+      "step": 2984
+    },
+    {
+      "epoch": 0.44638851502916105,
+      "grad_norm": 1.4545067247689016,
+      "learning_rate": 6.102483255253397e-06,
+      "loss": 0.3052,
+      "step": 2985
+    },
+    {
+      "epoch": 0.4465380589202931,
+      "grad_norm": 1.4865245714334914,
+      "learning_rate": 6.100120906676068e-06,
+      "loss": 0.215,
+      "step": 2986
+    },
+    {
+      "epoch": 0.44668760281142517,
+      "grad_norm": 1.5606516109454205,
+      "learning_rate": 6.097758299999771e-06,
+      "loss": 0.3224,
+      "step": 2987
+    },
+    {
+      "epoch": 0.4468371467025572,
+      "grad_norm": 1.6831853862543933,
+      "learning_rate": 6.095395435778797e-06,
+      "loss": 0.4388,
+      "step": 2988
+    },
+    {
+      "epoch": 0.44698669059368923,
+      "grad_norm": 1.6134233850855104,
+      "learning_rate": 6.093032314567498e-06,
+      "loss": 0.3387,
+      "step": 2989
+    },
+    {
+      "epoch": 0.4471362344848213,
+      "grad_norm": 1.8479914330951175,
+      "learning_rate": 6.0906689369202835e-06,
+      "loss": 0.5229,
+      "step": 2990
+    },
+    {
+      "epoch": 0.44728577837595335,
+      "grad_norm": 1.697848699434105,
+      "learning_rate": 6.088305303391625e-06,
+      "loss": 0.3263,
+      "step": 2991
+    },
+    {
+      "epoch": 0.4474353222670854,
+      "grad_norm": 2.340324861868485,
+      "learning_rate": 6.085941414536054e-06,
+      "loss": 0.5403,
+      "step": 2992
+    },
+    {
+      "epoch": 0.4475848661582174,
+      "grad_norm": 1.366594482966174,
+      "learning_rate": 6.083577270908163e-06,
+      "loss": 0.2498,
+      "step": 2993
+    },
+    {
+      "epoch": 0.4477344100493495,
+      "grad_norm": 2.003441472279418,
+      "learning_rate": 6.0812128730626e-06,
+      "loss": 0.2421,
+      "step": 2994
+    },
+    {
+      "epoch": 0.4478839539404815,
+      "grad_norm": 1.4812098336709252,
+      "learning_rate": 6.078848221554077e-06,
+      "loss": 0.3513,
+      "step": 2995
+    },
+    {
+      "epoch": 0.44803349783161356,
+      "grad_norm": 1.121351564649134,
+      "learning_rate": 6.076483316937364e-06,
+      "loss": 0.1646,
+      "step": 2996
+    },
+    {
+      "epoch": 0.44818304172274565,
+      "grad_norm": 1.3593452823065253,
+      "learning_rate": 6.07411815976729e-06,
+      "loss": 0.3391,
+      "step": 2997
+    },
+    {
+      "epoch": 0.4483325856138777,
+      "grad_norm": 1.6796848208383883,
+      "learning_rate": 6.071752750598745e-06,
+      "loss": 0.5086,
+      "step": 2998
+    },
+    {
+      "epoch": 0.4484821295050097,
+      "grad_norm": 1.4048595972263043,
+      "learning_rate": 6.0693870899866756e-06,
+      "loss": 0.306,
+      "step": 2999
+    },
+    {
+      "epoch": 0.4486316733961418,
+      "grad_norm": 1.3825869686134655,
+      "learning_rate": 6.067021178486088e-06,
+      "loss": 0.207,
+      "step": 3000
+    },
+    {
+      "epoch": 0.4487812172872738,
+      "grad_norm": 1.5962644544557885,
+      "learning_rate": 6.06465501665205e-06,
+      "loss": 0.401,
+      "step": 3001
+    },
+    {
+      "epoch": 0.44893076117840586,
+      "grad_norm": 1.4371440397490116,
+      "learning_rate": 6.062288605039683e-06,
+      "loss": 0.1915,
+      "step": 3002
+    },
+    {
+      "epoch": 0.4490803050695379,
+      "grad_norm": 1.968364897408132,
+      "learning_rate": 6.059921944204173e-06,
+      "loss": 0.4246,
+      "step": 3003
+    },
+    {
+      "epoch": 0.44922984896067,
+      "grad_norm": 1.8653866690731011,
+      "learning_rate": 6.057555034700759e-06,
+      "loss": 0.427,
+      "step": 3004
+    },
+    {
+      "epoch": 0.449379392851802,
+      "grad_norm": 1.7168140345422174,
+      "learning_rate": 6.055187877084743e-06,
+      "loss": 0.3281,
+      "step": 3005
+    },
+    {
+      "epoch": 0.44952893674293404,
+      "grad_norm": 1.5060475228175942,
+      "learning_rate": 6.052820471911481e-06,
+      "loss": 0.3382,
+      "step": 3006
+    },
+    {
+      "epoch": 0.4496784806340661,
+      "grad_norm": 1.4572399542192809,
+      "learning_rate": 6.05045281973639e-06,
+      "loss": 0.326,
+      "step": 3007
+    },
+    {
+      "epoch": 0.44982802452519816,
+      "grad_norm": 1.9104089940276523,
+      "learning_rate": 6.048084921114944e-06,
+      "loss": 0.3776,
+      "step": 3008
+    },
+    {
+      "epoch": 0.4499775684163302,
+      "grad_norm": 1.6392841422899054,
+      "learning_rate": 6.045716776602674e-06,
+      "loss": 0.3508,
+      "step": 3009
+    },
+    {
+      "epoch": 0.4501271123074622,
+      "grad_norm": 1.795922624052802,
+      "learning_rate": 6.043348386755172e-06,
+      "loss": 0.5236,
+      "step": 3010
+    },
+    {
+      "epoch": 0.4502766561985943,
+      "grad_norm": 1.4038039245527818,
+      "learning_rate": 6.040979752128082e-06,
+      "loss": 0.1898,
+      "step": 3011
+    },
+    {
+      "epoch": 0.45042620008972634,
+      "grad_norm": 1.34690239318641,
+      "learning_rate": 6.038610873277109e-06,
+      "loss": 0.322,
+      "step": 3012
+    },
+    {
+      "epoch": 0.45057574398085837,
+      "grad_norm": 1.938382182478136,
+      "learning_rate": 6.036241750758018e-06,
+      "loss": 0.3155,
+      "step": 3013
+    },
+    {
+      "epoch": 0.45072528787199045,
+      "grad_norm": 1.5315561280490633,
+      "learning_rate": 6.033872385126625e-06,
+      "loss": 0.2086,
+      "step": 3014
+    },
+    {
+      "epoch": 0.4508748317631225,
+      "grad_norm": 1.4664133436532398,
+      "learning_rate": 6.031502776938806e-06,
+      "loss": 0.339,
+      "step": 3015
+    },
+    {
+      "epoch": 0.4510243756542545,
+      "grad_norm": 1.3762760635076952,
+      "learning_rate": 6.029132926750494e-06,
+      "loss": 0.2114,
+      "step": 3016
+    },
+    {
+      "epoch": 0.45117391954538655,
+      "grad_norm": 1.3005324933468017,
+      "learning_rate": 6.0267628351176796e-06,
+      "loss": 0.1908,
+      "step": 3017
+    },
+    {
+      "epoch": 0.45132346343651863,
+      "grad_norm": 1.0685843417404295,
+      "learning_rate": 6.02439250259641e-06,
+      "loss": 0.1924,
+      "step": 3018
+    },
+    {
+      "epoch": 0.45147300732765067,
+      "grad_norm": 1.771550384435644,
+      "learning_rate": 6.022021929742785e-06,
+      "loss": 0.2815,
+      "step": 3019
+    },
+    {
+      "epoch": 0.4516225512187827,
+      "grad_norm": 2.383086764076166,
+      "learning_rate": 6.019651117112967e-06,
+      "loss": 0.3847,
+      "step": 3020
+    },
+    {
+      "epoch": 0.4517720951099148,
+      "grad_norm": 1.4770169366158958,
+      "learning_rate": 6.0172800652631706e-06,
+      "loss": 0.3493,
+      "step": 3021
+    },
+    {
+      "epoch": 0.4519216390010468,
+      "grad_norm": 1.3322870296999558,
+      "learning_rate": 6.0149087747496645e-06,
+      "loss": 0.1781,
+      "step": 3022
+    },
+    {
+      "epoch": 0.45207118289217885,
+      "grad_norm": 1.0795861187632947,
+      "learning_rate": 6.0125372461287814e-06,
+      "loss": 0.2391,
+      "step": 3023
+    },
+    {
+      "epoch": 0.4522207267833109,
+      "grad_norm": 1.8068779784303053,
+      "learning_rate": 6.010165479956901e-06,
+      "loss": 0.418,
+      "step": 3024
+    },
+    {
+      "epoch": 0.45237027067444296,
+      "grad_norm": 1.9286914389868077,
+      "learning_rate": 6.007793476790463e-06,
+      "loss": 0.1949,
+      "step": 3025
+    },
+    {
+      "epoch": 0.452519814565575,
+      "grad_norm": 1.146144637893908,
+      "learning_rate": 6.005421237185965e-06,
+      "loss": 0.1524,
+      "step": 3026
+    },
+    {
+      "epoch": 0.452669358456707,
+      "grad_norm": 1.0973657922867939,
+      "learning_rate": 6.003048761699953e-06,
+      "loss": 0.3547,
+      "step": 3027
+    },
+    {
+      "epoch": 0.4528189023478391,
+      "grad_norm": 1.4581605124819323,
+      "learning_rate": 6.000676050889036e-06,
+      "loss": 0.1848,
+      "step": 3028
+    },
+    {
+      "epoch": 0.45296844623897115,
+      "grad_norm": 1.269034060908923,
+      "learning_rate": 5.998303105309873e-06,
+      "loss": 0.199,
+      "step": 3029
+    },
+    {
+      "epoch": 0.4531179901301032,
+      "grad_norm": 1.5171368496196267,
+      "learning_rate": 5.995929925519181e-06,
+      "loss": 0.212,
+      "step": 3030
+    },
+    {
+      "epoch": 0.4532675340212352,
+      "grad_norm": 1.9339847824352052,
+      "learning_rate": 5.99355651207373e-06,
+      "loss": 0.3863,
+      "step": 3031
+    },
+    {
+      "epoch": 0.4534170779123673,
+      "grad_norm": 1.6847447888231042,
+      "learning_rate": 5.991182865530346e-06,
+      "loss": 0.4776,
+      "step": 3032
+    },
+    {
+      "epoch": 0.4535666218034993,
+      "grad_norm": 1.7472594442925573,
+      "learning_rate": 5.988808986445911e-06,
+      "loss": 0.1947,
+      "step": 3033
+    },
+    {
+      "epoch": 0.45371616569463136,
+      "grad_norm": 1.1215834395483242,
+      "learning_rate": 5.986434875377356e-06,
+      "loss": 0.1856,
+      "step": 3034
+    },
+    {
+      "epoch": 0.45386570958576344,
+      "grad_norm": 1.31004013577431,
+      "learning_rate": 5.984060532881673e-06,
+      "loss": 0.1829,
+      "step": 3035
+    },
+    {
+      "epoch": 0.4540152534768955,
+      "grad_norm": 1.1780108416771637,
+      "learning_rate": 5.981685959515905e-06,
+      "loss": 0.2096,
+      "step": 3036
+    },
+    {
+      "epoch": 0.4541647973680275,
+      "grad_norm": 1.9874221617073415,
+      "learning_rate": 5.979311155837151e-06,
+      "loss": 0.3116,
+      "step": 3037
+    },
+    {
+      "epoch": 0.45431434125915954,
+      "grad_norm": 1.3795207278158967,
+      "learning_rate": 5.976936122402562e-06,
+      "loss": 0.1797,
+      "step": 3038
+    },
+    {
+      "epoch": 0.4544638851502916,
+      "grad_norm": 1.1174423153921418,
+      "learning_rate": 5.974560859769343e-06,
+      "loss": 0.1965,
+      "step": 3039
+    },
+    {
+      "epoch": 0.45461342904142366,
+      "grad_norm": 1.4270442541948576,
+      "learning_rate": 5.972185368494754e-06,
+      "loss": 0.194,
+      "step": 3040
+    },
+    {
+      "epoch": 0.4547629729325557,
+      "grad_norm": 1.6346945251535583,
+      "learning_rate": 5.9698096491361095e-06,
+      "loss": 0.4558,
+      "step": 3041
+    },
+    {
+      "epoch": 0.4549125168236878,
+      "grad_norm": 1.726471907193774,
+      "learning_rate": 5.967433702250772e-06,
+      "loss": 0.3279,
+      "step": 3042
+    },
+    {
+      "epoch": 0.4550620607148198,
+      "grad_norm": 1.4908257407248642,
+      "learning_rate": 5.965057528396166e-06,
+      "loss": 0.3607,
+      "step": 3043
+    },
+    {
+      "epoch": 0.45521160460595184,
+      "grad_norm": 1.9466145535642352,
+      "learning_rate": 5.962681128129764e-06,
+      "loss": 0.3927,
+      "step": 3044
+    },
+    {
+      "epoch": 0.45536114849708387,
+      "grad_norm": 1.790484795898369,
+      "learning_rate": 5.960304502009091e-06,
+      "loss": 0.5027,
+      "step": 3045
+    },
+    {
+      "epoch": 0.45551069238821595,
+      "grad_norm": 2.1466817305191337,
+      "learning_rate": 5.957927650591727e-06,
+      "loss": 0.5421,
+      "step": 3046
+    },
+    {
+      "epoch": 0.455660236279348,
+      "grad_norm": 1.7488963785457814,
+      "learning_rate": 5.9555505744353045e-06,
+      "loss": 0.5295,
+      "step": 3047
+    },
+    {
+      "epoch": 0.45580978017048,
+      "grad_norm": 1.3663500809725344,
+      "learning_rate": 5.953173274097508e-06,
+      "loss": 0.3152,
+      "step": 3048
+    },
+    {
+      "epoch": 0.4559593240616121,
+      "grad_norm": 1.3010976352200727,
+      "learning_rate": 5.950795750136075e-06,
+      "loss": 0.2174,
+      "step": 3049
+    },
+    {
+      "epoch": 0.45610886795274413,
+      "grad_norm": 1.8659509131326897,
+      "learning_rate": 5.948418003108795e-06,
+      "loss": 0.418,
+      "step": 3050
+    },
+    {
+      "epoch": 0.45625841184387617,
+      "grad_norm": 2.351423158034144,
+      "learning_rate": 5.946040033573514e-06,
+      "loss": 0.5714,
+      "step": 3051
+    },
+    {
+      "epoch": 0.4564079557350082,
+      "grad_norm": 1.7203691095571323,
+      "learning_rate": 5.943661842088122e-06,
+      "loss": 0.3555,
+      "step": 3052
+    },
+    {
+      "epoch": 0.4565574996261403,
+      "grad_norm": 1.6103446132123642,
+      "learning_rate": 5.941283429210568e-06,
+      "loss": 0.3919,
+      "step": 3053
+    },
+    {
+      "epoch": 0.4567070435172723,
+      "grad_norm": 1.634616029154534,
+      "learning_rate": 5.93890479549885e-06,
+      "loss": 0.4904,
+      "step": 3054
+    },
+    {
+      "epoch": 0.45685658740840435,
+      "grad_norm": 1.2519583084546646,
+      "learning_rate": 5.936525941511017e-06,
+      "loss": 0.2004,
+      "step": 3055
+    },
+    {
+      "epoch": 0.45700613129953643,
+      "grad_norm": 1.8604210227354285,
+      "learning_rate": 5.934146867805175e-06,
+      "loss": 0.4454,
+      "step": 3056
+    },
+    {
+      "epoch": 0.45715567519066846,
+      "grad_norm": 1.5966577837546947,
+      "learning_rate": 5.931767574939473e-06,
+      "loss": 0.3326,
+      "step": 3057
+    },
+    {
+      "epoch": 0.4573052190818005,
+      "grad_norm": 1.7485226713542816,
+      "learning_rate": 5.929388063472119e-06,
+      "loss": 0.3078,
+      "step": 3058
+    },
+    {
+      "epoch": 0.4574547629729326,
+      "grad_norm": 1.8101199489307283,
+      "learning_rate": 5.9270083339613695e-06,
+      "loss": 0.5576,
+      "step": 3059
+    },
+    {
+      "epoch": 0.4576043068640646,
+      "grad_norm": 1.1197623012447884,
+      "learning_rate": 5.924628386965529e-06,
+      "loss": 0.1692,
+      "step": 3060
+    },
+    {
+      "epoch": 0.45775385075519665,
+      "grad_norm": 1.5496112484027527,
+      "learning_rate": 5.922248223042959e-06,
+      "loss": 0.3748,
+      "step": 3061
+    },
+    {
+      "epoch": 0.4579033946463287,
+      "grad_norm": 1.4213073823646392,
+      "learning_rate": 5.919867842752067e-06,
+      "loss": 0.2585,
+      "step": 3062
+    },
+    {
+      "epoch": 0.45805293853746076,
+      "grad_norm": 1.6766072796105778,
+      "learning_rate": 5.917487246651314e-06,
+      "loss": 0.3301,
+      "step": 3063
+    },
+    {
+      "epoch": 0.4582024824285928,
+      "grad_norm": 1.1877206484440475,
+      "learning_rate": 5.915106435299212e-06,
+      "loss": 0.2118,
+      "step": 3064
+    },
+    {
+      "epoch": 0.4583520263197248,
+      "grad_norm": 1.4732071986791726,
+      "learning_rate": 5.912725409254321e-06,
+      "loss": 0.1794,
+      "step": 3065
+    },
+    {
+      "epoch": 0.4585015702108569,
+      "grad_norm": 1.761728417820423,
+      "learning_rate": 5.910344169075252e-06,
+      "loss": 0.3489,
+      "step": 3066
+    },
+    {
+      "epoch": 0.45865111410198894,
+      "grad_norm": 1.356911951793647,
+      "learning_rate": 5.907962715320668e-06,
+      "loss": 0.325,
+      "step": 3067
+    },
+    {
+      "epoch": 0.458800657993121,
+      "grad_norm": 1.2177589250447785,
+      "learning_rate": 5.905581048549279e-06,
+      "loss": 0.1797,
+      "step": 3068
+    },
+    {
+      "epoch": 0.458950201884253,
+      "grad_norm": 1.683972181197731,
+      "learning_rate": 5.90319916931985e-06,
+      "loss": 0.3762,
+      "step": 3069
+    },
+    {
+      "epoch": 0.4590997457753851,
+      "grad_norm": 1.6629827017110606,
+      "learning_rate": 5.900817078191191e-06,
+      "loss": 0.1876,
+      "step": 3070
+    },
+    {
+      "epoch": 0.4592492896665171,
+      "grad_norm": 1.990952022645022,
+      "learning_rate": 5.898434775722165e-06,
+      "loss": 0.4711,
+      "step": 3071
+    },
+    {
+      "epoch": 0.45939883355764916,
+      "grad_norm": 1.3405923448155683,
+      "learning_rate": 5.896052262471681e-06,
+      "loss": 0.3233,
+      "step": 3072
+    },
+    {
+      "epoch": 0.45954837744878124,
+      "grad_norm": 1.242924374220424,
+      "learning_rate": 5.893669538998701e-06,
+      "loss": 0.2268,
+      "step": 3073
+    },
+    {
+      "epoch": 0.4596979213399133,
+      "grad_norm": 1.6674961385367664,
+      "learning_rate": 5.891286605862236e-06,
+      "loss": 0.4254,
+      "step": 3074
+    },
+    {
+      "epoch": 0.4598474652310453,
+      "grad_norm": 1.9155364958899164,
+      "learning_rate": 5.888903463621342e-06,
+      "loss": 0.3741,
+      "step": 3075
+    },
+    {
+      "epoch": 0.45999700912217734,
+      "grad_norm": 1.467167356716408,
+      "learning_rate": 5.886520112835128e-06,
+      "loss": 0.4894,
+      "step": 3076
+    },
+    {
+      "epoch": 0.4601465530133094,
+      "grad_norm": 1.528239460716899,
+      "learning_rate": 5.884136554062753e-06,
+      "loss": 0.2938,
+      "step": 3077
+    },
+    {
+      "epoch": 0.46029609690444145,
+      "grad_norm": 1.2857242621803937,
+      "learning_rate": 5.881752787863421e-06,
+      "loss": 0.3372,
+      "step": 3078
+    },
+    {
+      "epoch": 0.4604456407955735,
+      "grad_norm": 1.566112761935387,
+      "learning_rate": 5.879368814796388e-06,
+      "loss": 0.3513,
+      "step": 3079
+    },
+    {
+      "epoch": 0.46059518468670557,
+      "grad_norm": 1.829132984093117,
+      "learning_rate": 5.876984635420956e-06,
+      "loss": 0.4273,
+      "step": 3080
+    },
+    {
+      "epoch": 0.4607447285778376,
+      "grad_norm": 1.3703912717781694,
+      "learning_rate": 5.874600250296477e-06,
+      "loss": 0.2825,
+      "step": 3081
+    },
+    {
+      "epoch": 0.46089427246896963,
+      "grad_norm": 1.102997889638414,
+      "learning_rate": 5.87221565998235e-06,
+      "loss": 0.182,
+      "step": 3082
+    },
+    {
+      "epoch": 0.46104381636010167,
+      "grad_norm": 1.2503343656492412,
+      "learning_rate": 5.869830865038022e-06,
+      "loss": 0.1864,
+      "step": 3083
+    },
+    {
+      "epoch": 0.46119336025123375,
+      "grad_norm": 1.3521302194014102,
+      "learning_rate": 5.867445866022992e-06,
+      "loss": 0.3524,
+      "step": 3084
+    },
+    {
+      "epoch": 0.4613429041423658,
+      "grad_norm": 1.4419400149521013,
+      "learning_rate": 5.865060663496801e-06,
+      "loss": 0.2945,
+      "step": 3085
+    },
+    {
+      "epoch": 0.4614924480334978,
+      "grad_norm": 0.8534550219741183,
+      "learning_rate": 5.862675258019042e-06,
+      "loss": 0.2017,
+      "step": 3086
+    },
+    {
+      "epoch": 0.4616419919246299,
+      "grad_norm": 1.1420661007230424,
+      "learning_rate": 5.860289650149353e-06,
+      "loss": 0.1907,
+      "step": 3087
+    },
+    {
+      "epoch": 0.46179153581576193,
+      "grad_norm": 1.2419372462457205,
+      "learning_rate": 5.857903840447422e-06,
+      "loss": 0.3804,
+      "step": 3088
+    },
+    {
+      "epoch": 0.46194107970689396,
+      "grad_norm": 1.0043595882377505,
+      "learning_rate": 5.855517829472981e-06,
+      "loss": 0.181,
+      "step": 3089
+    },
+    {
+      "epoch": 0.462090623598026,
+      "grad_norm": 1.4634097900797467,
+      "learning_rate": 5.853131617785813e-06,
+      "loss": 0.3246,
+      "step": 3090
+    },
+    {
+      "epoch": 0.4622401674891581,
+      "grad_norm": 1.587493020890882,
+      "learning_rate": 5.850745205945745e-06,
+      "loss": 0.2042,
+      "step": 3091
+    },
+    {
+      "epoch": 0.4623897113802901,
+      "grad_norm": 1.7333993057463715,
+      "learning_rate": 5.848358594512654e-06,
+      "loss": 0.325,
+      "step": 3092
+    },
+    {
+      "epoch": 0.46253925527142215,
+      "grad_norm": 1.616159656091411,
+      "learning_rate": 5.84597178404646e-06,
+      "loss": 0.2417,
+      "step": 3093
+    },
+    {
+      "epoch": 0.46268879916255423,
+      "grad_norm": 1.2656140621324898,
+      "learning_rate": 5.8435847751071325e-06,
+      "loss": 0.2453,
+      "step": 3094
+    },
+    {
+      "epoch": 0.46283834305368626,
+      "grad_norm": 1.074466172247766,
+      "learning_rate": 5.841197568254687e-06,
+      "loss": 0.1631,
+      "step": 3095
+    },
+    {
+      "epoch": 0.4629878869448183,
+      "grad_norm": 1.0979861038745287,
+      "learning_rate": 5.838810164049185e-06,
+      "loss": 0.1769,
+      "step": 3096
+    },
+    {
+      "epoch": 0.4631374308359503,
+      "grad_norm": 1.2926739611623306,
+      "learning_rate": 5.836422563050734e-06,
+      "loss": 0.2501,
+      "step": 3097
+    },
+    {
+      "epoch": 0.4632869747270824,
+      "grad_norm": 1.727710286265107,
+      "learning_rate": 5.834034765819489e-06,
+      "loss": 0.5367,
+      "step": 3098
+    },
+    {
+      "epoch": 0.46343651861821444,
+      "grad_norm": 1.4761492716100206,
+      "learning_rate": 5.831646772915651e-06,
+      "loss": 0.3464,
+      "step": 3099
+    },
+    {
+      "epoch": 0.4635860625093465,
+      "grad_norm": 1.2203558008080015,
+      "learning_rate": 5.829258584899464e-06,
+      "loss": 0.1696,
+      "step": 3100
+    },
+    {
+      "epoch": 0.46373560640047856,
+      "grad_norm": 2.010676119341004,
+      "learning_rate": 5.826870202331221e-06,
+      "loss": 0.373,
+      "step": 3101
+    },
+    {
+      "epoch": 0.4638851502916106,
+      "grad_norm": 1.691843238482411,
+      "learning_rate": 5.824481625771262e-06,
+      "loss": 0.3301,
+      "step": 3102
+    },
+    {
+      "epoch": 0.4640346941827426,
+      "grad_norm": 1.1543076074169056,
+      "learning_rate": 5.822092855779966e-06,
+      "loss": 0.1873,
+      "step": 3103
+    },
+    {
+      "epoch": 0.46418423807387466,
+      "grad_norm": 1.2233599414089642,
+      "learning_rate": 5.819703892917763e-06,
+      "loss": 0.1865,
+      "step": 3104
+    },
+    {
+      "epoch": 0.46433378196500674,
+      "grad_norm": 2.210291468184369,
+      "learning_rate": 5.817314737745127e-06,
+      "loss": 0.4183,
+      "step": 3105
+    },
+    {
+      "epoch": 0.4644833258561388,
+      "grad_norm": 1.4383760211249257,
+      "learning_rate": 5.8149253908225766e-06,
+      "loss": 0.2193,
+      "step": 3106
+    },
+    {
+      "epoch": 0.4646328697472708,
+      "grad_norm": 1.3428933717745177,
+      "learning_rate": 5.812535852710677e-06,
+      "loss": 0.4662,
+      "step": 3107
+    },
+    {
+      "epoch": 0.4647824136384029,
+      "grad_norm": 1.523129681520151,
+      "learning_rate": 5.810146123970034e-06,
+      "loss": 0.5232,
+      "step": 3108
+    },
+    {
+      "epoch": 0.4649319575295349,
+      "grad_norm": 1.378941458503838,
+      "learning_rate": 5.807756205161304e-06,
+      "loss": 0.1905,
+      "step": 3109
+    },
+    {
+      "epoch": 0.46508150142066695,
+      "grad_norm": 2.2157282770241844,
+      "learning_rate": 5.8053660968451815e-06,
+      "loss": 0.3278,
+      "step": 3110
+    },
+    {
+      "epoch": 0.46523104531179904,
+      "grad_norm": 2.1780194356833182,
+      "learning_rate": 5.802975799582413e-06,
+      "loss": 0.4822,
+      "step": 3111
+    },
+    {
+      "epoch": 0.46538058920293107,
+      "grad_norm": 1.1575081784389918,
+      "learning_rate": 5.800585313933784e-06,
+      "loss": 0.1978,
+      "step": 3112
+    },
+    {
+      "epoch": 0.4655301330940631,
+      "grad_norm": 1.9490148059207726,
+      "learning_rate": 5.798194640460124e-06,
+      "loss": 0.5377,
+      "step": 3113
+    },
+    {
+      "epoch": 0.46567967698519513,
+      "grad_norm": 1.0308478611703875,
+      "learning_rate": 5.795803779722309e-06,
+      "loss": 0.1269,
+      "step": 3114
+    },
+    {
+      "epoch": 0.4658292208763272,
+      "grad_norm": 1.456905452881718,
+      "learning_rate": 5.793412732281258e-06,
+      "loss": 0.2402,
+      "step": 3115
+    },
+    {
+      "epoch": 0.46597876476745925,
+      "grad_norm": 1.865438890479385,
+      "learning_rate": 5.791021498697933e-06,
+      "loss": 0.2259,
+      "step": 3116
+    },
+    {
+      "epoch": 0.4661283086585913,
+      "grad_norm": 1.5482682061442332,
+      "learning_rate": 5.7886300795333405e-06,
+      "loss": 0.2349,
+      "step": 3117
+    },
+    {
+      "epoch": 0.46627785254972337,
+      "grad_norm": 1.168054204385437,
+      "learning_rate": 5.786238475348531e-06,
+      "loss": 0.1822,
+      "step": 3118
+    },
+    {
+      "epoch": 0.4664273964408554,
+      "grad_norm": 1.5800515492094007,
+      "learning_rate": 5.7838466867046e-06,
+      "loss": 0.2249,
+      "step": 3119
+    },
+    {
+      "epoch": 0.46657694033198743,
+      "grad_norm": 1.585867798029898,
+      "learning_rate": 5.78145471416268e-06,
+      "loss": 0.3313,
+      "step": 3120
+    },
+    {
+      "epoch": 0.46672648422311946,
+      "grad_norm": 1.2912421381268824,
+      "learning_rate": 5.779062558283954e-06,
+      "loss": 0.1931,
+      "step": 3121
+    },
+    {
+      "epoch": 0.46687602811425155,
+      "grad_norm": 1.303872839200776,
+      "learning_rate": 5.776670219629643e-06,
+      "loss": 0.2294,
+      "step": 3122
+    },
+    {
+      "epoch": 0.4670255720053836,
+      "grad_norm": 1.32116962155212,
+      "learning_rate": 5.7742776987610135e-06,
+      "loss": 0.2108,
+      "step": 3123
+    },
+    {
+      "epoch": 0.4671751158965156,
+      "grad_norm": 1.0010399849500764,
+      "learning_rate": 5.771884996239373e-06,
+      "loss": 0.154,
+      "step": 3124
+    },
+    {
+      "epoch": 0.4673246597876477,
+      "grad_norm": 1.4031763925464884,
+      "learning_rate": 5.7694921126260765e-06,
+      "loss": 0.2299,
+      "step": 3125
+    },
+    {
+      "epoch": 0.46747420367877973,
+      "grad_norm": 1.9032263969945602,
+      "learning_rate": 5.767099048482512e-06,
+      "loss": 0.2426,
+      "step": 3126
+    },
+    {
+      "epoch": 0.46762374756991176,
+      "grad_norm": 1.0697134527322825,
+      "learning_rate": 5.764705804370119e-06,
+      "loss": 0.2073,
+      "step": 3127
+    },
+    {
+      "epoch": 0.4677732914610438,
+      "grad_norm": 1.42472795223601,
+      "learning_rate": 5.762312380850373e-06,
+      "loss": 0.3811,
+      "step": 3128
+    },
+    {
+      "epoch": 0.4679228353521759,
+      "grad_norm": 1.312885115809687,
+      "learning_rate": 5.759918778484797e-06,
+      "loss": 0.2057,
+      "step": 3129
+    },
+    {
+      "epoch": 0.4680723792433079,
+      "grad_norm": 1.5752141659312524,
+      "learning_rate": 5.757524997834951e-06,
+      "loss": 0.2032,
+      "step": 3130
+    },
+    {
+      "epoch": 0.46822192313443994,
+      "grad_norm": 1.9755154425978854,
+      "learning_rate": 5.755131039462441e-06,
+      "loss": 0.2944,
+      "step": 3131
+    },
+    {
+      "epoch": 0.46837146702557203,
+      "grad_norm": 1.5857308751132788,
+      "learning_rate": 5.752736903928912e-06,
+      "loss": 0.3172,
+      "step": 3132
+    },
+    {
+      "epoch": 0.46852101091670406,
+      "grad_norm": 1.276569604328609,
+      "learning_rate": 5.75034259179605e-06,
+      "loss": 0.1794,
+      "step": 3133
+    },
+    {
+      "epoch": 0.4686705548078361,
+      "grad_norm": 1.4874225477081564,
+      "learning_rate": 5.747948103625585e-06,
+      "loss": 0.3464,
+      "step": 3134
+    },
+    {
+      "epoch": 0.4688200986989681,
+      "grad_norm": 1.8677226819900095,
+      "learning_rate": 5.745553439979287e-06,
+      "loss": 0.3892,
+      "step": 3135
+    },
+    {
+      "epoch": 0.4689696425901002,
+      "grad_norm": 1.9234557105998888,
+      "learning_rate": 5.743158601418967e-06,
+      "loss": 0.3449,
+      "step": 3136
+    },
+    {
+      "epoch": 0.46911918648123224,
+      "grad_norm": 1.4122630414281268,
+      "learning_rate": 5.740763588506475e-06,
+      "loss": 0.2106,
+      "step": 3137
+    },
+    {
+      "epoch": 0.4692687303723643,
+      "grad_norm": 2.120747227051215,
+      "learning_rate": 5.738368401803708e-06,
+      "loss": 0.4917,
+      "step": 3138
+    },
+    {
+      "epoch": 0.46941827426349636,
+      "grad_norm": 1.5280456344885767,
+      "learning_rate": 5.735973041872597e-06,
+      "loss": 0.2088,
+      "step": 3139
+    },
+    {
+      "epoch": 0.4695678181546284,
+      "grad_norm": 2.1695026435655715,
+      "learning_rate": 5.733577509275119e-06,
+      "loss": 0.5766,
+      "step": 3140
+    },
+    {
+      "epoch": 0.4697173620457604,
+      "grad_norm": 1.9039804361985175,
+      "learning_rate": 5.731181804573286e-06,
+      "loss": 0.3801,
+      "step": 3141
+    },
+    {
+      "epoch": 0.46986690593689245,
+      "grad_norm": 1.6142863728603896,
+      "learning_rate": 5.728785928329157e-06,
+      "loss": 0.261,
+      "step": 3142
+    },
+    {
+      "epoch": 0.47001644982802454,
+      "grad_norm": 1.6235559241321793,
+      "learning_rate": 5.726389881104824e-06,
+      "loss": 0.3315,
+      "step": 3143
+    },
+    {
+      "epoch": 0.47016599371915657,
+      "grad_norm": 2.715507431689836,
+      "learning_rate": 5.7239936634624225e-06,
+      "loss": 0.3486,
+      "step": 3144
+    },
+    {
+      "epoch": 0.4703155376102886,
+      "grad_norm": 1.8605569560890904,
+      "learning_rate": 5.7215972759641335e-06,
+      "loss": 0.4886,
+      "step": 3145
+    },
+    {
+      "epoch": 0.4704650815014207,
+      "grad_norm": 1.410091594154271,
+      "learning_rate": 5.719200719172168e-06,
+      "loss": 0.2126,
+      "step": 3146
+    },
+    {
+      "epoch": 0.4706146253925527,
+      "grad_norm": 1.2071331191053063,
+      "learning_rate": 5.716803993648784e-06,
+      "loss": 0.242,
+      "step": 3147
+    },
+    {
+      "epoch": 0.47076416928368475,
+      "grad_norm": 1.743927253751831,
+      "learning_rate": 5.7144070999562726e-06,
+      "loss": 0.4621,
+      "step": 3148
+    },
+    {
+      "epoch": 0.4709137131748168,
+      "grad_norm": 1.1162822964236923,
+      "learning_rate": 5.712010038656972e-06,
+      "loss": 0.1878,
+      "step": 3149
+    },
+    {
+      "epoch": 0.47106325706594887,
+      "grad_norm": 1.4397665068284078,
+      "learning_rate": 5.709612810313253e-06,
+      "loss": 0.244,
+      "step": 3150
+    },
+    {
+      "epoch": 0.4712128009570809,
+      "grad_norm": 1.2689183560373911,
+      "learning_rate": 5.707215415487532e-06,
+      "loss": 0.2142,
+      "step": 3151
+    },
+    {
+      "epoch": 0.47136234484821293,
+      "grad_norm": 1.2429339367434729,
+      "learning_rate": 5.704817854742257e-06,
+      "loss": 0.2077,
+      "step": 3152
+    },
+    {
+      "epoch": 0.471511888739345,
+      "grad_norm": 1.4973747904307906,
+      "learning_rate": 5.7024201286399215e-06,
+      "loss": 0.1525,
+      "step": 3153
+    },
+    {
+      "epoch": 0.47166143263047705,
+      "grad_norm": 1.5335821278579143,
+      "learning_rate": 5.700022237743055e-06,
+      "loss": 0.2315,
+      "step": 3154
+    },
+    {
+      "epoch": 0.4718109765216091,
+      "grad_norm": 1.8352653937040397,
+      "learning_rate": 5.697624182614225e-06,
+      "loss": 0.4395,
+      "step": 3155
+    },
+    {
+      "epoch": 0.4719605204127411,
+      "grad_norm": 1.9162160008419673,
+      "learning_rate": 5.695225963816038e-06,
+      "loss": 0.4633,
+      "step": 3156
+    },
+    {
+      "epoch": 0.4721100643038732,
+      "grad_norm": 1.3144121886364732,
+      "learning_rate": 5.692827581911139e-06,
+      "loss": 0.3179,
+      "step": 3157
+    },
+    {
+      "epoch": 0.47225960819500523,
+      "grad_norm": 1.6775962097688284,
+      "learning_rate": 5.690429037462213e-06,
+      "loss": 0.3079,
+      "step": 3158
+    },
+    {
+      "epoch": 0.47240915208613726,
+      "grad_norm": 1.5606712917150614,
+      "learning_rate": 5.688030331031981e-06,
+      "loss": 0.391,
+      "step": 3159
+    },
+    {
+      "epoch": 0.47255869597726935,
+      "grad_norm": 1.8951438306712434,
+      "learning_rate": 5.685631463183204e-06,
+      "loss": 0.238,
+      "step": 3160
+    },
+    {
+      "epoch": 0.4727082398684014,
+      "grad_norm": 1.4421516858300434,
+      "learning_rate": 5.683232434478675e-06,
+      "loss": 0.3104,
+      "step": 3161
+    },
+    {
+      "epoch": 0.4728577837595334,
+      "grad_norm": 1.516615298295563,
+      "learning_rate": 5.680833245481234e-06,
+      "loss": 0.2329,
+      "step": 3162
+    },
+    {
+      "epoch": 0.47300732765066544,
+      "grad_norm": 1.4316839755417867,
+      "learning_rate": 5.6784338967537524e-06,
+      "loss": 0.1996,
+      "step": 3163
+    },
+    {
+      "epoch": 0.47315687154179753,
+      "grad_norm": 1.426150093967574,
+      "learning_rate": 5.67603438885914e-06,
+      "loss": 0.2305,
+      "step": 3164
+    },
+    {
+      "epoch": 0.47330641543292956,
+      "grad_norm": 1.9888683194342018,
+      "learning_rate": 5.673634722360344e-06,
+      "loss": 0.5061,
+      "step": 3165
+    },
+    {
+      "epoch": 0.4734559593240616,
+      "grad_norm": 1.364501688300759,
+      "learning_rate": 5.6712348978203504e-06,
+      "loss": 0.2131,
+      "step": 3166
+    },
+    {
+      "epoch": 0.4736055032151937,
+      "grad_norm": 1.7957506725761962,
+      "learning_rate": 5.66883491580218e-06,
+      "loss": 0.3892,
+      "step": 3167
+    },
+    {
+      "epoch": 0.4737550471063257,
+      "grad_norm": 1.6274375743803688,
+      "learning_rate": 5.666434776868895e-06,
+      "loss": 0.2701,
+      "step": 3168
+    },
+    {
+      "epoch": 0.47390459099745774,
+      "grad_norm": 1.4211736786097835,
+      "learning_rate": 5.6640344815835866e-06,
+      "loss": 0.1947,
+      "step": 3169
+    },
+    {
+      "epoch": 0.47405413488858983,
+      "grad_norm": 1.8576338132648311,
+      "learning_rate": 5.6616340305093905e-06,
+      "loss": 0.3133,
+      "step": 3170
+    },
+    {
+      "epoch": 0.47420367877972186,
+      "grad_norm": 1.8516954583504217,
+      "learning_rate": 5.659233424209474e-06,
+      "loss": 0.3246,
+      "step": 3171
+    },
+    {
+      "epoch": 0.4743532226708539,
+      "grad_norm": 1.5765166821357868,
+      "learning_rate": 5.6568326632470424e-06,
+      "loss": 0.288,
+      "step": 3172
+    },
+    {
+      "epoch": 0.4745027665619859,
+      "grad_norm": 1.5049818589318757,
+      "learning_rate": 5.654431748185339e-06,
+      "loss": 0.2027,
+      "step": 3173
+    },
+    {
+      "epoch": 0.474652310453118,
+      "grad_norm": 1.6563418650791573,
+      "learning_rate": 5.65203067958764e-06,
+      "loss": 0.3274,
+      "step": 3174
+    },
+    {
+      "epoch": 0.47480185434425004,
+      "grad_norm": 1.4060135417396495,
+      "learning_rate": 5.649629458017261e-06,
+      "loss": 0.1846,
+      "step": 3175
+    },
+    {
+      "epoch": 0.47495139823538207,
+      "grad_norm": 1.2055453379363343,
+      "learning_rate": 5.647228084037548e-06,
+      "loss": 0.1944,
+      "step": 3176
+    },
+    {
+      "epoch": 0.47510094212651416,
+      "grad_norm": 1.32347197279457,
+      "learning_rate": 5.644826558211889e-06,
+      "loss": 0.2127,
+      "step": 3177
+    },
+    {
+      "epoch": 0.4752504860176462,
+      "grad_norm": 1.847868151137175,
+      "learning_rate": 5.642424881103706e-06,
+      "loss": 0.5842,
+      "step": 3178
+    },
+    {
+      "epoch": 0.4754000299087782,
+      "grad_norm": 1.674408388100392,
+      "learning_rate": 5.6400230532764545e-06,
+      "loss": 0.3812,
+      "step": 3179
+    },
+    {
+      "epoch": 0.47554957379991025,
+      "grad_norm": 1.757199701941517,
+      "learning_rate": 5.637621075293627e-06,
+      "loss": 0.4498,
+      "step": 3180
+    },
+    {
+      "epoch": 0.47569911769104234,
+      "grad_norm": 1.5559941460663576,
+      "learning_rate": 5.63521894771875e-06,
+      "loss": 0.2237,
+      "step": 3181
+    },
+    {
+      "epoch": 0.47584866158217437,
+      "grad_norm": 1.5462633932338483,
+      "learning_rate": 5.632816671115385e-06,
+      "loss": 0.2413,
+      "step": 3182
+    },
+    {
+      "epoch": 0.4759982054733064,
+      "grad_norm": 1.5498398967317741,
+      "learning_rate": 5.630414246047132e-06,
+      "loss": 0.3177,
+      "step": 3183
+    },
+    {
+      "epoch": 0.4761477493644385,
+      "grad_norm": 1.6167447628978575,
+      "learning_rate": 5.62801167307762e-06,
+      "loss": 0.2376,
+      "step": 3184
+    },
+    {
+      "epoch": 0.4762972932555705,
+      "grad_norm": 1.4919899689914684,
+      "learning_rate": 5.625608952770518e-06,
+      "loss": 0.1693,
+      "step": 3185
+    },
+    {
+      "epoch": 0.47644683714670255,
+      "grad_norm": 1.5067704088223868,
+      "learning_rate": 5.623206085689525e-06,
+      "loss": 0.2136,
+      "step": 3186
+    },
+    {
+      "epoch": 0.4765963810378346,
+      "grad_norm": 1.1449557548932223,
+      "learning_rate": 5.62080307239838e-06,
+      "loss": 0.1582,
+      "step": 3187
+    },
+    {
+      "epoch": 0.47674592492896667,
+      "grad_norm": 1.3257476397757155,
+      "learning_rate": 5.6183999134608514e-06,
+      "loss": 0.3345,
+      "step": 3188
+    },
+    {
+      "epoch": 0.4768954688200987,
+      "grad_norm": 1.4360228664127825,
+      "learning_rate": 5.6159966094407426e-06,
+      "loss": 0.2616,
+      "step": 3189
+    },
+    {
+      "epoch": 0.47704501271123073,
+      "grad_norm": 0.9734180829666479,
+      "learning_rate": 5.613593160901893e-06,
+      "loss": 0.1593,
+      "step": 3190
+    },
+    {
+      "epoch": 0.4771945566023628,
+      "grad_norm": 1.4248599328670821,
+      "learning_rate": 5.611189568408173e-06,
+      "loss": 0.1703,
+      "step": 3191
+    },
+    {
+      "epoch": 0.47734410049349485,
+      "grad_norm": 1.0590562637029552,
+      "learning_rate": 5.60878583252349e-06,
+      "loss": 0.2125,
+      "step": 3192
+    },
+    {
+      "epoch": 0.4774936443846269,
+      "grad_norm": 1.4611566756560919,
+      "learning_rate": 5.606381953811786e-06,
+      "loss": 0.3466,
+      "step": 3193
+    },
+    {
+      "epoch": 0.4776431882757589,
+      "grad_norm": 1.3346184917090114,
+      "learning_rate": 5.60397793283703e-06,
+      "loss": 0.2076,
+      "step": 3194
+    },
+    {
+      "epoch": 0.477792732166891,
+      "grad_norm": 1.566654209478883,
+      "learning_rate": 5.601573770163231e-06,
+      "loss": 0.3383,
+      "step": 3195
+    },
+    {
+      "epoch": 0.47794227605802303,
+      "grad_norm": 1.5414225158374537,
+      "learning_rate": 5.599169466354427e-06,
+      "loss": 0.3466,
+      "step": 3196
+    },
+    {
+      "epoch": 0.47809181994915506,
+      "grad_norm": 1.4350966493610382,
+      "learning_rate": 5.596765021974693e-06,
+      "loss": 0.2875,
+      "step": 3197
+    },
+    {
+      "epoch": 0.47824136384028715,
+      "grad_norm": 1.689845874095469,
+      "learning_rate": 5.594360437588132e-06,
+      "loss": 0.426,
+      "step": 3198
+    },
+    {
+      "epoch": 0.4783909077314192,
+      "grad_norm": 1.0977797048520335,
+      "learning_rate": 5.591955713758885e-06,
+      "loss": 0.2392,
+      "step": 3199
+    },
+    {
+      "epoch": 0.4785404516225512,
+      "grad_norm": 1.7612756142198773,
+      "learning_rate": 5.589550851051123e-06,
+      "loss": 0.5625,
+      "step": 3200
+    },
+    {
+      "epoch": 0.47868999551368324,
+      "grad_norm": 1.9558696306956667,
+      "learning_rate": 5.58714585002905e-06,
+      "loss": 0.4112,
+      "step": 3201
+    },
+    {
+      "epoch": 0.47883953940481533,
+      "grad_norm": 1.6349407082918148,
+      "learning_rate": 5.584740711256901e-06,
+      "loss": 0.2047,
+      "step": 3202
+    },
+    {
+      "epoch": 0.47898908329594736,
+      "grad_norm": 0.9418582681860653,
+      "learning_rate": 5.5823354352989465e-06,
+      "loss": 0.1597,
+      "step": 3203
+    },
+    {
+      "epoch": 0.4791386271870794,
+      "grad_norm": 1.4766455127450906,
+      "learning_rate": 5.579930022719486e-06,
+      "loss": 0.2793,
+      "step": 3204
+    },
+    {
+      "epoch": 0.4792881710782115,
+      "grad_norm": 1.654406135220705,
+      "learning_rate": 5.577524474082853e-06,
+      "loss": 0.2065,
+      "step": 3205
+    },
+    {
+      "epoch": 0.4794377149693435,
+      "grad_norm": 1.6127674887270247,
+      "learning_rate": 5.575118789953414e-06,
+      "loss": 0.3119,
+      "step": 3206
+    },
+    {
+      "epoch": 0.47958725886047554,
+      "grad_norm": 1.3677004533072454,
+      "learning_rate": 5.572712970895563e-06,
+      "loss": 0.2697,
+      "step": 3207
+    },
+    {
+      "epoch": 0.47973680275160757,
+      "grad_norm": 1.3293568676356242,
+      "learning_rate": 5.5703070174737305e-06,
+      "loss": 0.3024,
+      "step": 3208
+    },
+    {
+      "epoch": 0.47988634664273966,
+      "grad_norm": 1.5812284057258443,
+      "learning_rate": 5.567900930252375e-06,
+      "loss": 0.2966,
+      "step": 3209
+    },
+    {
+      "epoch": 0.4800358905338717,
+      "grad_norm": 1.3637746072600263,
+      "learning_rate": 5.565494709795988e-06,
+      "loss": 0.2399,
+      "step": 3210
+    },
+    {
+      "epoch": 0.4801854344250037,
+      "grad_norm": 1.2403212239051626,
+      "learning_rate": 5.5630883566690915e-06,
+      "loss": 0.2032,
+      "step": 3211
+    },
+    {
+      "epoch": 0.4803349783161358,
+      "grad_norm": 1.3275458152883233,
+      "learning_rate": 5.5606818714362406e-06,
+      "loss": 0.1753,
+      "step": 3212
+    },
+    {
+      "epoch": 0.48048452220726784,
+      "grad_norm": 1.6827685464628404,
+      "learning_rate": 5.55827525466202e-06,
+      "loss": 0.3836,
+      "step": 3213
+    },
+    {
+      "epoch": 0.48063406609839987,
+      "grad_norm": 1.5010871722356298,
+      "learning_rate": 5.5558685069110444e-06,
+      "loss": 0.3594,
+      "step": 3214
+    },
+    {
+      "epoch": 0.4807836099895319,
+      "grad_norm": 1.7500323777950166,
+      "learning_rate": 5.5534616287479585e-06,
+      "loss": 0.4294,
+      "step": 3215
+    },
+    {
+      "epoch": 0.480933153880664,
+      "grad_norm": 1.3283210890130643,
+      "learning_rate": 5.5510546207374415e-06,
+      "loss": 0.26,
+      "step": 3216
+    },
+    {
+      "epoch": 0.481082697771796,
+      "grad_norm": 1.6922301032567173,
+      "learning_rate": 5.548647483444199e-06,
+      "loss": 0.2203,
+      "step": 3217
+    },
+    {
+      "epoch": 0.48123224166292805,
+      "grad_norm": 1.3062615721477455,
+      "learning_rate": 5.54624021743297e-06,
+      "loss": 0.3097,
+      "step": 3218
+    },
+    {
+      "epoch": 0.48138178555406014,
+      "grad_norm": 1.346469662217694,
+      "learning_rate": 5.543832823268522e-06,
+      "loss": 0.2069,
+      "step": 3219
+    },
+    {
+      "epoch": 0.48153132944519217,
+      "grad_norm": 0.9633365182885122,
+      "learning_rate": 5.541425301515652e-06,
+      "loss": 0.1913,
+      "step": 3220
+    },
+    {
+      "epoch": 0.4816808733363242,
+      "grad_norm": 1.9220717530990858,
+      "learning_rate": 5.5390176527391895e-06,
+      "loss": 0.4657,
+      "step": 3221
+    },
+    {
+      "epoch": 0.48183041722745623,
+      "grad_norm": 1.5111193343359333,
+      "learning_rate": 5.536609877503989e-06,
+      "loss": 0.3328,
+      "step": 3222
+    },
+    {
+      "epoch": 0.4819799611185883,
+      "grad_norm": 1.952680430360637,
+      "learning_rate": 5.534201976374941e-06,
+      "loss": 0.5342,
+      "step": 3223
+    },
+    {
+      "epoch": 0.48212950500972035,
+      "grad_norm": 1.5983344593704398,
+      "learning_rate": 5.531793949916961e-06,
+      "loss": 0.2425,
+      "step": 3224
+    },
+    {
+      "epoch": 0.4822790489008524,
+      "grad_norm": 1.8572599659337972,
+      "learning_rate": 5.529385798694995e-06,
+      "loss": 0.5123,
+      "step": 3225
+    },
+    {
+      "epoch": 0.48242859279198447,
+      "grad_norm": 1.424752198495054,
+      "learning_rate": 5.526977523274017e-06,
+      "loss": 0.1894,
+      "step": 3226
+    },
+    {
+      "epoch": 0.4825781366831165,
+      "grad_norm": 1.3362880994059592,
+      "learning_rate": 5.524569124219036e-06,
+      "loss": 0.2178,
+      "step": 3227
+    },
+    {
+      "epoch": 0.48272768057424853,
+      "grad_norm": 1.5846787140906795,
+      "learning_rate": 5.522160602095082e-06,
+      "loss": 0.3475,
+      "step": 3228
+    },
+    {
+      "epoch": 0.4828772244653806,
+      "grad_norm": 1.9665234251579087,
+      "learning_rate": 5.5197519574672185e-06,
+      "loss": 0.5254,
+      "step": 3229
+    },
+    {
+      "epoch": 0.48302676835651265,
+      "grad_norm": 1.2406935735339304,
+      "learning_rate": 5.517343190900537e-06,
+      "loss": 0.196,
+      "step": 3230
+    },
+    {
+      "epoch": 0.4831763122476447,
+      "grad_norm": 1.2840122911547907,
+      "learning_rate": 5.514934302960156e-06,
+      "loss": 0.3584,
+      "step": 3231
+    },
+    {
+      "epoch": 0.4833258561387767,
+      "grad_norm": 1.7114869197966367,
+      "learning_rate": 5.512525294211226e-06,
+      "loss": 0.5087,
+      "step": 3232
+    },
+    {
+      "epoch": 0.4834754000299088,
+      "grad_norm": 1.6167473170926372,
+      "learning_rate": 5.5101161652189225e-06,
+      "loss": 0.3207,
+      "step": 3233
+    },
+    {
+      "epoch": 0.48362494392104083,
+      "grad_norm": 1.1172175670125728,
+      "learning_rate": 5.507706916548451e-06,
+      "loss": 0.197,
+      "step": 3234
+    },
+    {
+      "epoch": 0.48377448781217286,
+      "grad_norm": 1.0454088270739283,
+      "learning_rate": 5.505297548765043e-06,
+      "loss": 0.244,
+      "step": 3235
+    },
+    {
+      "epoch": 0.48392403170330495,
+      "grad_norm": 2.1427540459127905,
+      "learning_rate": 5.502888062433961e-06,
+      "loss": 0.401,
+      "step": 3236
+    },
+    {
+      "epoch": 0.484073575594437,
+      "grad_norm": 1.483691258814316,
+      "learning_rate": 5.500478458120493e-06,
+      "loss": 0.3204,
+      "step": 3237
+    },
+    {
+      "epoch": 0.484223119485569,
+      "grad_norm": 1.4985182367477778,
+      "learning_rate": 5.4980687363899545e-06,
+      "loss": 0.1841,
+      "step": 3238
+    },
+    {
+      "epoch": 0.48437266337670104,
+      "grad_norm": 1.5306187126370105,
+      "learning_rate": 5.495658897807691e-06,
+      "loss": 0.3307,
+      "step": 3239
+    },
+    {
+      "epoch": 0.4845222072678331,
+      "grad_norm": 1.9059741922872409,
+      "learning_rate": 5.4932489429390725e-06,
+      "loss": 0.4811,
+      "step": 3240
+    },
+    {
+      "epoch": 0.48467175115896516,
+      "grad_norm": 1.5997735336338483,
+      "learning_rate": 5.4908388723495e-06,
+      "loss": 0.4974,
+      "step": 3241
+    },
+    {
+      "epoch": 0.4848212950500972,
+      "grad_norm": 1.1574319053701934,
+      "learning_rate": 5.488428686604397e-06,
+      "loss": 0.1849,
+      "step": 3242
+    },
+    {
+      "epoch": 0.4849708389412293,
+      "grad_norm": 1.7158993652318315,
+      "learning_rate": 5.486018386269216e-06,
+      "loss": 0.1957,
+      "step": 3243
+    },
+    {
+      "epoch": 0.4851203828323613,
+      "grad_norm": 1.3803613849267424,
+      "learning_rate": 5.4836079719094395e-06,
+      "loss": 0.229,
+      "step": 3244
+    },
+    {
+      "epoch": 0.48526992672349334,
+      "grad_norm": 1.40606103461945,
+      "learning_rate": 5.4811974440905705e-06,
+      "loss": 0.1864,
+      "step": 3245
+    },
+    {
+      "epoch": 0.48541947061462537,
+      "grad_norm": 0.9610421038503347,
+      "learning_rate": 5.478786803378143e-06,
+      "loss": 0.2373,
+      "step": 3246
+    },
+    {
+      "epoch": 0.48556901450575746,
+      "grad_norm": 1.415230316232099,
+      "learning_rate": 5.476376050337719e-06,
+      "loss": 0.3828,
+      "step": 3247
+    },
+    {
+      "epoch": 0.4857185583968895,
+      "grad_norm": 2.307432712150402,
+      "learning_rate": 5.473965185534882e-06,
+      "loss": 0.3894,
+      "step": 3248
+    },
+    {
+      "epoch": 0.4858681022880215,
+      "grad_norm": 1.4880888031014317,
+      "learning_rate": 5.471554209535244e-06,
+      "loss": 0.2822,
+      "step": 3249
+    },
+    {
+      "epoch": 0.4860176461791536,
+      "grad_norm": 1.2863532256194963,
+      "learning_rate": 5.469143122904444e-06,
+      "loss": 0.2129,
+      "step": 3250
+    },
+    {
+      "epoch": 0.48616719007028564,
+      "grad_norm": 1.6475188307887112,
+      "learning_rate": 5.466731926208148e-06,
+      "loss": 0.2088,
+      "step": 3251
+    },
+    {
+      "epoch": 0.48631673396141767,
+      "grad_norm": 1.5875387799039424,
+      "learning_rate": 5.46432062001204e-06,
+      "loss": 0.3401,
+      "step": 3252
+    },
+    {
+      "epoch": 0.4864662778525497,
+      "grad_norm": 1.8220640993048132,
+      "learning_rate": 5.461909204881842e-06,
+      "loss": 0.2542,
+      "step": 3253
+    },
+    {
+      "epoch": 0.4866158217436818,
+      "grad_norm": 1.233777671255329,
+      "learning_rate": 5.459497681383295e-06,
+      "loss": 0.2139,
+      "step": 3254
+    },
+    {
+      "epoch": 0.4867653656348138,
+      "grad_norm": 1.734386302292572,
+      "learning_rate": 5.457086050082161e-06,
+      "loss": 0.2322,
+      "step": 3255
+    },
+    {
+      "epoch": 0.48691490952594585,
+      "grad_norm": 1.2129989858188304,
+      "learning_rate": 5.454674311544236e-06,
+      "loss": 0.2004,
+      "step": 3256
+    },
+    {
+      "epoch": 0.48706445341707794,
+      "grad_norm": 1.5232752004392356,
+      "learning_rate": 5.4522624663353356e-06,
+      "loss": 0.3088,
+      "step": 3257
+    },
+    {
+      "epoch": 0.48721399730820997,
+      "grad_norm": 1.0342508423040848,
+      "learning_rate": 5.4498505150213e-06,
+      "loss": 0.19,
+      "step": 3258
+    },
+    {
+      "epoch": 0.487363541199342,
+      "grad_norm": 1.2149206479073549,
+      "learning_rate": 5.447438458168e-06,
+      "loss": 0.1917,
+      "step": 3259
+    },
+    {
+      "epoch": 0.48751308509047403,
+      "grad_norm": 1.8186709488406252,
+      "learning_rate": 5.445026296341325e-06,
+      "loss": 0.4124,
+      "step": 3260
+    },
+    {
+      "epoch": 0.4876626289816061,
+      "grad_norm": 1.4871462802100281,
+      "learning_rate": 5.442614030107192e-06,
+      "loss": 0.3458,
+      "step": 3261
+    },
+    {
+      "epoch": 0.48781217287273815,
+      "grad_norm": 1.3202962648646308,
+      "learning_rate": 5.440201660031542e-06,
+      "loss": 0.3653,
+      "step": 3262
+    },
+    {
+      "epoch": 0.4879617167638702,
+      "grad_norm": 2.336083852527424,
+      "learning_rate": 5.4377891866803405e-06,
+      "loss": 0.2729,
+      "step": 3263
+    },
+    {
+      "epoch": 0.48811126065500227,
+      "grad_norm": 1.2930049790322846,
+      "learning_rate": 5.435376610619577e-06,
+      "loss": 0.2439,
+      "step": 3264
+    },
+    {
+      "epoch": 0.4882608045461343,
+      "grad_norm": 1.557892290174439,
+      "learning_rate": 5.432963932415264e-06,
+      "loss": 0.3053,
+      "step": 3265
+    },
+    {
+      "epoch": 0.48841034843726633,
+      "grad_norm": 1.3905071948011742,
+      "learning_rate": 5.430551152633438e-06,
+      "loss": 0.214,
+      "step": 3266
+    },
+    {
+      "epoch": 0.48855989232839836,
+      "grad_norm": 1.8231907139006023,
+      "learning_rate": 5.428138271840165e-06,
+      "loss": 0.4806,
+      "step": 3267
+    },
+    {
+      "epoch": 0.48870943621953045,
+      "grad_norm": 1.5732546852605194,
+      "learning_rate": 5.425725290601527e-06,
+      "loss": 0.3985,
+      "step": 3268
+    },
+    {
+      "epoch": 0.4888589801106625,
+      "grad_norm": 2.677639259079234,
+      "learning_rate": 5.423312209483632e-06,
+      "loss": 0.4894,
+      "step": 3269
+    },
+    {
+      "epoch": 0.4890085240017945,
+      "grad_norm": 1.3541375970165608,
+      "learning_rate": 5.420899029052614e-06,
+      "loss": 0.1714,
+      "step": 3270
+    },
+    {
+      "epoch": 0.4891580678929266,
+      "grad_norm": 1.7576357519511228,
+      "learning_rate": 5.418485749874628e-06,
+      "loss": 0.2243,
+      "step": 3271
+    },
+    {
+      "epoch": 0.4893076117840586,
+      "grad_norm": 1.4302229998846714,
+      "learning_rate": 5.41607237251585e-06,
+      "loss": 0.3575,
+      "step": 3272
+    },
+    {
+      "epoch": 0.48945715567519066,
+      "grad_norm": 1.357123434068504,
+      "learning_rate": 5.413658897542483e-06,
+      "loss": 0.2934,
+      "step": 3273
+    },
+    {
+      "epoch": 0.4896066995663227,
+      "grad_norm": 1.620224094370209,
+      "learning_rate": 5.411245325520754e-06,
+      "loss": 0.3828,
+      "step": 3274
+    },
+    {
+      "epoch": 0.4897562434574548,
+      "grad_norm": 1.4222644545024612,
+      "learning_rate": 5.408831657016908e-06,
+      "loss": 0.2534,
+      "step": 3275
+    },
+    {
+      "epoch": 0.4899057873485868,
+      "grad_norm": 1.6038088280717635,
+      "learning_rate": 5.4064178925972155e-06,
+      "loss": 0.3653,
+      "step": 3276
+    },
+    {
+      "epoch": 0.49005533123971884,
+      "grad_norm": 6.225442055629013,
+      "learning_rate": 5.4040040328279684e-06,
+      "loss": 0.4034,
+      "step": 3277
+    },
+    {
+      "epoch": 0.4902048751308509,
+      "grad_norm": 1.3073622659475577,
+      "learning_rate": 5.40159007827548e-06,
+      "loss": 0.2864,
+      "step": 3278
+    },
+    {
+      "epoch": 0.49035441902198296,
+      "grad_norm": 2.0634805595333963,
+      "learning_rate": 5.399176029506091e-06,
+      "loss": 0.3318,
+      "step": 3279
+    },
+    {
+      "epoch": 0.490503962913115,
+      "grad_norm": 1.6410728978525977,
+      "learning_rate": 5.396761887086157e-06,
+      "loss": 0.3503,
+      "step": 3280
+    },
+    {
+      "epoch": 0.490653506804247,
+      "grad_norm": 1.1385654609121518,
+      "learning_rate": 5.394347651582063e-06,
+      "loss": 0.1922,
+      "step": 3281
+    },
+    {
+      "epoch": 0.4908030506953791,
+      "grad_norm": 1.7005263723270547,
+      "learning_rate": 5.391933323560208e-06,
+      "loss": 0.441,
+      "step": 3282
+    },
+    {
+      "epoch": 0.49095259458651114,
+      "grad_norm": 1.770272350276855,
+      "learning_rate": 5.389518903587016e-06,
+      "loss": 0.3787,
+      "step": 3283
+    },
+    {
+      "epoch": 0.49110213847764317,
+      "grad_norm": 1.3506689259925078,
+      "learning_rate": 5.387104392228939e-06,
+      "loss": 0.2533,
+      "step": 3284
+    },
+    {
+      "epoch": 0.49125168236877526,
+      "grad_norm": 1.3898273923789393,
+      "learning_rate": 5.384689790052439e-06,
+      "loss": 0.3693,
+      "step": 3285
+    },
+    {
+      "epoch": 0.4914012262599073,
+      "grad_norm": 1.9579535041564384,
+      "learning_rate": 5.3822750976240065e-06,
+      "loss": 0.4891,
+      "step": 3286
+    },
+    {
+      "epoch": 0.4915507701510393,
+      "grad_norm": 1.6436155900647393,
+      "learning_rate": 5.379860315510153e-06,
+      "loss": 0.441,
+      "step": 3287
+    },
+    {
+      "epoch": 0.4917003140421714,
+      "grad_norm": 1.5550148010776086,
+      "learning_rate": 5.377445444277407e-06,
+      "loss": 0.3143,
+      "step": 3288
+    },
+    {
+      "epoch": 0.49184985793330344,
+      "grad_norm": 1.849270778547779,
+      "learning_rate": 5.375030484492324e-06,
+      "loss": 0.2501,
+      "step": 3289
+    },
+    {
+      "epoch": 0.49199940182443547,
+      "grad_norm": 1.5318913479060694,
+      "learning_rate": 5.372615436721474e-06,
+      "loss": 0.3657,
+      "step": 3290
+    },
+    {
+      "epoch": 0.4921489457155675,
+      "grad_norm": 2.1556436681340028,
+      "learning_rate": 5.370200301531453e-06,
+      "loss": 0.541,
+      "step": 3291
+    },
+    {
+      "epoch": 0.4922984896066996,
+      "grad_norm": 1.5606790446050516,
+      "learning_rate": 5.3677850794888704e-06,
+      "loss": 0.3051,
+      "step": 3292
+    },
+    {
+      "epoch": 0.4924480334978316,
+      "grad_norm": 1.384057987332389,
+      "learning_rate": 5.365369771160366e-06,
+      "loss": 0.1973,
+      "step": 3293
+    },
+    {
+      "epoch": 0.49259757738896365,
+      "grad_norm": 1.15291598303946,
+      "learning_rate": 5.362954377112593e-06,
+      "loss": 0.1559,
+      "step": 3294
+    },
+    {
+      "epoch": 0.49274712128009573,
+      "grad_norm": 1.265068756127247,
+      "learning_rate": 5.360538897912223e-06,
+      "loss": 0.2327,
+      "step": 3295
+    },
+    {
+      "epoch": 0.49289666517122777,
+      "grad_norm": 1.6658809938402388,
+      "learning_rate": 5.358123334125953e-06,
+      "loss": 0.3179,
+      "step": 3296
+    },
+    {
+      "epoch": 0.4930462090623598,
+      "grad_norm": 1.8651590421952458,
+      "learning_rate": 5.355707686320499e-06,
+      "loss": 0.3796,
+      "step": 3297
+    },
+    {
+      "epoch": 0.49319575295349183,
+      "grad_norm": 2.0781278430947956,
+      "learning_rate": 5.353291955062594e-06,
+      "loss": 0.3463,
+      "step": 3298
+    },
+    {
+      "epoch": 0.4933452968446239,
+      "grad_norm": 1.3667567419203674,
+      "learning_rate": 5.350876140918989e-06,
+      "loss": 0.1763,
+      "step": 3299
+    },
+    {
+      "epoch": 0.49349484073575595,
+      "grad_norm": 1.5793291775977985,
+      "learning_rate": 5.348460244456462e-06,
+      "loss": 0.3639,
+      "step": 3300
+    },
+    {
+      "epoch": 0.493644384626888,
+      "grad_norm": 1.436284245662927,
+      "learning_rate": 5.346044266241802e-06,
+      "loss": 0.2976,
+      "step": 3301
+    },
+    {
+      "epoch": 0.49379392851802006,
+      "grad_norm": 1.573325582954275,
+      "learning_rate": 5.3436282068418245e-06,
+      "loss": 0.3671,
+      "step": 3302
+    },
+    {
+      "epoch": 0.4939434724091521,
+      "grad_norm": 1.4650670097603558,
+      "learning_rate": 5.341212066823356e-06,
+      "loss": 0.197,
+      "step": 3303
+    },
+    {
+      "epoch": 0.4940930163002841,
+      "grad_norm": 1.9818504468304579,
+      "learning_rate": 5.338795846753249e-06,
+      "loss": 0.2051,
+      "step": 3304
+    },
+    {
+      "epoch": 0.49424256019141616,
+      "grad_norm": 2.0426688110857087,
+      "learning_rate": 5.336379547198371e-06,
+      "loss": 0.4497,
+      "step": 3305
+    },
+    {
+      "epoch": 0.49439210408254824,
+      "grad_norm": 1.9574028012684885,
+      "learning_rate": 5.3339631687256085e-06,
+      "loss": 0.4708,
+      "step": 3306
+    },
+    {
+      "epoch": 0.4945416479736803,
+      "grad_norm": 1.575961759570781,
+      "learning_rate": 5.331546711901869e-06,
+      "loss": 0.3804,
+      "step": 3307
+    },
+    {
+      "epoch": 0.4946911918648123,
+      "grad_norm": 1.414716802354248,
+      "learning_rate": 5.3291301772940755e-06,
+      "loss": 0.2485,
+      "step": 3308
+    },
+    {
+      "epoch": 0.4948407357559444,
+      "grad_norm": 1.240412229265847,
+      "learning_rate": 5.32671356546917e-06,
+      "loss": 0.2153,
+      "step": 3309
+    },
+    {
+      "epoch": 0.4949902796470764,
+      "grad_norm": 1.8834615593768038,
+      "learning_rate": 5.324296876994115e-06,
+      "loss": 0.4081,
+      "step": 3310
+    },
+    {
+      "epoch": 0.49513982353820846,
+      "grad_norm": 2.0275002053515876,
+      "learning_rate": 5.321880112435885e-06,
+      "loss": 0.5748,
+      "step": 3311
+    },
+    {
+      "epoch": 0.4952893674293405,
+      "grad_norm": 1.1895625858020245,
+      "learning_rate": 5.31946327236148e-06,
+      "loss": 0.1955,
+      "step": 3312
+    },
+    {
+      "epoch": 0.4954389113204726,
+      "grad_norm": 1.6692053077282905,
+      "learning_rate": 5.3170463573379115e-06,
+      "loss": 0.3985,
+      "step": 3313
+    },
+    {
+      "epoch": 0.4955884552116046,
+      "grad_norm": 1.1693085573761919,
+      "learning_rate": 5.314629367932213e-06,
+      "loss": 0.201,
+      "step": 3314
+    },
+    {
+      "epoch": 0.49573799910273664,
+      "grad_norm": 1.607561425215323,
+      "learning_rate": 5.3122123047114334e-06,
+      "loss": 0.4755,
+      "step": 3315
+    },
+    {
+      "epoch": 0.4958875429938687,
+      "grad_norm": 1.6244993103305614,
+      "learning_rate": 5.3097951682426375e-06,
+      "loss": 0.2829,
+      "step": 3316
+    },
+    {
+      "epoch": 0.49603708688500076,
+      "grad_norm": 1.335438428761969,
+      "learning_rate": 5.30737795909291e-06,
+      "loss": 0.3357,
+      "step": 3317
+    },
+    {
+      "epoch": 0.4961866307761328,
+      "grad_norm": 1.1870501993360763,
+      "learning_rate": 5.304960677829349e-06,
+      "loss": 0.2321,
+      "step": 3318
+    },
+    {
+      "epoch": 0.4963361746672648,
+      "grad_norm": 1.2459003700531992,
+      "learning_rate": 5.302543325019075e-06,
+      "loss": 0.1832,
+      "step": 3319
+    },
+    {
+      "epoch": 0.4964857185583969,
+      "grad_norm": 1.174928428747797,
+      "learning_rate": 5.300125901229222e-06,
+      "loss": 0.1906,
+      "step": 3320
+    },
+    {
+      "epoch": 0.49663526244952894,
+      "grad_norm": 1.312638962292064,
+      "learning_rate": 5.297708407026939e-06,
+      "loss": 0.1821,
+      "step": 3321
+    },
+    {
+      "epoch": 0.49678480634066097,
+      "grad_norm": 1.8127687148152816,
+      "learning_rate": 5.295290842979396e-06,
+      "loss": 0.3657,
+      "step": 3322
+    },
+    {
+      "epoch": 0.49693435023179305,
+      "grad_norm": 1.5846846945588664,
+      "learning_rate": 5.292873209653774e-06,
+      "loss": 0.4235,
+      "step": 3323
+    },
+    {
+      "epoch": 0.4970838941229251,
+      "grad_norm": 1.4706523678799026,
+      "learning_rate": 5.290455507617276e-06,
+      "loss": 0.3334,
+      "step": 3324
+    },
+    {
+      "epoch": 0.4972334380140571,
+      "grad_norm": 1.703735745140162,
+      "learning_rate": 5.288037737437115e-06,
+      "loss": 0.3885,
+      "step": 3325
+    },
+    {
+      "epoch": 0.49738298190518915,
+      "grad_norm": 1.376769352546737,
+      "learning_rate": 5.285619899680527e-06,
+      "loss": 0.3308,
+      "step": 3326
+    },
+    {
+      "epoch": 0.49753252579632123,
+      "grad_norm": 1.317011044800177,
+      "learning_rate": 5.283201994914755e-06,
+      "loss": 0.2099,
+      "step": 3327
+    },
+    {
+      "epoch": 0.49768206968745327,
+      "grad_norm": 1.681669403343869,
+      "learning_rate": 5.280784023707067e-06,
+      "loss": 0.208,
+      "step": 3328
+    },
+    {
+      "epoch": 0.4978316135785853,
+      "grad_norm": 1.5061732433872248,
+      "learning_rate": 5.278365986624743e-06,
+      "loss": 0.1759,
+      "step": 3329
+    },
+    {
+      "epoch": 0.4979811574697174,
+      "grad_norm": 1.356987051993483,
+      "learning_rate": 5.275947884235073e-06,
+      "loss": 0.2116,
+      "step": 3330
+    },
+    {
+      "epoch": 0.4981307013608494,
+      "grad_norm": 1.307261358026284,
+      "learning_rate": 5.2735297171053714e-06,
+      "loss": 0.2082,
+      "step": 3331
+    },
+    {
+      "epoch": 0.49828024525198145,
+      "grad_norm": 1.799388885073472,
+      "learning_rate": 5.271111485802962e-06,
+      "loss": 0.4406,
+      "step": 3332
+    },
+    {
+      "epoch": 0.4984297891431135,
+      "grad_norm": 1.5618056900457256,
+      "learning_rate": 5.2686931908951835e-06,
+      "loss": 0.3651,
+      "step": 3333
+    },
+    {
+      "epoch": 0.49857933303424556,
+      "grad_norm": 1.6901599045091793,
+      "learning_rate": 5.266274832949395e-06,
+      "loss": 0.4992,
+      "step": 3334
+    },
+    {
+      "epoch": 0.4987288769253776,
+      "grad_norm": 1.4954460375400984,
+      "learning_rate": 5.263856412532964e-06,
+      "loss": 0.3267,
+      "step": 3335
+    },
+    {
+      "epoch": 0.4988784208165096,
+      "grad_norm": 1.2517059338039354,
+      "learning_rate": 5.261437930213275e-06,
+      "loss": 0.178,
+      "step": 3336
+    },
+    {
+      "epoch": 0.4990279647076417,
+      "grad_norm": 1.1344813829644422,
+      "learning_rate": 5.259019386557729e-06,
+      "loss": 0.2142,
+      "step": 3337
+    },
+    {
+      "epoch": 0.49917750859877374,
+      "grad_norm": 2.101616933181502,
+      "learning_rate": 5.256600782133738e-06,
+      "loss": 0.5949,
+      "step": 3338
+    },
+    {
+      "epoch": 0.4993270524899058,
+      "grad_norm": 1.4383620186111616,
+      "learning_rate": 5.25418211750873e-06,
+      "loss": 0.349,
+      "step": 3339
+    },
+    {
+      "epoch": 0.4994765963810378,
+      "grad_norm": 1.699016778277064,
+      "learning_rate": 5.251763393250149e-06,
+      "loss": 0.401,
+      "step": 3340
+    },
+    {
+      "epoch": 0.4996261402721699,
+      "grad_norm": 1.0482102710690397,
+      "learning_rate": 5.249344609925449e-06,
+      "loss": 0.1841,
+      "step": 3341
+    },
+    {
+      "epoch": 0.4997756841633019,
+      "grad_norm": 1.7173296868671082,
+      "learning_rate": 5.246925768102101e-06,
+      "loss": 0.4094,
+      "step": 3342
+    },
+    {
+      "epoch": 0.49992522805443396,
+      "grad_norm": 1.4164669080686887,
+      "learning_rate": 5.244506868347588e-06,
+      "loss": 0.2179,
+      "step": 3343
+    },
+    {
+      "epoch": 0.500074771945566,
+      "grad_norm": 1.531411863063166,
+      "learning_rate": 5.242087911229405e-06,
+      "loss": 0.3356,
+      "step": 3344
+    },
+    {
+      "epoch": 0.500224315836698,
+      "grad_norm": 1.135792810651757,
+      "learning_rate": 5.239668897315068e-06,
+      "loss": 0.1927,
+      "step": 3345
+    },
+    {
+      "epoch": 0.5003738597278301,
+      "grad_norm": 1.6441864050801602,
+      "learning_rate": 5.237249827172096e-06,
+      "loss": 0.4575,
+      "step": 3346
+    },
+    {
+      "epoch": 0.5005234036189622,
+      "grad_norm": 1.162852279035988,
+      "learning_rate": 5.234830701368029e-06,
+      "loss": 0.1545,
+      "step": 3347
+    },
+    {
+      "epoch": 0.5006729475100942,
+      "grad_norm": 1.410269599279374,
+      "learning_rate": 5.232411520470416e-06,
+      "loss": 0.2162,
+      "step": 3348
+    },
+    {
+      "epoch": 0.5008224914012263,
+      "grad_norm": 1.745801598261214,
+      "learning_rate": 5.2299922850468195e-06,
+      "loss": 0.5337,
+      "step": 3349
+    },
+    {
+      "epoch": 0.5009720352923583,
+      "grad_norm": 2.176023856699294,
+      "learning_rate": 5.227572995664819e-06,
+      "loss": 0.4114,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5011215791834903,
+      "grad_norm": 2.3007951519065246,
+      "learning_rate": 5.225153652891999e-06,
+      "loss": 0.4568,
+      "step": 3351
+    },
+    {
+      "epoch": 0.5012711230746224,
+      "grad_norm": 1.8594233682896886,
+      "learning_rate": 5.222734257295963e-06,
+      "loss": 0.4813,
+      "step": 3352
+    },
+    {
+      "epoch": 0.5014206669657545,
+      "grad_norm": 1.4068832031512468,
+      "learning_rate": 5.220314809444321e-06,
+      "loss": 0.3328,
+      "step": 3353
+    },
+    {
+      "epoch": 0.5015702108568865,
+      "grad_norm": 1.67853079379061,
+      "learning_rate": 5.217895309904703e-06,
+      "loss": 0.2757,
+      "step": 3354
+    },
+    {
+      "epoch": 0.5017197547480186,
+      "grad_norm": 1.6234161746374836,
+      "learning_rate": 5.215475759244746e-06,
+      "loss": 0.2088,
+      "step": 3355
+    },
+    {
+      "epoch": 0.5018692986391506,
+      "grad_norm": 1.4803686485569791,
+      "learning_rate": 5.2130561580321e-06,
+      "loss": 0.3738,
+      "step": 3356
+    },
+    {
+      "epoch": 0.5020188425302826,
+      "grad_norm": 1.7103662059610552,
+      "learning_rate": 5.210636506834425e-06,
+      "loss": 0.2087,
+      "step": 3357
+    },
+    {
+      "epoch": 0.5021683864214147,
+      "grad_norm": 1.7934033193420236,
+      "learning_rate": 5.208216806219395e-06,
+      "loss": 0.5237,
+      "step": 3358
+    },
+    {
+      "epoch": 0.5023179303125467,
+      "grad_norm": 2.0124390173788775,
+      "learning_rate": 5.205797056754696e-06,
+      "loss": 0.3567,
+      "step": 3359
+    },
+    {
+      "epoch": 0.5024674742036788,
+      "grad_norm": 1.644508950632731,
+      "learning_rate": 5.203377259008024e-06,
+      "loss": 0.3472,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5026170180948109,
+      "grad_norm": 1.4189985308607627,
+      "learning_rate": 5.200957413547086e-06,
+      "loss": 0.1959,
+      "step": 3361
+    },
+    {
+      "epoch": 0.5027665619859428,
+      "grad_norm": 2.013951064845426,
+      "learning_rate": 5.1985375209396035e-06,
+      "loss": 0.5964,
+      "step": 3362
+    },
+    {
+      "epoch": 0.5029161058770749,
+      "grad_norm": 1.063791871070481,
+      "learning_rate": 5.196117581753305e-06,
+      "loss": 0.171,
+      "step": 3363
+    },
+    {
+      "epoch": 0.503065649768207,
+      "grad_norm": 1.8308256692715812,
+      "learning_rate": 5.193697596555932e-06,
+      "loss": 0.44,
+      "step": 3364
+    },
+    {
+      "epoch": 0.503215193659339,
+      "grad_norm": 1.8386517447639408,
+      "learning_rate": 5.191277565915236e-06,
+      "loss": 0.6374,
+      "step": 3365
+    },
+    {
+      "epoch": 0.5033647375504711,
+      "grad_norm": 1.9183016447353525,
+      "learning_rate": 5.18885749039898e-06,
+      "loss": 0.3919,
+      "step": 3366
+    },
+    {
+      "epoch": 0.5035142814416032,
+      "grad_norm": 2.0616577191442653,
+      "learning_rate": 5.186437370574935e-06,
+      "loss": 0.2469,
+      "step": 3367
+    },
+    {
+      "epoch": 0.5036638253327351,
+      "grad_norm": 1.7802773589123986,
+      "learning_rate": 5.184017207010887e-06,
+      "loss": 0.2157,
+      "step": 3368
+    },
+    {
+      "epoch": 0.5038133692238672,
+      "grad_norm": 1.7787158574348014,
+      "learning_rate": 5.181597000274628e-06,
+      "loss": 0.2512,
+      "step": 3369
+    },
+    {
+      "epoch": 0.5039629131149993,
+      "grad_norm": 1.5473819922931655,
+      "learning_rate": 5.179176750933965e-06,
+      "loss": 0.4162,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5041124570061313,
+      "grad_norm": 2.1562775008894244,
+      "learning_rate": 5.176756459556709e-06,
+      "loss": 0.3656,
+      "step": 3371
+    },
+    {
+      "epoch": 0.5042620008972634,
+      "grad_norm": 1.4364564466383367,
+      "learning_rate": 5.174336126710683e-06,
+      "loss": 0.3441,
+      "step": 3372
+    },
+    {
+      "epoch": 0.5044115447883954,
+      "grad_norm": 1.6456126737151073,
+      "learning_rate": 5.171915752963721e-06,
+      "loss": 0.3596,
+      "step": 3373
+    },
+    {
+      "epoch": 0.5045610886795274,
+      "grad_norm": 1.9483420592597502,
+      "learning_rate": 5.169495338883666e-06,
+      "loss": 0.4743,
+      "step": 3374
+    },
+    {
+      "epoch": 0.5047106325706595,
+      "grad_norm": 1.649657426297314,
+      "learning_rate": 5.1670748850383734e-06,
+      "loss": 0.2528,
+      "step": 3375
+    },
+    {
+      "epoch": 0.5048601764617915,
+      "grad_norm": 1.6332145086519905,
+      "learning_rate": 5.164654391995702e-06,
+      "loss": 0.4456,
+      "step": 3376
+    },
+    {
+      "epoch": 0.5050097203529236,
+      "grad_norm": 1.6522420757205838,
+      "learning_rate": 5.162233860323523e-06,
+      "loss": 0.3963,
+      "step": 3377
+    },
+    {
+      "epoch": 0.5051592642440557,
+      "grad_norm": 1.2741254428541653,
+      "learning_rate": 5.159813290589717e-06,
+      "loss": 0.2927,
+      "step": 3378
+    },
+    {
+      "epoch": 0.5053088081351876,
+      "grad_norm": 1.0038407207877114,
+      "learning_rate": 5.1573926833621724e-06,
+      "loss": 0.1635,
+      "step": 3379
+    },
+    {
+      "epoch": 0.5054583520263197,
+      "grad_norm": 1.1748485985062411,
+      "learning_rate": 5.1549720392087865e-06,
+      "loss": 0.1937,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5056078959174518,
+      "grad_norm": 1.453744768955761,
+      "learning_rate": 5.152551358697468e-06,
+      "loss": 0.2786,
+      "step": 3381
+    },
+    {
+      "epoch": 0.5057574398085838,
+      "grad_norm": 1.4933599639180963,
+      "learning_rate": 5.150130642396129e-06,
+      "loss": 0.2105,
+      "step": 3382
+    },
+    {
+      "epoch": 0.5059069836997159,
+      "grad_norm": 1.2741666356322288,
+      "learning_rate": 5.147709890872697e-06,
+      "loss": 0.1731,
+      "step": 3383
+    },
+    {
+      "epoch": 0.506056527590848,
+      "grad_norm": 2.2491770404898217,
+      "learning_rate": 5.145289104695101e-06,
+      "loss": 0.5676,
+      "step": 3384
+    },
+    {
+      "epoch": 0.5062060714819799,
+      "grad_norm": 1.9408723876207343,
+      "learning_rate": 5.14286828443128e-06,
+      "loss": 0.4774,
+      "step": 3385
+    },
+    {
+      "epoch": 0.506355615373112,
+      "grad_norm": 1.4212514196361943,
+      "learning_rate": 5.1404474306491816e-06,
+      "loss": 0.214,
+      "step": 3386
+    },
+    {
+      "epoch": 0.5065051592642441,
+      "grad_norm": 1.9045071920993695,
+      "learning_rate": 5.138026543916763e-06,
+      "loss": 0.3409,
+      "step": 3387
+    },
+    {
+      "epoch": 0.5066547031553761,
+      "grad_norm": 2.155362762135023,
+      "learning_rate": 5.135605624801987e-06,
+      "loss": 0.5308,
+      "step": 3388
+    },
+    {
+      "epoch": 0.5068042470465082,
+      "grad_norm": 1.7219918801573693,
+      "learning_rate": 5.133184673872824e-06,
+      "loss": 0.2495,
+      "step": 3389
+    },
+    {
+      "epoch": 0.5069537909376401,
+      "grad_norm": 1.1700199480933995,
+      "learning_rate": 5.130763691697254e-06,
+      "loss": 0.1942,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5071033348287722,
+      "grad_norm": 1.7548276948843984,
+      "learning_rate": 5.128342678843261e-06,
+      "loss": 0.4727,
+      "step": 3391
+    },
+    {
+      "epoch": 0.5072528787199043,
+      "grad_norm": 1.4512806488385455,
+      "learning_rate": 5.125921635878839e-06,
+      "loss": 0.3694,
+      "step": 3392
+    },
+    {
+      "epoch": 0.5074024226110363,
+      "grad_norm": 1.4994540711193551,
+      "learning_rate": 5.123500563371988e-06,
+      "loss": 0.2164,
+      "step": 3393
+    },
+    {
+      "epoch": 0.5075519665021684,
+      "grad_norm": 1.5807710125666976,
+      "learning_rate": 5.1210794618907135e-06,
+      "loss": 0.3291,
+      "step": 3394
+    },
+    {
+      "epoch": 0.5077015103933005,
+      "grad_norm": 1.5954515312505237,
+      "learning_rate": 5.118658332003031e-06,
+      "loss": 0.4411,
+      "step": 3395
+    },
+    {
+      "epoch": 0.5078510542844324,
+      "grad_norm": 1.9192671256027398,
+      "learning_rate": 5.116237174276961e-06,
+      "loss": 0.5256,
+      "step": 3396
+    },
+    {
+      "epoch": 0.5080005981755645,
+      "grad_norm": 1.4114290051973426,
+      "learning_rate": 5.113815989280528e-06,
+      "loss": 0.218,
+      "step": 3397
+    },
+    {
+      "epoch": 0.5081501420666966,
+      "grad_norm": 1.8229179727458815,
+      "learning_rate": 5.111394777581769e-06,
+      "loss": 0.433,
+      "step": 3398
+    },
+    {
+      "epoch": 0.5082996859578286,
+      "grad_norm": 1.0778267074582055,
+      "learning_rate": 5.108973539748721e-06,
+      "loss": 0.1962,
+      "step": 3399
+    },
+    {
+      "epoch": 0.5084492298489607,
+      "grad_norm": 1.5454148953863824,
+      "learning_rate": 5.10655227634943e-06,
+      "loss": 0.3592,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5085987737400928,
+      "grad_norm": 1.74859452433122,
+      "learning_rate": 5.104130987951947e-06,
+      "loss": 0.496,
+      "step": 3401
+    },
+    {
+      "epoch": 0.5087483176312247,
+      "grad_norm": 1.362679322662736,
+      "learning_rate": 5.101709675124332e-06,
+      "loss": 0.2273,
+      "step": 3402
+    },
+    {
+      "epoch": 0.5088978615223568,
+      "grad_norm": 2.070075014276669,
+      "learning_rate": 5.0992883384346485e-06,
+      "loss": 0.2638,
+      "step": 3403
+    },
+    {
+      "epoch": 0.5090474054134888,
+      "grad_norm": 1.5938286200934397,
+      "learning_rate": 5.096866978450962e-06,
+      "loss": 0.5063,
+      "step": 3404
+    },
+    {
+      "epoch": 0.5091969493046209,
+      "grad_norm": 1.16516888607397,
+      "learning_rate": 5.09444559574135e-06,
+      "loss": 0.2357,
+      "step": 3405
+    },
+    {
+      "epoch": 0.509346493195753,
+      "grad_norm": 2.029114392825952,
+      "learning_rate": 5.0920241908738885e-06,
+      "loss": 0.5142,
+      "step": 3406
+    },
+    {
+      "epoch": 0.509496037086885,
+      "grad_norm": 1.5198340641323709,
+      "learning_rate": 5.089602764416667e-06,
+      "loss": 0.236,
+      "step": 3407
+    },
+    {
+      "epoch": 0.509645580978017,
+      "grad_norm": 1.7318426021724316,
+      "learning_rate": 5.087181316937773e-06,
+      "loss": 0.342,
+      "step": 3408
+    },
+    {
+      "epoch": 0.5097951248691491,
+      "grad_norm": 1.412619025249756,
+      "learning_rate": 5.084759849005301e-06,
+      "loss": 0.1432,
+      "step": 3409
+    },
+    {
+      "epoch": 0.5099446687602811,
+      "grad_norm": 1.5740314968616134,
+      "learning_rate": 5.082338361187354e-06,
+      "loss": 0.3604,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5100942126514132,
+      "grad_norm": 1.6106092571840667,
+      "learning_rate": 5.079916854052031e-06,
+      "loss": 0.3553,
+      "step": 3411
+    },
+    {
+      "epoch": 0.5102437565425453,
+      "grad_norm": 1.3048853067534962,
+      "learning_rate": 5.077495328167446e-06,
+      "loss": 0.2093,
+      "step": 3412
+    },
+    {
+      "epoch": 0.5103933004336773,
+      "grad_norm": 1.3774154157596905,
+      "learning_rate": 5.07507378410171e-06,
+      "loss": 0.3105,
+      "step": 3413
+    },
+    {
+      "epoch": 0.5105428443248093,
+      "grad_norm": 1.0928212880625816,
+      "learning_rate": 5.072652222422938e-06,
+      "loss": 0.1978,
+      "step": 3414
+    },
+    {
+      "epoch": 0.5106923882159414,
+      "grad_norm": 1.148279407869326,
+      "learning_rate": 5.0702306436992575e-06,
+      "loss": 0.1998,
+      "step": 3415
+    },
+    {
+      "epoch": 0.5108419321070734,
+      "grad_norm": 2.167927897987719,
+      "learning_rate": 5.0678090484987895e-06,
+      "loss": 0.3238,
+      "step": 3416
+    },
+    {
+      "epoch": 0.5109914759982055,
+      "grad_norm": 1.5482664593136313,
+      "learning_rate": 5.065387437389666e-06,
+      "loss": 0.183,
+      "step": 3417
+    },
+    {
+      "epoch": 0.5111410198893375,
+      "grad_norm": 1.0394819657038659,
+      "learning_rate": 5.06296581094002e-06,
+      "loss": 0.1603,
+      "step": 3418
+    },
+    {
+      "epoch": 0.5112905637804696,
+      "grad_norm": 1.106087546826269,
+      "learning_rate": 5.060544169717987e-06,
+      "loss": 0.1787,
+      "step": 3419
+    },
+    {
+      "epoch": 0.5114401076716016,
+      "grad_norm": 1.0109749456896577,
+      "learning_rate": 5.058122514291709e-06,
+      "loss": 0.1278,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5115896515627336,
+      "grad_norm": 1.3052672308820343,
+      "learning_rate": 5.0557008452293275e-06,
+      "loss": 0.3621,
+      "step": 3421
+    },
+    {
+      "epoch": 0.5117391954538657,
+      "grad_norm": 1.0511096136393951,
+      "learning_rate": 5.053279163098992e-06,
+      "loss": 0.1874,
+      "step": 3422
+    },
+    {
+      "epoch": 0.5118887393449978,
+      "grad_norm": 1.5171824078027538,
+      "learning_rate": 5.050857468468853e-06,
+      "loss": 0.3404,
+      "step": 3423
+    },
+    {
+      "epoch": 0.5120382832361298,
+      "grad_norm": 1.050201412025008,
+      "learning_rate": 5.04843576190706e-06,
+      "loss": 0.1693,
+      "step": 3424
+    },
+    {
+      "epoch": 0.5121878271272619,
+      "grad_norm": 1.4748037147038833,
+      "learning_rate": 5.0460140439817695e-06,
+      "loss": 0.2263,
+      "step": 3425
+    },
+    {
+      "epoch": 0.5123373710183939,
+      "grad_norm": 1.3128576316535443,
+      "learning_rate": 5.043592315261143e-06,
+      "loss": 0.1389,
+      "step": 3426
+    },
+    {
+      "epoch": 0.5124869149095259,
+      "grad_norm": 1.284722286802869,
+      "learning_rate": 5.041170576313338e-06,
+      "loss": 0.209,
+      "step": 3427
+    },
+    {
+      "epoch": 0.512636458800658,
+      "grad_norm": 1.1356824496722022,
+      "learning_rate": 5.0387488277065186e-06,
+      "loss": 0.3192,
+      "step": 3428
+    },
+    {
+      "epoch": 0.5127860026917901,
+      "grad_norm": 1.3644018650118583,
+      "learning_rate": 5.036327070008852e-06,
+      "loss": 0.229,
+      "step": 3429
+    },
+    {
+      "epoch": 0.5129355465829221,
+      "grad_norm": 1.2797031965795196,
+      "learning_rate": 5.0339053037885035e-06,
+      "loss": 0.2217,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5130850904740542,
+      "grad_norm": 1.6886018581906976,
+      "learning_rate": 5.031483529613646e-06,
+      "loss": 0.4042,
+      "step": 3431
+    },
+    {
+      "epoch": 0.5132346343651862,
+      "grad_norm": 1.454212457336455,
+      "learning_rate": 5.029061748052446e-06,
+      "loss": 0.1886,
+      "step": 3432
+    },
+    {
+      "epoch": 0.5133841782563182,
+      "grad_norm": 1.4334168078083147,
+      "learning_rate": 5.026639959673083e-06,
+      "loss": 0.3052,
+      "step": 3433
+    },
+    {
+      "epoch": 0.5135337221474503,
+      "grad_norm": 1.745914040348649,
+      "learning_rate": 5.024218165043726e-06,
+      "loss": 0.3588,
+      "step": 3434
+    },
+    {
+      "epoch": 0.5136832660385823,
+      "grad_norm": 2.0781048544287293,
+      "learning_rate": 5.021796364732554e-06,
+      "loss": 0.2493,
+      "step": 3435
+    },
+    {
+      "epoch": 0.5138328099297144,
+      "grad_norm": 1.6255642021510852,
+      "learning_rate": 5.019374559307747e-06,
+      "loss": 0.346,
+      "step": 3436
+    },
+    {
+      "epoch": 0.5139823538208464,
+      "grad_norm": 1.7336607656638219,
+      "learning_rate": 5.01695274933748e-06,
+      "loss": 0.203,
+      "step": 3437
+    },
+    {
+      "epoch": 0.5141318977119784,
+      "grad_norm": 1.6058882608416245,
+      "learning_rate": 5.014530935389938e-06,
+      "loss": 0.3645,
+      "step": 3438
+    },
+    {
+      "epoch": 0.5142814416031105,
+      "grad_norm": 1.4498498807058657,
+      "learning_rate": 5.012109118033294e-06,
+      "loss": 0.2795,
+      "step": 3439
+    },
+    {
+      "epoch": 0.5144309854942426,
+      "grad_norm": 2.0694828386357873,
+      "learning_rate": 5.009687297835736e-06,
+      "loss": 0.5495,
+      "step": 3440
+    },
+    {
+      "epoch": 0.5145805293853746,
+      "grad_norm": 1.5377492094048737,
+      "learning_rate": 5.0072654753654444e-06,
+      "loss": 0.4609,
+      "step": 3441
+    },
+    {
+      "epoch": 0.5147300732765067,
+      "grad_norm": 1.4923232476167383,
+      "learning_rate": 5.004843651190602e-06,
+      "loss": 0.314,
+      "step": 3442
+    },
+    {
+      "epoch": 0.5148796171676387,
+      "grad_norm": 1.023025356909767,
+      "learning_rate": 5.002421825879394e-06,
+      "loss": 0.2076,
+      "step": 3443
+    },
+    {
+      "epoch": 0.5150291610587707,
+      "grad_norm": 1.4703535067143427,
+      "learning_rate": 5e-06,
+      "loss": 0.2911,
+      "step": 3444
+    },
+    {
+      "epoch": 0.5151787049499028,
+      "grad_norm": 1.635494712784053,
+      "learning_rate": 4.997578174120607e-06,
+      "loss": 0.1704,
+      "step": 3445
+    },
+    {
+      "epoch": 0.5153282488410349,
+      "grad_norm": 1.3501580885318492,
+      "learning_rate": 4.995156348809398e-06,
+      "loss": 0.3092,
+      "step": 3446
+    },
+    {
+      "epoch": 0.5154777927321669,
+      "grad_norm": 1.4552065824604172,
+      "learning_rate": 4.992734524634557e-06,
+      "loss": 0.3157,
+      "step": 3447
+    },
+    {
+      "epoch": 0.515627336623299,
+      "grad_norm": 1.6023986192308008,
+      "learning_rate": 4.990312702164265e-06,
+      "loss": 0.3258,
+      "step": 3448
+    },
+    {
+      "epoch": 0.5157768805144309,
+      "grad_norm": 1.135753254332216,
+      "learning_rate": 4.987890881966707e-06,
+      "loss": 0.2194,
+      "step": 3449
+    },
+    {
+      "epoch": 0.515926424405563,
+      "grad_norm": 1.7579710147169378,
+      "learning_rate": 4.985469064610065e-06,
+      "loss": 0.379,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5160759682966951,
+      "grad_norm": 1.2562049164914357,
+      "learning_rate": 4.983047250662519e-06,
+      "loss": 0.19,
+      "step": 3451
+    },
+    {
+      "epoch": 0.5162255121878271,
+      "grad_norm": 1.7306909181736667,
+      "learning_rate": 4.980625440692255e-06,
+      "loss": 0.5524,
+      "step": 3452
+    },
+    {
+      "epoch": 0.5163750560789592,
+      "grad_norm": 1.0200496356880533,
+      "learning_rate": 4.978203635267447e-06,
+      "loss": 0.2777,
+      "step": 3453
+    },
+    {
+      "epoch": 0.5165245999700913,
+      "grad_norm": 1.02643941739448,
+      "learning_rate": 4.9757818349562755e-06,
+      "loss": 0.2026,
+      "step": 3454
+    },
+    {
+      "epoch": 0.5166741438612232,
+      "grad_norm": 1.5225694971242867,
+      "learning_rate": 4.973360040326919e-06,
+      "loss": 0.3837,
+      "step": 3455
+    },
+    {
+      "epoch": 0.5168236877523553,
+      "grad_norm": 1.4438502518414775,
+      "learning_rate": 4.970938251947554e-06,
+      "loss": 0.4278,
+      "step": 3456
+    },
+    {
+      "epoch": 0.5169732316434874,
+      "grad_norm": 1.3218484273904558,
+      "learning_rate": 4.968516470386357e-06,
+      "loss": 0.2523,
+      "step": 3457
+    },
+    {
+      "epoch": 0.5171227755346194,
+      "grad_norm": 1.3619841304443039,
+      "learning_rate": 4.966094696211498e-06,
+      "loss": 0.2346,
+      "step": 3458
+    },
+    {
+      "epoch": 0.5172723194257515,
+      "grad_norm": 0.9202093983071812,
+      "learning_rate": 4.96367292999115e-06,
+      "loss": 0.1737,
+      "step": 3459
+    },
+    {
+      "epoch": 0.5174218633168836,
+      "grad_norm": 1.5024659381360044,
+      "learning_rate": 4.961251172293482e-06,
+      "loss": 0.2774,
+      "step": 3460
+    },
+    {
+      "epoch": 0.5175714072080155,
+      "grad_norm": 1.142291287339441,
+      "learning_rate": 4.958829423686663e-06,
+      "loss": 0.2169,
+      "step": 3461
+    },
+    {
+      "epoch": 0.5177209510991476,
+      "grad_norm": 1.1269059328349746,
+      "learning_rate": 4.956407684738859e-06,
+      "loss": 0.2414,
+      "step": 3462
+    },
+    {
+      "epoch": 0.5178704949902796,
+      "grad_norm": 1.4408048574377068,
+      "learning_rate": 4.953985956018231e-06,
+      "loss": 0.3941,
+      "step": 3463
+    },
+    {
+      "epoch": 0.5180200388814117,
+      "grad_norm": 1.1493374443284174,
+      "learning_rate": 4.951564238092942e-06,
+      "loss": 0.1601,
+      "step": 3464
+    },
+    {
+      "epoch": 0.5181695827725438,
+      "grad_norm": 1.543338492311835,
+      "learning_rate": 4.949142531531149e-06,
+      "loss": 0.26,
+      "step": 3465
+    },
+    {
+      "epoch": 0.5183191266636757,
+      "grad_norm": 1.0704201449260624,
+      "learning_rate": 4.946720836901008e-06,
+      "loss": 0.2098,
+      "step": 3466
+    },
+    {
+      "epoch": 0.5184686705548078,
+      "grad_norm": 1.1149725431271198,
+      "learning_rate": 4.944299154770673e-06,
+      "loss": 0.1564,
+      "step": 3467
+    },
+    {
+      "epoch": 0.5186182144459399,
+      "grad_norm": 1.58868666961587,
+      "learning_rate": 4.941877485708293e-06,
+      "loss": 0.4078,
+      "step": 3468
+    },
+    {
+      "epoch": 0.5187677583370719,
+      "grad_norm": 1.263877473847384,
+      "learning_rate": 4.939455830282014e-06,
+      "loss": 0.2109,
+      "step": 3469
+    },
+    {
+      "epoch": 0.518917302228204,
+      "grad_norm": 2.0767432150699503,
+      "learning_rate": 4.937034189059981e-06,
+      "loss": 0.4936,
+      "step": 3470
+    },
+    {
+      "epoch": 0.5190668461193361,
+      "grad_norm": 1.564988355482826,
+      "learning_rate": 4.9346125626103344e-06,
+      "loss": 0.2268,
+      "step": 3471
+    },
+    {
+      "epoch": 0.519216390010468,
+      "grad_norm": 1.7520264234912677,
+      "learning_rate": 4.932190951501212e-06,
+      "loss": 0.3324,
+      "step": 3472
+    },
+    {
+      "epoch": 0.5193659339016001,
+      "grad_norm": 0.9378784012067424,
+      "learning_rate": 4.929769356300745e-06,
+      "loss": 0.1623,
+      "step": 3473
+    },
+    {
+      "epoch": 0.5195154777927322,
+      "grad_norm": 1.3070459934481593,
+      "learning_rate": 4.9273477775770634e-06,
+      "loss": 0.3314,
+      "step": 3474
+    },
+    {
+      "epoch": 0.5196650216838642,
+      "grad_norm": 1.7416905476980176,
+      "learning_rate": 4.924926215898292e-06,
+      "loss": 0.4998,
+      "step": 3475
+    },
+    {
+      "epoch": 0.5198145655749963,
+      "grad_norm": 1.356929202517075,
+      "learning_rate": 4.922504671832555e-06,
+      "loss": 0.2937,
+      "step": 3476
+    },
+    {
+      "epoch": 0.5199641094661283,
+      "grad_norm": 1.4583002960913412,
+      "learning_rate": 4.920083145947971e-06,
+      "loss": 0.323,
+      "step": 3477
+    },
+    {
+      "epoch": 0.5201136533572603,
+      "grad_norm": 2.0884343537485734,
+      "learning_rate": 4.917661638812649e-06,
+      "loss": 0.5929,
+      "step": 3478
+    },
+    {
+      "epoch": 0.5202631972483924,
+      "grad_norm": 1.9246883502332144,
+      "learning_rate": 4.9152401509947e-06,
+      "loss": 0.5366,
+      "step": 3479
+    },
+    {
+      "epoch": 0.5204127411395244,
+      "grad_norm": 1.7689104275350418,
+      "learning_rate": 4.912818683062229e-06,
+      "loss": 0.2301,
+      "step": 3480
+    },
+    {
+      "epoch": 0.5205622850306565,
+      "grad_norm": 1.6135079839907187,
+      "learning_rate": 4.9103972355833345e-06,
+      "loss": 0.2101,
+      "step": 3481
+    },
+    {
+      "epoch": 0.5207118289217886,
+      "grad_norm": 1.5135656331832419,
+      "learning_rate": 4.907975809126113e-06,
+      "loss": 0.3627,
+      "step": 3482
+    },
+    {
+      "epoch": 0.5208613728129206,
+      "grad_norm": 1.1866081615701136,
+      "learning_rate": 4.905554404258654e-06,
+      "loss": 0.3292,
+      "step": 3483
+    },
+    {
+      "epoch": 0.5210109167040526,
+      "grad_norm": 1.6290902598708927,
+      "learning_rate": 4.90313302154904e-06,
+      "loss": 0.3552,
+      "step": 3484
+    },
+    {
+      "epoch": 0.5211604605951847,
+      "grad_norm": 1.3734572788893116,
+      "learning_rate": 4.900711661565353e-06,
+      "loss": 0.215,
+      "step": 3485
+    },
+    {
+      "epoch": 0.5213100044863167,
+      "grad_norm": 0.9577406608934755,
+      "learning_rate": 4.898290324875668e-06,
+      "loss": 0.1676,
+      "step": 3486
+    },
+    {
+      "epoch": 0.5214595483774488,
+      "grad_norm": 0.9118461606891877,
+      "learning_rate": 4.895869012048053e-06,
+      "loss": 0.1516,
+      "step": 3487
+    },
+    {
+      "epoch": 0.5216090922685809,
+      "grad_norm": 1.34978265933406,
+      "learning_rate": 4.893447723650572e-06,
+      "loss": 0.1731,
+      "step": 3488
+    },
+    {
+      "epoch": 0.5217586361597129,
+      "grad_norm": 1.3337609032683095,
+      "learning_rate": 4.89102646025128e-06,
+      "loss": 0.3653,
+      "step": 3489
+    },
+    {
+      "epoch": 0.5219081800508449,
+      "grad_norm": 1.1283458585985406,
+      "learning_rate": 4.888605222418232e-06,
+      "loss": 0.2716,
+      "step": 3490
+    },
+    {
+      "epoch": 0.522057723941977,
+      "grad_norm": 1.6956120283556566,
+      "learning_rate": 4.886184010719472e-06,
+      "loss": 0.4099,
+      "step": 3491
+    },
+    {
+      "epoch": 0.522207267833109,
+      "grad_norm": 1.63895974569,
+      "learning_rate": 4.88376282572304e-06,
+      "loss": 0.3172,
+      "step": 3492
+    },
+    {
+      "epoch": 0.5223568117242411,
+      "grad_norm": 1.5138032302467108,
+      "learning_rate": 4.88134166799697e-06,
+      "loss": 0.2708,
+      "step": 3493
+    },
+    {
+      "epoch": 0.5225063556153731,
+      "grad_norm": 1.7341059849859075,
+      "learning_rate": 4.878920538109288e-06,
+      "loss": 0.3361,
+      "step": 3494
+    },
+    {
+      "epoch": 0.5226558995065052,
+      "grad_norm": 1.4357879747612492,
+      "learning_rate": 4.876499436628013e-06,
+      "loss": 0.2236,
+      "step": 3495
+    },
+    {
+      "epoch": 0.5228054433976372,
+      "grad_norm": 1.9388442682852385,
+      "learning_rate": 4.874078364121162e-06,
+      "loss": 0.3892,
+      "step": 3496
+    },
+    {
+      "epoch": 0.5229549872887692,
+      "grad_norm": 1.3139109038725212,
+      "learning_rate": 4.871657321156738e-06,
+      "loss": 0.2246,
+      "step": 3497
+    },
+    {
+      "epoch": 0.5231045311799013,
+      "grad_norm": 1.0818048681042511,
+      "learning_rate": 4.869236308302748e-06,
+      "loss": 0.2252,
+      "step": 3498
+    },
+    {
+      "epoch": 0.5232540750710334,
+      "grad_norm": 1.3703939412577448,
+      "learning_rate": 4.866815326127177e-06,
+      "loss": 0.1899,
+      "step": 3499
+    },
+    {
+      "epoch": 0.5234036189621654,
+      "grad_norm": 1.4822403566706586,
+      "learning_rate": 4.864394375198015e-06,
+      "loss": 0.3076,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5235531628532974,
+      "grad_norm": 1.9706263153865453,
+      "learning_rate": 4.861973456083238e-06,
+      "loss": 0.4768,
+      "step": 3501
+    },
+    {
+      "epoch": 0.5237027067444295,
+      "grad_norm": 2.1181556891438085,
+      "learning_rate": 4.85955256935082e-06,
+      "loss": 0.2084,
+      "step": 3502
+    },
+    {
+      "epoch": 0.5238522506355615,
+      "grad_norm": 1.192035696381639,
+      "learning_rate": 4.857131715568723e-06,
+      "loss": 0.224,
+      "step": 3503
+    },
+    {
+      "epoch": 0.5240017945266936,
+      "grad_norm": 0.7719952352600978,
+      "learning_rate": 4.854710895304902e-06,
+      "loss": 0.1399,
+      "step": 3504
+    },
+    {
+      "epoch": 0.5241513384178257,
+      "grad_norm": 1.3107969810050264,
+      "learning_rate": 4.852290109127304e-06,
+      "loss": 0.2784,
+      "step": 3505
+    },
+    {
+      "epoch": 0.5243008823089577,
+      "grad_norm": 1.4710047540471154,
+      "learning_rate": 4.849869357603871e-06,
+      "loss": 0.4199,
+      "step": 3506
+    },
+    {
+      "epoch": 0.5244504262000897,
+      "grad_norm": 1.9363546523939164,
+      "learning_rate": 4.847448641302532e-06,
+      "loss": 0.3845,
+      "step": 3507
+    },
+    {
+      "epoch": 0.5245999700912217,
+      "grad_norm": 1.4299597363726546,
+      "learning_rate": 4.845027960791215e-06,
+      "loss": 0.1818,
+      "step": 3508
+    },
+    {
+      "epoch": 0.5247495139823538,
+      "grad_norm": 1.392803585853321,
+      "learning_rate": 4.842607316637829e-06,
+      "loss": 0.1893,
+      "step": 3509
+    },
+    {
+      "epoch": 0.5248990578734859,
+      "grad_norm": 1.4652214173791756,
+      "learning_rate": 4.840186709410285e-06,
+      "loss": 0.3097,
+      "step": 3510
+    },
+    {
+      "epoch": 0.5250486017646179,
+      "grad_norm": 0.9930788283031162,
+      "learning_rate": 4.837766139676478e-06,
+      "loss": 0.1814,
+      "step": 3511
+    },
+    {
+      "epoch": 0.52519814565575,
+      "grad_norm": 1.4409939401080238,
+      "learning_rate": 4.835345608004299e-06,
+      "loss": 0.2391,
+      "step": 3512
+    },
+    {
+      "epoch": 0.525347689546882,
+      "grad_norm": 1.136343022154102,
+      "learning_rate": 4.832925114961629e-06,
+      "loss": 0.3325,
+      "step": 3513
+    },
+    {
+      "epoch": 0.525497233438014,
+      "grad_norm": 2.351393910634896,
+      "learning_rate": 4.8305046611163345e-06,
+      "loss": 0.2251,
+      "step": 3514
+    },
+    {
+      "epoch": 0.5256467773291461,
+      "grad_norm": 1.5324676085801396,
+      "learning_rate": 4.828084247036281e-06,
+      "loss": 0.3119,
+      "step": 3515
+    },
+    {
+      "epoch": 0.5257963212202782,
+      "grad_norm": 1.8398614526208819,
+      "learning_rate": 4.825663873289318e-06,
+      "loss": 0.221,
+      "step": 3516
+    },
+    {
+      "epoch": 0.5259458651114102,
+      "grad_norm": 1.7561849264734717,
+      "learning_rate": 4.823243540443292e-06,
+      "loss": 0.4972,
+      "step": 3517
+    },
+    {
+      "epoch": 0.5260954090025423,
+      "grad_norm": 1.2526140804123165,
+      "learning_rate": 4.820823249066038e-06,
+      "loss": 0.1815,
+      "step": 3518
+    },
+    {
+      "epoch": 0.5262449528936743,
+      "grad_norm": 1.6554249201864173,
+      "learning_rate": 4.818402999725373e-06,
+      "loss": 0.3794,
+      "step": 3519
+    },
+    {
+      "epoch": 0.5263944967848063,
+      "grad_norm": 1.3316819776233528,
+      "learning_rate": 4.815982792989114e-06,
+      "loss": 0.2839,
+      "step": 3520
+    },
+    {
+      "epoch": 0.5265440406759384,
+      "grad_norm": 1.692015505122212,
+      "learning_rate": 4.8135626294250655e-06,
+      "loss": 0.3722,
+      "step": 3521
+    },
+    {
+      "epoch": 0.5266935845670704,
+      "grad_norm": 1.7190685063797913,
+      "learning_rate": 4.811142509601022e-06,
+      "loss": 0.5343,
+      "step": 3522
+    },
+    {
+      "epoch": 0.5268431284582025,
+      "grad_norm": 1.4706527377105485,
+      "learning_rate": 4.808722434084766e-06,
+      "loss": 0.3285,
+      "step": 3523
+    },
+    {
+      "epoch": 0.5269926723493346,
+      "grad_norm": 1.8774181151216505,
+      "learning_rate": 4.80630240344407e-06,
+      "loss": 0.238,
+      "step": 3524
+    },
+    {
+      "epoch": 0.5271422162404665,
+      "grad_norm": 1.7586866324598942,
+      "learning_rate": 4.803882418246696e-06,
+      "loss": 0.4316,
+      "step": 3525
+    },
+    {
+      "epoch": 0.5272917601315986,
+      "grad_norm": 1.7649578082955628,
+      "learning_rate": 4.801462479060397e-06,
+      "loss": 0.4658,
+      "step": 3526
+    },
+    {
+      "epoch": 0.5274413040227307,
+      "grad_norm": 1.4270224302991061,
+      "learning_rate": 4.799042586452915e-06,
+      "loss": 0.1993,
+      "step": 3527
+    },
+    {
+      "epoch": 0.5275908479138627,
+      "grad_norm": 1.597762726870796,
+      "learning_rate": 4.796622740991978e-06,
+      "loss": 0.3639,
+      "step": 3528
+    },
+    {
+      "epoch": 0.5277403918049948,
+      "grad_norm": 1.5745989361707744,
+      "learning_rate": 4.794202943245306e-06,
+      "loss": 0.2338,
+      "step": 3529
+    },
+    {
+      "epoch": 0.5278899356961269,
+      "grad_norm": 1.6119777317893236,
+      "learning_rate": 4.791783193780607e-06,
+      "loss": 0.3549,
+      "step": 3530
+    },
+    {
+      "epoch": 0.5280394795872588,
+      "grad_norm": 1.448369751737763,
+      "learning_rate": 4.789363493165577e-06,
+      "loss": 0.3254,
+      "step": 3531
+    },
+    {
+      "epoch": 0.5281890234783909,
+      "grad_norm": 1.9783620520561083,
+      "learning_rate": 4.786943841967901e-06,
+      "loss": 0.4244,
+      "step": 3532
+    },
+    {
+      "epoch": 0.528338567369523,
+      "grad_norm": 1.7086111696643427,
+      "learning_rate": 4.7845242407552564e-06,
+      "loss": 0.5218,
+      "step": 3533
+    },
+    {
+      "epoch": 0.528488111260655,
+      "grad_norm": 2.058530122472167,
+      "learning_rate": 4.782104690095298e-06,
+      "loss": 0.705,
+      "step": 3534
+    },
+    {
+      "epoch": 0.5286376551517871,
+      "grad_norm": 1.4083919927576292,
+      "learning_rate": 4.779685190555681e-06,
+      "loss": 0.2537,
+      "step": 3535
+    },
+    {
+      "epoch": 0.528787199042919,
+      "grad_norm": 1.4221251550637304,
+      "learning_rate": 4.777265742704039e-06,
+      "loss": 0.214,
+      "step": 3536
+    },
+    {
+      "epoch": 0.5289367429340511,
+      "grad_norm": 1.5537572598460359,
+      "learning_rate": 4.774846347108001e-06,
+      "loss": 0.4211,
+      "step": 3537
+    },
+    {
+      "epoch": 0.5290862868251832,
+      "grad_norm": 1.4170336569377495,
+      "learning_rate": 4.772427004335183e-06,
+      "loss": 0.3499,
+      "step": 3538
+    },
+    {
+      "epoch": 0.5292358307163152,
+      "grad_norm": 1.4275209656159527,
+      "learning_rate": 4.770007714953181e-06,
+      "loss": 0.341,
+      "step": 3539
+    },
+    {
+      "epoch": 0.5293853746074473,
+      "grad_norm": 1.6285920446138622,
+      "learning_rate": 4.7675884795295855e-06,
+      "loss": 0.1867,
+      "step": 3540
+    },
+    {
+      "epoch": 0.5295349184985794,
+      "grad_norm": 1.430494043188066,
+      "learning_rate": 4.765169298631973e-06,
+      "loss": 0.1935,
+      "step": 3541
+    },
+    {
+      "epoch": 0.5296844623897113,
+      "grad_norm": 1.2416404413194209,
+      "learning_rate": 4.762750172827905e-06,
+      "loss": 0.1899,
+      "step": 3542
+    },
+    {
+      "epoch": 0.5298340062808434,
+      "grad_norm": 1.4246138627274758,
+      "learning_rate": 4.760331102684934e-06,
+      "loss": 0.2316,
+      "step": 3543
+    },
+    {
+      "epoch": 0.5299835501719755,
+      "grad_norm": 1.6367006596769544,
+      "learning_rate": 4.757912088770596e-06,
+      "loss": 0.4669,
+      "step": 3544
+    },
+    {
+      "epoch": 0.5301330940631075,
+      "grad_norm": 1.366581512757693,
+      "learning_rate": 4.755493131652415e-06,
+      "loss": 0.1932,
+      "step": 3545
+    },
+    {
+      "epoch": 0.5302826379542396,
+      "grad_norm": 1.7191803566801642,
+      "learning_rate": 4.7530742318979e-06,
+      "loss": 0.484,
+      "step": 3546
+    },
+    {
+      "epoch": 0.5304321818453717,
+      "grad_norm": 1.552142422679913,
+      "learning_rate": 4.750655390074552e-06,
+      "loss": 0.265,
+      "step": 3547
+    },
+    {
+      "epoch": 0.5305817257365036,
+      "grad_norm": 1.1885077097062553,
+      "learning_rate": 4.7482366067498526e-06,
+      "loss": 0.1907,
+      "step": 3548
+    },
+    {
+      "epoch": 0.5307312696276357,
+      "grad_norm": 1.6116298485523732,
+      "learning_rate": 4.7458178824912704e-06,
+      "loss": 0.2614,
+      "step": 3549
+    },
+    {
+      "epoch": 0.5308808135187678,
+      "grad_norm": 1.6356455691825849,
+      "learning_rate": 4.743399217866263e-06,
+      "loss": 0.3475,
+      "step": 3550
+    },
+    {
+      "epoch": 0.5310303574098998,
+      "grad_norm": 1.4715217386600319,
+      "learning_rate": 4.740980613442272e-06,
+      "loss": 0.1646,
+      "step": 3551
+    },
+    {
+      "epoch": 0.5311799013010319,
+      "grad_norm": 1.416799241645883,
+      "learning_rate": 4.738562069786725e-06,
+      "loss": 0.3194,
+      "step": 3552
+    },
+    {
+      "epoch": 0.5313294451921639,
+      "grad_norm": 1.5866917428028944,
+      "learning_rate": 4.736143587467039e-06,
+      "loss": 0.1772,
+      "step": 3553
+    },
+    {
+      "epoch": 0.5314789890832959,
+      "grad_norm": 1.229115311730451,
+      "learning_rate": 4.733725167050607e-06,
+      "loss": 0.1627,
+      "step": 3554
+    },
+    {
+      "epoch": 0.531628532974428,
+      "grad_norm": 1.211183396876207,
+      "learning_rate": 4.731306809104818e-06,
+      "loss": 0.2267,
+      "step": 3555
+    },
+    {
+      "epoch": 0.53177807686556,
+      "grad_norm": 2.2369509200391455,
+      "learning_rate": 4.7288885141970396e-06,
+      "loss": 0.2039,
+      "step": 3556
+    },
+    {
+      "epoch": 0.5319276207566921,
+      "grad_norm": 1.8028493082061634,
+      "learning_rate": 4.726470282894629e-06,
+      "loss": 0.3871,
+      "step": 3557
+    },
+    {
+      "epoch": 0.5320771646478242,
+      "grad_norm": 1.969525411582806,
+      "learning_rate": 4.7240521157649295e-06,
+      "loss": 0.5476,
+      "step": 3558
+    },
+    {
+      "epoch": 0.5322267085389562,
+      "grad_norm": 1.6776763319782029,
+      "learning_rate": 4.7216340133752604e-06,
+      "loss": 0.4609,
+      "step": 3559
+    },
+    {
+      "epoch": 0.5323762524300882,
+      "grad_norm": 1.220727007823009,
+      "learning_rate": 4.719215976292934e-06,
+      "loss": 0.2119,
+      "step": 3560
+    },
+    {
+      "epoch": 0.5325257963212203,
+      "grad_norm": 1.7235272167160625,
+      "learning_rate": 4.716798005085246e-06,
+      "loss": 0.412,
+      "step": 3561
+    },
+    {
+      "epoch": 0.5326753402123523,
+      "grad_norm": 1.954837902451542,
+      "learning_rate": 4.714380100319476e-06,
+      "loss": 0.5354,
+      "step": 3562
+    },
+    {
+      "epoch": 0.5328248841034844,
+      "grad_norm": 1.5724258743980508,
+      "learning_rate": 4.711962262562887e-06,
+      "loss": 0.2358,
+      "step": 3563
+    },
+    {
+      "epoch": 0.5329744279946165,
+      "grad_norm": 2.0670603634468545,
+      "learning_rate": 4.709544492382726e-06,
+      "loss": 0.5506,
+      "step": 3564
+    },
+    {
+      "epoch": 0.5331239718857484,
+      "grad_norm": 1.657619094892483,
+      "learning_rate": 4.707126790346227e-06,
+      "loss": 0.3559,
+      "step": 3565
+    },
+    {
+      "epoch": 0.5332735157768805,
+      "grad_norm": 1.4867129285160177,
+      "learning_rate": 4.704709157020606e-06,
+      "loss": 0.3264,
+      "step": 3566
+    },
+    {
+      "epoch": 0.5334230596680125,
+      "grad_norm": 1.7017440972223712,
+      "learning_rate": 4.702291592973062e-06,
+      "loss": 0.2979,
+      "step": 3567
+    },
+    {
+      "epoch": 0.5335726035591446,
+      "grad_norm": 1.9673502426761393,
+      "learning_rate": 4.69987409877078e-06,
+      "loss": 0.5318,
+      "step": 3568
+    },
+    {
+      "epoch": 0.5337221474502767,
+      "grad_norm": 1.4977969143854684,
+      "learning_rate": 4.697456674980926e-06,
+      "loss": 0.3394,
+      "step": 3569
+    },
+    {
+      "epoch": 0.5338716913414087,
+      "grad_norm": 1.6784064708619297,
+      "learning_rate": 4.695039322170652e-06,
+      "loss": 0.3555,
+      "step": 3570
+    },
+    {
+      "epoch": 0.5340212352325407,
+      "grad_norm": 1.96234854429619,
+      "learning_rate": 4.692622040907092e-06,
+      "loss": 0.6075,
+      "step": 3571
+    },
+    {
+      "epoch": 0.5341707791236728,
+      "grad_norm": 1.4259119209086994,
+      "learning_rate": 4.690204831757363e-06,
+      "loss": 0.1889,
+      "step": 3572
+    },
+    {
+      "epoch": 0.5343203230148048,
+      "grad_norm": 1.7546548074341575,
+      "learning_rate": 4.687787695288567e-06,
+      "loss": 0.3012,
+      "step": 3573
+    },
+    {
+      "epoch": 0.5344698669059369,
+      "grad_norm": 1.8205519670137686,
+      "learning_rate": 4.685370632067788e-06,
+      "loss": 0.4875,
+      "step": 3574
+    },
+    {
+      "epoch": 0.534619410797069,
+      "grad_norm": 1.3195329317180426,
+      "learning_rate": 4.682953642662089e-06,
+      "loss": 0.2789,
+      "step": 3575
+    },
+    {
+      "epoch": 0.534768954688201,
+      "grad_norm": 1.1371430231198516,
+      "learning_rate": 4.680536727638522e-06,
+      "loss": 0.1545,
+      "step": 3576
+    },
+    {
+      "epoch": 0.534918498579333,
+      "grad_norm": 1.3375712737529661,
+      "learning_rate": 4.678119887564115e-06,
+      "loss": 0.1937,
+      "step": 3577
+    },
+    {
+      "epoch": 0.5350680424704651,
+      "grad_norm": 1.1773100320990963,
+      "learning_rate": 4.675703123005886e-06,
+      "loss": 0.2153,
+      "step": 3578
+    },
+    {
+      "epoch": 0.5352175863615971,
+      "grad_norm": 1.2694167100147624,
+      "learning_rate": 4.673286434530832e-06,
+      "loss": 0.3311,
+      "step": 3579
+    },
+    {
+      "epoch": 0.5353671302527292,
+      "grad_norm": 1.973094778134566,
+      "learning_rate": 4.670869822705926e-06,
+      "loss": 0.5673,
+      "step": 3580
+    },
+    {
+      "epoch": 0.5355166741438612,
+      "grad_norm": 1.4077081167408994,
+      "learning_rate": 4.668453288098132e-06,
+      "loss": 0.2774,
+      "step": 3581
+    },
+    {
+      "epoch": 0.5356662180349933,
+      "grad_norm": 1.516290100892926,
+      "learning_rate": 4.666036831274392e-06,
+      "loss": 0.2193,
+      "step": 3582
+    },
+    {
+      "epoch": 0.5358157619261253,
+      "grad_norm": 1.5345251033260943,
+      "learning_rate": 4.663620452801631e-06,
+      "loss": 0.3554,
+      "step": 3583
+    },
+    {
+      "epoch": 0.5359653058172573,
+      "grad_norm": 1.4266746622095299,
+      "learning_rate": 4.661204153246753e-06,
+      "loss": 0.3471,
+      "step": 3584
+    },
+    {
+      "epoch": 0.5361148497083894,
+      "grad_norm": 1.959088255237877,
+      "learning_rate": 4.6587879331766465e-06,
+      "loss": 0.4178,
+      "step": 3585
+    },
+    {
+      "epoch": 0.5362643935995215,
+      "grad_norm": 1.2738818990475473,
+      "learning_rate": 4.656371793158178e-06,
+      "loss": 0.2069,
+      "step": 3586
+    },
+    {
+      "epoch": 0.5364139374906535,
+      "grad_norm": 1.9050636351962995,
+      "learning_rate": 4.653955733758198e-06,
+      "loss": 0.3567,
+      "step": 3587
+    },
+    {
+      "epoch": 0.5365634813817856,
+      "grad_norm": 1.6779491300436666,
+      "learning_rate": 4.651539755543539e-06,
+      "loss": 0.1506,
+      "step": 3588
+    },
+    {
+      "epoch": 0.5367130252729176,
+      "grad_norm": 1.7278703163893183,
+      "learning_rate": 4.6491238590810114e-06,
+      "loss": 0.5122,
+      "step": 3589
+    },
+    {
+      "epoch": 0.5368625691640496,
+      "grad_norm": 1.285149964140387,
+      "learning_rate": 4.646708044937408e-06,
+      "loss": 0.2056,
+      "step": 3590
+    },
+    {
+      "epoch": 0.5370121130551817,
+      "grad_norm": 1.0933373626532004,
+      "learning_rate": 4.644292313679502e-06,
+      "loss": 0.2019,
+      "step": 3591
+    },
+    {
+      "epoch": 0.5371616569463138,
+      "grad_norm": 1.4578130787462922,
+      "learning_rate": 4.6418766658740464e-06,
+      "loss": 0.191,
+      "step": 3592
+    },
+    {
+      "epoch": 0.5373112008374458,
+      "grad_norm": 1.5712976259860016,
+      "learning_rate": 4.639461102087777e-06,
+      "loss": 0.3531,
+      "step": 3593
+    },
+    {
+      "epoch": 0.5374607447285779,
+      "grad_norm": 1.4827323156288543,
+      "learning_rate": 4.63704562288741e-06,
+      "loss": 0.1951,
+      "step": 3594
+    },
+    {
+      "epoch": 0.5376102886197098,
+      "grad_norm": 1.115006492033478,
+      "learning_rate": 4.634630228839635e-06,
+      "loss": 0.2213,
+      "step": 3595
+    },
+    {
+      "epoch": 0.5377598325108419,
+      "grad_norm": 1.461958976876183,
+      "learning_rate": 4.63221492051113e-06,
+      "loss": 0.4051,
+      "step": 3596
+    },
+    {
+      "epoch": 0.537909376401974,
+      "grad_norm": 1.6258766326897567,
+      "learning_rate": 4.629799698468549e-06,
+      "loss": 0.1955,
+      "step": 3597
+    },
+    {
+      "epoch": 0.538058920293106,
+      "grad_norm": 1.4054556934541154,
+      "learning_rate": 4.6273845632785256e-06,
+      "loss": 0.2125,
+      "step": 3598
+    },
+    {
+      "epoch": 0.5382084641842381,
+      "grad_norm": 1.041762193432224,
+      "learning_rate": 4.624969515507679e-06,
+      "loss": 0.1924,
+      "step": 3599
+    },
+    {
+      "epoch": 0.5383580080753702,
+      "grad_norm": 1.4542501769456762,
+      "learning_rate": 4.6225545557225945e-06,
+      "loss": 0.3441,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5385075519665021,
+      "grad_norm": 1.4919518389528756,
+      "learning_rate": 4.620139684489849e-06,
+      "loss": 0.2355,
+      "step": 3601
+    },
+    {
+      "epoch": 0.5386570958576342,
+      "grad_norm": 1.3783031852299132,
+      "learning_rate": 4.617724902375994e-06,
+      "loss": 0.2304,
+      "step": 3602
+    },
+    {
+      "epoch": 0.5388066397487663,
+      "grad_norm": 1.4168701539968467,
+      "learning_rate": 4.615310209947562e-06,
+      "loss": 0.2107,
+      "step": 3603
+    },
+    {
+      "epoch": 0.5389561836398983,
+      "grad_norm": 1.644250447091797,
+      "learning_rate": 4.612895607771064e-06,
+      "loss": 0.3363,
+      "step": 3604
+    },
+    {
+      "epoch": 0.5391057275310304,
+      "grad_norm": 1.7003203843951784,
+      "learning_rate": 4.610481096412985e-06,
+      "loss": 0.2935,
+      "step": 3605
+    },
+    {
+      "epoch": 0.5392552714221625,
+      "grad_norm": 1.9608275622063933,
+      "learning_rate": 4.608066676439794e-06,
+      "loss": 0.5222,
+      "step": 3606
+    },
+    {
+      "epoch": 0.5394048153132944,
+      "grad_norm": 1.6707441302180261,
+      "learning_rate": 4.605652348417938e-06,
+      "loss": 0.2279,
+      "step": 3607
+    },
+    {
+      "epoch": 0.5395543592044265,
+      "grad_norm": 1.3570608454800357,
+      "learning_rate": 4.603238112913842e-06,
+      "loss": 0.2076,
+      "step": 3608
+    },
+    {
+      "epoch": 0.5397039030955586,
+      "grad_norm": 1.6140789387503762,
+      "learning_rate": 4.60082397049391e-06,
+      "loss": 0.2158,
+      "step": 3609
+    },
+    {
+      "epoch": 0.5398534469866906,
+      "grad_norm": 2.129218227380532,
+      "learning_rate": 4.59840992172452e-06,
+      "loss": 0.6327,
+      "step": 3610
+    },
+    {
+      "epoch": 0.5400029908778227,
+      "grad_norm": 2.1454675155835017,
+      "learning_rate": 4.595995967172033e-06,
+      "loss": 0.6113,
+      "step": 3611
+    },
+    {
+      "epoch": 0.5401525347689546,
+      "grad_norm": 1.7145729893928194,
+      "learning_rate": 4.593582107402785e-06,
+      "loss": 0.3341,
+      "step": 3612
+    },
+    {
+      "epoch": 0.5403020786600867,
+      "grad_norm": 1.9012732049082708,
+      "learning_rate": 4.591168342983093e-06,
+      "loss": 0.2404,
+      "step": 3613
+    },
+    {
+      "epoch": 0.5404516225512188,
+      "grad_norm": 1.4498433671817794,
+      "learning_rate": 4.5887546744792476e-06,
+      "loss": 0.2012,
+      "step": 3614
+    },
+    {
+      "epoch": 0.5406011664423508,
+      "grad_norm": 1.7171612592641192,
+      "learning_rate": 4.586341102457518e-06,
+      "loss": 0.3926,
+      "step": 3615
+    },
+    {
+      "epoch": 0.5407507103334829,
+      "grad_norm": 1.6371246462787574,
+      "learning_rate": 4.583927627484153e-06,
+      "loss": 0.2237,
+      "step": 3616
+    },
+    {
+      "epoch": 0.540900254224615,
+      "grad_norm": 1.7189037373450475,
+      "learning_rate": 4.581514250125374e-06,
+      "loss": 0.1945,
+      "step": 3617
+    },
+    {
+      "epoch": 0.5410497981157469,
+      "grad_norm": 1.621330478342808,
+      "learning_rate": 4.579100970947387e-06,
+      "loss": 0.4194,
+      "step": 3618
+    },
+    {
+      "epoch": 0.541199342006879,
+      "grad_norm": 2.20237075123333,
+      "learning_rate": 4.57668779051637e-06,
+      "loss": 0.4388,
+      "step": 3619
+    },
+    {
+      "epoch": 0.5413488858980111,
+      "grad_norm": 2.5514565197624197,
+      "learning_rate": 4.574274709398475e-06,
+      "loss": 0.2575,
+      "step": 3620
+    },
+    {
+      "epoch": 0.5414984297891431,
+      "grad_norm": 1.5656114450122998,
+      "learning_rate": 4.571861728159836e-06,
+      "loss": 0.2278,
+      "step": 3621
+    },
+    {
+      "epoch": 0.5416479736802752,
+      "grad_norm": 1.9158231714402134,
+      "learning_rate": 4.5694488473665625e-06,
+      "loss": 0.241,
+      "step": 3622
+    },
+    {
+      "epoch": 0.5417975175714073,
+      "grad_norm": 1.8500844874936493,
+      "learning_rate": 4.567036067584738e-06,
+      "loss": 0.4013,
+      "step": 3623
+    },
+    {
+      "epoch": 0.5419470614625392,
+      "grad_norm": 1.536728323367019,
+      "learning_rate": 4.564623389380426e-06,
+      "loss": 0.2078,
+      "step": 3624
+    },
+    {
+      "epoch": 0.5420966053536713,
+      "grad_norm": 1.7953428909943243,
+      "learning_rate": 4.562210813319661e-06,
+      "loss": 0.2134,
+      "step": 3625
+    },
+    {
+      "epoch": 0.5422461492448033,
+      "grad_norm": 2.018110871488534,
+      "learning_rate": 4.559798339968459e-06,
+      "loss": 0.3426,
+      "step": 3626
+    },
+    {
+      "epoch": 0.5423956931359354,
+      "grad_norm": 1.3138686548561962,
+      "learning_rate": 4.557385969892809e-06,
+      "loss": 0.2118,
+      "step": 3627
+    },
+    {
+      "epoch": 0.5425452370270675,
+      "grad_norm": 2.3800526504743544,
+      "learning_rate": 4.554973703658676e-06,
+      "loss": 0.1592,
+      "step": 3628
+    },
+    {
+      "epoch": 0.5426947809181994,
+      "grad_norm": 1.1674207977655517,
+      "learning_rate": 4.552561541832002e-06,
+      "loss": 0.1865,
+      "step": 3629
+    },
+    {
+      "epoch": 0.5428443248093315,
+      "grad_norm": 1.6296355903967825,
+      "learning_rate": 4.550149484978701e-06,
+      "loss": 0.2245,
+      "step": 3630
+    },
+    {
+      "epoch": 0.5429938687004636,
+      "grad_norm": 1.546275603117866,
+      "learning_rate": 4.547737533664667e-06,
+      "loss": 0.1881,
+      "step": 3631
+    },
+    {
+      "epoch": 0.5431434125915956,
+      "grad_norm": 1.8563039566315576,
+      "learning_rate": 4.545325688455766e-06,
+      "loss": 0.2158,
+      "step": 3632
+    },
+    {
+      "epoch": 0.5432929564827277,
+      "grad_norm": 1.909384983273893,
+      "learning_rate": 4.54291394991784e-06,
+      "loss": 0.1928,
+      "step": 3633
+    },
+    {
+      "epoch": 0.5434425003738598,
+      "grad_norm": 1.7306613061519398,
+      "learning_rate": 4.540502318616708e-06,
+      "loss": 0.2385,
+      "step": 3634
+    },
+    {
+      "epoch": 0.5435920442649917,
+      "grad_norm": 1.5487667458820433,
+      "learning_rate": 4.538090795118159e-06,
+      "loss": 0.2039,
+      "step": 3635
+    },
+    {
+      "epoch": 0.5437415881561238,
+      "grad_norm": 1.8407684602188683,
+      "learning_rate": 4.535679379987961e-06,
+      "loss": 0.4428,
+      "step": 3636
+    },
+    {
+      "epoch": 0.5438911320472559,
+      "grad_norm": 1.8693971467850499,
+      "learning_rate": 4.533268073791854e-06,
+      "loss": 0.3359,
+      "step": 3637
+    },
+    {
+      "epoch": 0.5440406759383879,
+      "grad_norm": 1.540410429421942,
+      "learning_rate": 4.530856877095557e-06,
+      "loss": 0.3461,
+      "step": 3638
+    },
+    {
+      "epoch": 0.54419021982952,
+      "grad_norm": 1.4587300754337025,
+      "learning_rate": 4.5284457904647585e-06,
+      "loss": 0.3748,
+      "step": 3639
+    },
+    {
+      "epoch": 0.544339763720652,
+      "grad_norm": 1.574943101324438,
+      "learning_rate": 4.526034814465121e-06,
+      "loss": 0.3262,
+      "step": 3640
+    },
+    {
+      "epoch": 0.544489307611784,
+      "grad_norm": 1.675685128431864,
+      "learning_rate": 4.523623949662284e-06,
+      "loss": 0.302,
+      "step": 3641
+    },
+    {
+      "epoch": 0.5446388515029161,
+      "grad_norm": 1.2926390795199603,
+      "learning_rate": 4.521213196621858e-06,
+      "loss": 0.2833,
+      "step": 3642
+    },
+    {
+      "epoch": 0.5447883953940481,
+      "grad_norm": 1.4062065912540513,
+      "learning_rate": 4.518802555909432e-06,
+      "loss": 0.2638,
+      "step": 3643
+    },
+    {
+      "epoch": 0.5449379392851802,
+      "grad_norm": 1.5376659922043572,
+      "learning_rate": 4.516392028090564e-06,
+      "loss": 0.3906,
+      "step": 3644
+    },
+    {
+      "epoch": 0.5450874831763123,
+      "grad_norm": 1.4973884587866484,
+      "learning_rate": 4.513981613730786e-06,
+      "loss": 0.2047,
+      "step": 3645
+    },
+    {
+      "epoch": 0.5452370270674443,
+      "grad_norm": 1.3383658073421694,
+      "learning_rate": 4.511571313395605e-06,
+      "loss": 0.3175,
+      "step": 3646
+    },
+    {
+      "epoch": 0.5453865709585763,
+      "grad_norm": 1.8556149033656897,
+      "learning_rate": 4.509161127650502e-06,
+      "loss": 0.3392,
+      "step": 3647
+    },
+    {
+      "epoch": 0.5455361148497084,
+      "grad_norm": 1.80199563542535,
+      "learning_rate": 4.506751057060928e-06,
+      "loss": 0.2334,
+      "step": 3648
+    },
+    {
+      "epoch": 0.5456856587408404,
+      "grad_norm": 1.5120205146776637,
+      "learning_rate": 4.504341102192311e-06,
+      "loss": 0.2754,
+      "step": 3649
+    },
+    {
+      "epoch": 0.5458352026319725,
+      "grad_norm": 1.2694153160565378,
+      "learning_rate": 4.501931263610047e-06,
+      "loss": 0.1759,
+      "step": 3650
+    },
+    {
+      "epoch": 0.5459847465231046,
+      "grad_norm": 2.11280085732116,
+      "learning_rate": 4.499521541879508e-06,
+      "loss": 0.3389,
+      "step": 3651
+    },
+    {
+      "epoch": 0.5461342904142366,
+      "grad_norm": 2.684582629406241,
+      "learning_rate": 4.49711193756604e-06,
+      "loss": 0.2197,
+      "step": 3652
+    },
+    {
+      "epoch": 0.5462838343053686,
+      "grad_norm": 2.1375854659143543,
+      "learning_rate": 4.494702451234958e-06,
+      "loss": 0.4994,
+      "step": 3653
+    },
+    {
+      "epoch": 0.5464333781965006,
+      "grad_norm": 1.7511040236402404,
+      "learning_rate": 4.492293083451551e-06,
+      "loss": 0.4358,
+      "step": 3654
+    },
+    {
+      "epoch": 0.5465829220876327,
+      "grad_norm": 1.9004355383149747,
+      "learning_rate": 4.489883834781079e-06,
+      "loss": 0.3991,
+      "step": 3655
+    },
+    {
+      "epoch": 0.5467324659787648,
+      "grad_norm": 1.9508214870790095,
+      "learning_rate": 4.4874747057887756e-06,
+      "loss": 0.2123,
+      "step": 3656
+    },
+    {
+      "epoch": 0.5468820098698968,
+      "grad_norm": 1.393009322340977,
+      "learning_rate": 4.4850656970398445e-06,
+      "loss": 0.1935,
+      "step": 3657
+    },
+    {
+      "epoch": 0.5470315537610289,
+      "grad_norm": 2.2204350802082264,
+      "learning_rate": 4.4826568090994635e-06,
+      "loss": 0.5944,
+      "step": 3658
+    },
+    {
+      "epoch": 0.5471810976521609,
+      "grad_norm": 1.3876428453270266,
+      "learning_rate": 4.4802480425327815e-06,
+      "loss": 0.2091,
+      "step": 3659
+    },
+    {
+      "epoch": 0.5473306415432929,
+      "grad_norm": 1.5091104368877524,
+      "learning_rate": 4.47783939790492e-06,
+      "loss": 0.2027,
+      "step": 3660
+    },
+    {
+      "epoch": 0.547480185434425,
+      "grad_norm": 1.4782934563639556,
+      "learning_rate": 4.475430875780966e-06,
+      "loss": 0.1921,
+      "step": 3661
+    },
+    {
+      "epoch": 0.5476297293255571,
+      "grad_norm": 1.3476666823273606,
+      "learning_rate": 4.4730224767259835e-06,
+      "loss": 0.2987,
+      "step": 3662
+    },
+    {
+      "epoch": 0.5477792732166891,
+      "grad_norm": 1.497176430979183,
+      "learning_rate": 4.470614201305007e-06,
+      "loss": 0.1607,
+      "step": 3663
+    },
+    {
+      "epoch": 0.5479288171078212,
+      "grad_norm": 1.3796593740972627,
+      "learning_rate": 4.46820605008304e-06,
+      "loss": 0.1941,
+      "step": 3664
+    },
+    {
+      "epoch": 0.5480783609989532,
+      "grad_norm": 1.6361789769261383,
+      "learning_rate": 4.465798023625062e-06,
+      "loss": 0.2117,
+      "step": 3665
+    },
+    {
+      "epoch": 0.5482279048900852,
+      "grad_norm": 1.7401255092969485,
+      "learning_rate": 4.463390122496012e-06,
+      "loss": 0.3168,
+      "step": 3666
+    },
+    {
+      "epoch": 0.5483774487812173,
+      "grad_norm": 1.6768308255934834,
+      "learning_rate": 4.460982347260813e-06,
+      "loss": 0.4315,
+      "step": 3667
+    },
+    {
+      "epoch": 0.5485269926723494,
+      "grad_norm": 1.1769237227237928,
+      "learning_rate": 4.458574698484349e-06,
+      "loss": 0.1979,
+      "step": 3668
+    },
+    {
+      "epoch": 0.5486765365634814,
+      "grad_norm": 1.8181887340685052,
+      "learning_rate": 4.456167176731478e-06,
+      "loss": 0.3617,
+      "step": 3669
+    },
+    {
+      "epoch": 0.5488260804546135,
+      "grad_norm": 1.8597276898084434,
+      "learning_rate": 4.453759782567031e-06,
+      "loss": 0.4951,
+      "step": 3670
+    },
+    {
+      "epoch": 0.5489756243457454,
+      "grad_norm": 2.0804206601929747,
+      "learning_rate": 4.451352516555802e-06,
+      "loss": 0.3081,
+      "step": 3671
+    },
+    {
+      "epoch": 0.5491251682368775,
+      "grad_norm": 1.2510363455073479,
+      "learning_rate": 4.44894537926256e-06,
+      "loss": 0.1942,
+      "step": 3672
+    },
+    {
+      "epoch": 0.5492747121280096,
+      "grad_norm": 1.7610349744949918,
+      "learning_rate": 4.446538371252042e-06,
+      "loss": 0.3415,
+      "step": 3673
+    },
+    {
+      "epoch": 0.5494242560191416,
+      "grad_norm": 1.4549183546588127,
+      "learning_rate": 4.444131493088956e-06,
+      "loss": 0.3443,
+      "step": 3674
+    },
+    {
+      "epoch": 0.5495737999102737,
+      "grad_norm": 1.5300081335892495,
+      "learning_rate": 4.441724745337982e-06,
+      "loss": 0.1803,
+      "step": 3675
+    },
+    {
+      "epoch": 0.5497233438014058,
+      "grad_norm": 1.8473890728356732,
+      "learning_rate": 4.43931812856376e-06,
+      "loss": 0.4096,
+      "step": 3676
+    },
+    {
+      "epoch": 0.5498728876925377,
+      "grad_norm": 1.082783619934534,
+      "learning_rate": 4.436911643330909e-06,
+      "loss": 0.1938,
+      "step": 3677
+    },
+    {
+      "epoch": 0.5500224315836698,
+      "grad_norm": 1.8449280983656142,
+      "learning_rate": 4.434505290204013e-06,
+      "loss": 0.2759,
+      "step": 3678
+    },
+    {
+      "epoch": 0.5501719754748019,
+      "grad_norm": 1.8510421150420215,
+      "learning_rate": 4.432099069747625e-06,
+      "loss": 0.3467,
+      "step": 3679
+    },
+    {
+      "epoch": 0.5503215193659339,
+      "grad_norm": 1.8763952805675503,
+      "learning_rate": 4.429692982526272e-06,
+      "loss": 0.3869,
+      "step": 3680
+    },
+    {
+      "epoch": 0.550471063257066,
+      "grad_norm": 1.8434682323650826,
+      "learning_rate": 4.4272870291044385e-06,
+      "loss": 0.3162,
+      "step": 3681
+    },
+    {
+      "epoch": 0.550620607148198,
+      "grad_norm": 1.619398372644164,
+      "learning_rate": 4.424881210046588e-06,
+      "loss": 0.3891,
+      "step": 3682
+    },
+    {
+      "epoch": 0.55077015103933,
+      "grad_norm": 1.0183619154944885,
+      "learning_rate": 4.4224755259171474e-06,
+      "loss": 0.2017,
+      "step": 3683
+    },
+    {
+      "epoch": 0.5509196949304621,
+      "grad_norm": 1.4729547476353326,
+      "learning_rate": 4.420069977280515e-06,
+      "loss": 0.182,
+      "step": 3684
+    },
+    {
+      "epoch": 0.5510692388215941,
+      "grad_norm": 1.5871225167325878,
+      "learning_rate": 4.417664564701056e-06,
+      "loss": 0.1883,
+      "step": 3685
+    },
+    {
+      "epoch": 0.5512187827127262,
+      "grad_norm": 1.6250502184297433,
+      "learning_rate": 4.415259288743101e-06,
+      "loss": 0.2383,
+      "step": 3686
+    },
+    {
+      "epoch": 0.5513683266038583,
+      "grad_norm": 1.1626293846136282,
+      "learning_rate": 4.412854149970952e-06,
+      "loss": 0.2129,
+      "step": 3687
+    },
+    {
+      "epoch": 0.5515178704949902,
+      "grad_norm": 2.12434434593106,
+      "learning_rate": 4.410449148948878e-06,
+      "loss": 0.6634,
+      "step": 3688
+    },
+    {
+      "epoch": 0.5516674143861223,
+      "grad_norm": 1.779837424684205,
+      "learning_rate": 4.4080442862411155e-06,
+      "loss": 0.492,
+      "step": 3689
+    },
+    {
+      "epoch": 0.5518169582772544,
+      "grad_norm": 1.6280486779070482,
+      "learning_rate": 4.405639562411869e-06,
+      "loss": 0.3824,
+      "step": 3690
+    },
+    {
+      "epoch": 0.5519665021683864,
+      "grad_norm": 1.2269128736362505,
+      "learning_rate": 4.4032349780253094e-06,
+      "loss": 0.1843,
+      "step": 3691
+    },
+    {
+      "epoch": 0.5521160460595185,
+      "grad_norm": 1.5707493169747542,
+      "learning_rate": 4.400830533645575e-06,
+      "loss": 0.2989,
+      "step": 3692
+    },
+    {
+      "epoch": 0.5522655899506506,
+      "grad_norm": 1.5702555536980602,
+      "learning_rate": 4.3984262298367706e-06,
+      "loss": 0.3377,
+      "step": 3693
+    },
+    {
+      "epoch": 0.5524151338417825,
+      "grad_norm": 2.039220723411418,
+      "learning_rate": 4.3960220671629704e-06,
+      "loss": 0.2134,
+      "step": 3694
+    },
+    {
+      "epoch": 0.5525646777329146,
+      "grad_norm": 2.02853402713955,
+      "learning_rate": 4.393618046188217e-06,
+      "loss": 0.2223,
+      "step": 3695
+    },
+    {
+      "epoch": 0.5527142216240467,
+      "grad_norm": 2.1312606881172904,
+      "learning_rate": 4.391214167476511e-06,
+      "loss": 0.58,
+      "step": 3696
+    },
+    {
+      "epoch": 0.5528637655151787,
+      "grad_norm": 1.5335152861023273,
+      "learning_rate": 4.388810431591829e-06,
+      "loss": 0.3053,
+      "step": 3697
+    },
+    {
+      "epoch": 0.5530133094063108,
+      "grad_norm": 1.895118001155134,
+      "learning_rate": 4.386406839098108e-06,
+      "loss": 0.3881,
+      "step": 3698
+    },
+    {
+      "epoch": 0.5531628532974427,
+      "grad_norm": 1.3992028246664994,
+      "learning_rate": 4.384003390559258e-06,
+      "loss": 0.196,
+      "step": 3699
+    },
+    {
+      "epoch": 0.5533123971885748,
+      "grad_norm": 1.421234940471731,
+      "learning_rate": 4.381600086539151e-06,
+      "loss": 0.1791,
+      "step": 3700
+    },
+    {
+      "epoch": 0.5534619410797069,
+      "grad_norm": 1.7926872298354135,
+      "learning_rate": 4.379196927601622e-06,
+      "loss": 0.4514,
+      "step": 3701
+    },
+    {
+      "epoch": 0.5536114849708389,
+      "grad_norm": 1.5046636185379967,
+      "learning_rate": 4.376793914310476e-06,
+      "loss": 0.3117,
+      "step": 3702
+    },
+    {
+      "epoch": 0.553761028861971,
+      "grad_norm": 1.4200654672168842,
+      "learning_rate": 4.3743910472294835e-06,
+      "loss": 0.147,
+      "step": 3703
+    },
+    {
+      "epoch": 0.5539105727531031,
+      "grad_norm": 2.124991711432056,
+      "learning_rate": 4.371988326922381e-06,
+      "loss": 0.711,
+      "step": 3704
+    },
+    {
+      "epoch": 0.554060116644235,
+      "grad_norm": 1.2259532651560952,
+      "learning_rate": 4.3695857539528705e-06,
+      "loss": 0.167,
+      "step": 3705
+    },
+    {
+      "epoch": 0.5542096605353671,
+      "grad_norm": 1.71466572734677,
+      "learning_rate": 4.367183328884616e-06,
+      "loss": 0.2966,
+      "step": 3706
+    },
+    {
+      "epoch": 0.5543592044264992,
+      "grad_norm": 1.4687978199012193,
+      "learning_rate": 4.364781052281252e-06,
+      "loss": 0.3417,
+      "step": 3707
+    },
+    {
+      "epoch": 0.5545087483176312,
+      "grad_norm": 1.4014204433835475,
+      "learning_rate": 4.362378924706374e-06,
+      "loss": 0.219,
+      "step": 3708
+    },
+    {
+      "epoch": 0.5546582922087633,
+      "grad_norm": 1.3260680988141613,
+      "learning_rate": 4.3599769467235455e-06,
+      "loss": 0.2324,
+      "step": 3709
+    },
+    {
+      "epoch": 0.5548078360998954,
+      "grad_norm": 1.9733255365162365,
+      "learning_rate": 4.357575118896296e-06,
+      "loss": 0.5631,
+      "step": 3710
+    },
+    {
+      "epoch": 0.5549573799910273,
+      "grad_norm": 1.704368041443617,
+      "learning_rate": 4.355173441788112e-06,
+      "loss": 0.2482,
+      "step": 3711
+    },
+    {
+      "epoch": 0.5551069238821594,
+      "grad_norm": 1.3602084531631606,
+      "learning_rate": 4.352771915962454e-06,
+      "loss": 0.2085,
+      "step": 3712
+    },
+    {
+      "epoch": 0.5552564677732914,
+      "grad_norm": 1.1924270602927478,
+      "learning_rate": 4.350370541982742e-06,
+      "loss": 0.1791,
+      "step": 3713
+    },
+    {
+      "epoch": 0.5554060116644235,
+      "grad_norm": 1.6609565690705255,
+      "learning_rate": 4.3479693204123604e-06,
+      "loss": 0.4581,
+      "step": 3714
+    },
+    {
+      "epoch": 0.5555555555555556,
+      "grad_norm": 1.3086925554693893,
+      "learning_rate": 4.345568251814663e-06,
+      "loss": 0.1964,
+      "step": 3715
+    },
+    {
+      "epoch": 0.5557050994466876,
+      "grad_norm": 1.459973940584078,
+      "learning_rate": 4.343167336752959e-06,
+      "loss": 0.1694,
+      "step": 3716
+    },
+    {
+      "epoch": 0.5558546433378196,
+      "grad_norm": 1.0197965232991946,
+      "learning_rate": 4.340766575790528e-06,
+      "loss": 0.1545,
+      "step": 3717
+    },
+    {
+      "epoch": 0.5560041872289517,
+      "grad_norm": 1.710086881050594,
+      "learning_rate": 4.338365969490611e-06,
+      "loss": 0.479,
+      "step": 3718
+    },
+    {
+      "epoch": 0.5561537311200837,
+      "grad_norm": 1.2458190047063138,
+      "learning_rate": 4.3359655184164125e-06,
+      "loss": 0.2591,
+      "step": 3719
+    },
+    {
+      "epoch": 0.5563032750112158,
+      "grad_norm": 1.7265111518577587,
+      "learning_rate": 4.333565223131107e-06,
+      "loss": 0.3026,
+      "step": 3720
+    },
+    {
+      "epoch": 0.5564528189023479,
+      "grad_norm": 1.7006783672890977,
+      "learning_rate": 4.331165084197821e-06,
+      "loss": 0.4155,
+      "step": 3721
+    },
+    {
+      "epoch": 0.5566023627934799,
+      "grad_norm": 2.24584551559027,
+      "learning_rate": 4.328765102179651e-06,
+      "loss": 0.6092,
+      "step": 3722
+    },
+    {
+      "epoch": 0.5567519066846119,
+      "grad_norm": 1.5686699024208746,
+      "learning_rate": 4.326365277639657e-06,
+      "loss": 0.3227,
+      "step": 3723
+    },
+    {
+      "epoch": 0.556901450575744,
+      "grad_norm": 1.216697878660411,
+      "learning_rate": 4.323965611140863e-06,
+      "loss": 0.3495,
+      "step": 3724
+    },
+    {
+      "epoch": 0.557050994466876,
+      "grad_norm": 2.1454702555477727,
+      "learning_rate": 4.32156610324625e-06,
+      "loss": 0.2151,
+      "step": 3725
+    },
+    {
+      "epoch": 0.5572005383580081,
+      "grad_norm": 1.4848464819357048,
+      "learning_rate": 4.319166754518768e-06,
+      "loss": 0.4727,
+      "step": 3726
+    },
+    {
+      "epoch": 0.5573500822491402,
+      "grad_norm": 3.270072864805568,
+      "learning_rate": 4.316767565521327e-06,
+      "loss": 0.2948,
+      "step": 3727
+    },
+    {
+      "epoch": 0.5574996261402722,
+      "grad_norm": 1.5653364995014118,
+      "learning_rate": 4.314368536816799e-06,
+      "loss": 0.4562,
+      "step": 3728
+    },
+    {
+      "epoch": 0.5576491700314042,
+      "grad_norm": 2.0500993739442412,
+      "learning_rate": 4.31196966896802e-06,
+      "loss": 0.3331,
+      "step": 3729
+    },
+    {
+      "epoch": 0.5577987139225362,
+      "grad_norm": 2.201295915170263,
+      "learning_rate": 4.3095709625377885e-06,
+      "loss": 0.3756,
+      "step": 3730
+    },
+    {
+      "epoch": 0.5579482578136683,
+      "grad_norm": 1.6854473805153678,
+      "learning_rate": 4.307172418088863e-06,
+      "loss": 0.353,
+      "step": 3731
+    },
+    {
+      "epoch": 0.5580978017048004,
+      "grad_norm": 1.911466231230202,
+      "learning_rate": 4.304774036183964e-06,
+      "loss": 0.5574,
+      "step": 3732
+    },
+    {
+      "epoch": 0.5582473455959324,
+      "grad_norm": 1.929179136209855,
+      "learning_rate": 4.3023758173857764e-06,
+      "loss": 0.3294,
+      "step": 3733
+    },
+    {
+      "epoch": 0.5583968894870645,
+      "grad_norm": 1.3667926899468956,
+      "learning_rate": 4.299977762256946e-06,
+      "loss": 0.2124,
+      "step": 3734
+    },
+    {
+      "epoch": 0.5585464333781965,
+      "grad_norm": 2.697037628503975,
+      "learning_rate": 4.29757987136008e-06,
+      "loss": 0.2051,
+      "step": 3735
+    },
+    {
+      "epoch": 0.5586959772693285,
+      "grad_norm": 1.6135252013042591,
+      "learning_rate": 4.295182145257744e-06,
+      "loss": 0.3395,
+      "step": 3736
+    },
+    {
+      "epoch": 0.5588455211604606,
+      "grad_norm": 1.508358238955974,
+      "learning_rate": 4.292784584512471e-06,
+      "loss": 0.1759,
+      "step": 3737
+    },
+    {
+      "epoch": 0.5589950650515927,
+      "grad_norm": 1.8796805295236263,
+      "learning_rate": 4.290387189686748e-06,
+      "loss": 0.476,
+      "step": 3738
+    },
+    {
+      "epoch": 0.5591446089427247,
+      "grad_norm": 1.5709579431840082,
+      "learning_rate": 4.287989961343029e-06,
+      "loss": 0.3692,
+      "step": 3739
+    },
+    {
+      "epoch": 0.5592941528338568,
+      "grad_norm": 1.5271978772058465,
+      "learning_rate": 4.28559290004373e-06,
+      "loss": 0.2455,
+      "step": 3740
+    },
+    {
+      "epoch": 0.5594436967249888,
+      "grad_norm": 1.5965798116355054,
+      "learning_rate": 4.283196006351219e-06,
+      "loss": 0.3172,
+      "step": 3741
+    },
+    {
+      "epoch": 0.5595932406161208,
+      "grad_norm": 1.5747075138534872,
+      "learning_rate": 4.280799280827833e-06,
+      "loss": 0.3397,
+      "step": 3742
+    },
+    {
+      "epoch": 0.5597427845072529,
+      "grad_norm": 2.348694227529921,
+      "learning_rate": 4.278402724035868e-06,
+      "loss": 0.221,
+      "step": 3743
+    },
+    {
+      "epoch": 0.5598923283983849,
+      "grad_norm": 1.459683116901361,
+      "learning_rate": 4.276006336537577e-06,
+      "loss": 0.319,
+      "step": 3744
+    },
+    {
+      "epoch": 0.560041872289517,
+      "grad_norm": 1.759239579641639,
+      "learning_rate": 4.273610118895178e-06,
+      "loss": 0.2453,
+      "step": 3745
+    },
+    {
+      "epoch": 0.560191416180649,
+      "grad_norm": 1.744850048587745,
+      "learning_rate": 4.2712140716708465e-06,
+      "loss": 0.3606,
+      "step": 3746
+    },
+    {
+      "epoch": 0.560340960071781,
+      "grad_norm": 1.7930423330861387,
+      "learning_rate": 4.268818195426715e-06,
+      "loss": 0.3549,
+      "step": 3747
+    },
+    {
+      "epoch": 0.5604905039629131,
+      "grad_norm": 1.8638965502443476,
+      "learning_rate": 4.266422490724883e-06,
+      "loss": 0.2108,
+      "step": 3748
+    },
+    {
+      "epoch": 0.5606400478540452,
+      "grad_norm": 1.6929676082309733,
+      "learning_rate": 4.264026958127404e-06,
+      "loss": 0.3148,
+      "step": 3749
+    },
+    {
+      "epoch": 0.5607895917451772,
+      "grad_norm": 1.7573506914261001,
+      "learning_rate": 4.261631598196293e-06,
+      "loss": 0.3503,
+      "step": 3750
+    },
+    {
+      "epoch": 0.5609391356363093,
+      "grad_norm": 1.2550353673685644,
+      "learning_rate": 4.259236411493526e-06,
+      "loss": 0.2267,
+      "step": 3751
+    },
+    {
+      "epoch": 0.5610886795274413,
+      "grad_norm": 1.396447043820233,
+      "learning_rate": 4.256841398581035e-06,
+      "loss": 0.338,
+      "step": 3752
+    },
+    {
+      "epoch": 0.5612382234185733,
+      "grad_norm": 1.4047858629698926,
+      "learning_rate": 4.254446560020714e-06,
+      "loss": 0.1939,
+      "step": 3753
+    },
+    {
+      "epoch": 0.5613877673097054,
+      "grad_norm": 1.7448713388804253,
+      "learning_rate": 4.252051896374416e-06,
+      "loss": 0.2867,
+      "step": 3754
+    },
+    {
+      "epoch": 0.5615373112008375,
+      "grad_norm": 1.190396991434064,
+      "learning_rate": 4.24965740820395e-06,
+      "loss": 0.1743,
+      "step": 3755
+    },
+    {
+      "epoch": 0.5616868550919695,
+      "grad_norm": 1.4830547911619545,
+      "learning_rate": 4.2472630960710906e-06,
+      "loss": 0.2371,
+      "step": 3756
+    },
+    {
+      "epoch": 0.5618363989831016,
+      "grad_norm": 1.6803677800346988,
+      "learning_rate": 4.24486896053756e-06,
+      "loss": 0.4105,
+      "step": 3757
+    },
+    {
+      "epoch": 0.5619859428742335,
+      "grad_norm": 1.7843402963560555,
+      "learning_rate": 4.2424750021650505e-06,
+      "loss": 0.1911,
+      "step": 3758
+    },
+    {
+      "epoch": 0.5621354867653656,
+      "grad_norm": 1.3070197531860872,
+      "learning_rate": 4.240081221515203e-06,
+      "loss": 0.2109,
+      "step": 3759
+    },
+    {
+      "epoch": 0.5622850306564977,
+      "grad_norm": 1.4646566398646295,
+      "learning_rate": 4.237687619149627e-06,
+      "loss": 0.2391,
+      "step": 3760
+    },
+    {
+      "epoch": 0.5624345745476297,
+      "grad_norm": 1.3749863369219528,
+      "learning_rate": 4.2352941956298835e-06,
+      "loss": 0.3234,
+      "step": 3761
+    },
+    {
+      "epoch": 0.5625841184387618,
+      "grad_norm": 1.6256377087587262,
+      "learning_rate": 4.23290095151749e-06,
+      "loss": 0.3059,
+      "step": 3762
+    },
+    {
+      "epoch": 0.5627336623298939,
+      "grad_norm": 1.0526622135292967,
+      "learning_rate": 4.230507887373926e-06,
+      "loss": 0.1609,
+      "step": 3763
+    },
+    {
+      "epoch": 0.5628832062210258,
+      "grad_norm": 1.1601833699970812,
+      "learning_rate": 4.228115003760627e-06,
+      "loss": 0.2103,
+      "step": 3764
+    },
+    {
+      "epoch": 0.5630327501121579,
+      "grad_norm": 1.3805756715861772,
+      "learning_rate": 4.225722301238987e-06,
+      "loss": 0.2133,
+      "step": 3765
+    },
+    {
+      "epoch": 0.56318229400329,
+      "grad_norm": 1.2285578315653152,
+      "learning_rate": 4.223329780370359e-06,
+      "loss": 0.2906,
+      "step": 3766
+    },
+    {
+      "epoch": 0.563331837894422,
+      "grad_norm": 1.0545194721595736,
+      "learning_rate": 4.220937441716048e-06,
+      "loss": 0.2419,
+      "step": 3767
+    },
+    {
+      "epoch": 0.5634813817855541,
+      "grad_norm": 1.8754558366761425,
+      "learning_rate": 4.218545285837321e-06,
+      "loss": 0.3342,
+      "step": 3768
+    },
+    {
+      "epoch": 0.5636309256766862,
+      "grad_norm": 1.879834063009346,
+      "learning_rate": 4.216153313295401e-06,
+      "loss": 0.3649,
+      "step": 3769
+    },
+    {
+      "epoch": 0.5637804695678181,
+      "grad_norm": 1.373634860925274,
+      "learning_rate": 4.213761524651469e-06,
+      "loss": 0.2021,
+      "step": 3770
+    },
+    {
+      "epoch": 0.5639300134589502,
+      "grad_norm": 1.6925748087720431,
+      "learning_rate": 4.211369920466661e-06,
+      "loss": 0.5069,
+      "step": 3771
+    },
+    {
+      "epoch": 0.5640795573500822,
+      "grad_norm": 1.217939544459708,
+      "learning_rate": 4.2089785013020686e-06,
+      "loss": 0.1998,
+      "step": 3772
+    },
+    {
+      "epoch": 0.5642291012412143,
+      "grad_norm": 1.926661562944481,
+      "learning_rate": 4.206587267718743e-06,
+      "loss": 0.4489,
+      "step": 3773
+    },
+    {
+      "epoch": 0.5643786451323464,
+      "grad_norm": 1.6974792521617525,
+      "learning_rate": 4.204196220277692e-06,
+      "loss": 0.3709,
+      "step": 3774
+    },
+    {
+      "epoch": 0.5645281890234783,
+      "grad_norm": 1.3914955473555144,
+      "learning_rate": 4.2018053595398766e-06,
+      "loss": 0.3334,
+      "step": 3775
+    },
+    {
+      "epoch": 0.5646777329146104,
+      "grad_norm": 1.8360783705539188,
+      "learning_rate": 4.199414686066219e-06,
+      "loss": 0.3313,
+      "step": 3776
+    },
+    {
+      "epoch": 0.5648272768057425,
+      "grad_norm": 1.623276695386809,
+      "learning_rate": 4.1970242004175884e-06,
+      "loss": 0.372,
+      "step": 3777
+    },
+    {
+      "epoch": 0.5649768206968745,
+      "grad_norm": 1.6066517149442552,
+      "learning_rate": 4.194633903154819e-06,
+      "loss": 0.1992,
+      "step": 3778
+    },
+    {
+      "epoch": 0.5651263645880066,
+      "grad_norm": 1.5792435940178833,
+      "learning_rate": 4.192243794838698e-06,
+      "loss": 0.4063,
+      "step": 3779
+    },
+    {
+      "epoch": 0.5652759084791387,
+      "grad_norm": 0.9997425194899473,
+      "learning_rate": 4.189853876029966e-06,
+      "loss": 0.3228,
+      "step": 3780
+    },
+    {
+      "epoch": 0.5654254523702706,
+      "grad_norm": 2.063913929954077,
+      "learning_rate": 4.187464147289326e-06,
+      "loss": 0.3574,
+      "step": 3781
+    },
+    {
+      "epoch": 0.5655749962614027,
+      "grad_norm": 1.635138880368131,
+      "learning_rate": 4.185074609177425e-06,
+      "loss": 0.3418,
+      "step": 3782
+    },
+    {
+      "epoch": 0.5657245401525348,
+      "grad_norm": 1.7099814378347198,
+      "learning_rate": 4.182685262254875e-06,
+      "loss": 0.3688,
+      "step": 3783
+    },
+    {
+      "epoch": 0.5658740840436668,
+      "grad_norm": 1.048172395904991,
+      "learning_rate": 4.180296107082238e-06,
+      "loss": 0.1967,
+      "step": 3784
+    },
+    {
+      "epoch": 0.5660236279347989,
+      "grad_norm": 1.4773704699879262,
+      "learning_rate": 4.177907144220036e-06,
+      "loss": 0.2355,
+      "step": 3785
+    },
+    {
+      "epoch": 0.566173171825931,
+      "grad_norm": 1.7211220022506921,
+      "learning_rate": 4.1755183742287404e-06,
+      "loss": 0.4811,
+      "step": 3786
+    },
+    {
+      "epoch": 0.5663227157170629,
+      "grad_norm": 1.5450919500167295,
+      "learning_rate": 4.173129797668779e-06,
+      "loss": 0.3029,
+      "step": 3787
+    },
+    {
+      "epoch": 0.566472259608195,
+      "grad_norm": 1.3702653273245642,
+      "learning_rate": 4.1707414151005374e-06,
+      "loss": 0.2632,
+      "step": 3788
+    },
+    {
+      "epoch": 0.566621803499327,
+      "grad_norm": 1.6529737223038268,
+      "learning_rate": 4.1683532270843505e-06,
+      "loss": 0.4421,
+      "step": 3789
+    },
+    {
+      "epoch": 0.5667713473904591,
+      "grad_norm": 1.727830283297176,
+      "learning_rate": 4.1659652341805115e-06,
+      "loss": 0.4467,
+      "step": 3790
+    },
+    {
+      "epoch": 0.5669208912815912,
+      "grad_norm": 1.7423370579387567,
+      "learning_rate": 4.163577436949267e-06,
+      "loss": 0.344,
+      "step": 3791
+    },
+    {
+      "epoch": 0.5670704351727232,
+      "grad_norm": 1.839263370923161,
+      "learning_rate": 4.161189835950816e-06,
+      "loss": 0.4408,
+      "step": 3792
+    },
+    {
+      "epoch": 0.5672199790638552,
+      "grad_norm": 1.2231225047916932,
+      "learning_rate": 4.158802431745314e-06,
+      "loss": 0.2086,
+      "step": 3793
+    },
+    {
+      "epoch": 0.5673695229549873,
+      "grad_norm": 1.6834296845774626,
+      "learning_rate": 4.156415224892868e-06,
+      "loss": 0.2075,
+      "step": 3794
+    },
+    {
+      "epoch": 0.5675190668461193,
+      "grad_norm": 1.63695734693028,
+      "learning_rate": 4.1540282159535405e-06,
+      "loss": 0.3609,
+      "step": 3795
+    },
+    {
+      "epoch": 0.5676686107372514,
+      "grad_norm": 4.050254883222083,
+      "learning_rate": 4.151641405487348e-06,
+      "loss": 0.4249,
+      "step": 3796
+    },
+    {
+      "epoch": 0.5678181546283835,
+      "grad_norm": 2.012583096932945,
+      "learning_rate": 4.149254794054256e-06,
+      "loss": 0.3465,
+      "step": 3797
+    },
+    {
+      "epoch": 0.5679676985195155,
+      "grad_norm": 1.3031350746325656,
+      "learning_rate": 4.146868382214188e-06,
+      "loss": 0.2125,
+      "step": 3798
+    },
+    {
+      "epoch": 0.5681172424106475,
+      "grad_norm": 1.132012454693644,
+      "learning_rate": 4.14448217052702e-06,
+      "loss": 0.2188,
+      "step": 3799
+    },
+    {
+      "epoch": 0.5682667863017796,
+      "grad_norm": 1.7236076504434965,
+      "learning_rate": 4.142096159552578e-06,
+      "loss": 0.3361,
+      "step": 3800
+    },
+    {
+      "epoch": 0.5684163301929116,
+      "grad_norm": 1.8633241948391013,
+      "learning_rate": 4.139710349850649e-06,
+      "loss": 0.2458,
+      "step": 3801
+    },
+    {
+      "epoch": 0.5685658740840437,
+      "grad_norm": 1.625642374423236,
+      "learning_rate": 4.13732474198096e-06,
+      "loss": 0.287,
+      "step": 3802
+    },
+    {
+      "epoch": 0.5687154179751757,
+      "grad_norm": 1.3510065589123672,
+      "learning_rate": 4.134939336503201e-06,
+      "loss": 0.2177,
+      "step": 3803
+    },
+    {
+      "epoch": 0.5688649618663078,
+      "grad_norm": 1.3967447851707402,
+      "learning_rate": 4.13255413397701e-06,
+      "loss": 0.2053,
+      "step": 3804
+    },
+    {
+      "epoch": 0.5690145057574398,
+      "grad_norm": 3.731031445311677,
+      "learning_rate": 4.130169134961979e-06,
+      "loss": 0.467,
+      "step": 3805
+    },
+    {
+      "epoch": 0.5691640496485718,
+      "grad_norm": 1.6356233741539357,
+      "learning_rate": 4.127784340017653e-06,
+      "loss": 0.367,
+      "step": 3806
+    },
+    {
+      "epoch": 0.5693135935397039,
+      "grad_norm": 1.6628010853628497,
+      "learning_rate": 4.125399749703526e-06,
+      "loss": 0.3327,
+      "step": 3807
+    },
+    {
+      "epoch": 0.569463137430836,
+      "grad_norm": 1.7948970149563248,
+      "learning_rate": 4.123015364579046e-06,
+      "loss": 0.2519,
+      "step": 3808
+    },
+    {
+      "epoch": 0.569612681321968,
+      "grad_norm": 1.5509060057609694,
+      "learning_rate": 4.120631185203613e-06,
+      "loss": 0.3288,
+      "step": 3809
+    },
+    {
+      "epoch": 0.5697622252131,
+      "grad_norm": 1.9889960830789197,
+      "learning_rate": 4.118247212136579e-06,
+      "loss": 0.2358,
+      "step": 3810
+    },
+    {
+      "epoch": 0.5699117691042321,
+      "grad_norm": 1.4594387383602159,
+      "learning_rate": 4.11586344593725e-06,
+      "loss": 0.2483,
+      "step": 3811
+    },
+    {
+      "epoch": 0.5700613129953641,
+      "grad_norm": 1.505789986482447,
+      "learning_rate": 4.113479887164873e-06,
+      "loss": 0.1721,
+      "step": 3812
+    },
+    {
+      "epoch": 0.5702108568864962,
+      "grad_norm": 1.5260414366564614,
+      "learning_rate": 4.111096536378661e-06,
+      "loss": 0.3631,
+      "step": 3813
+    },
+    {
+      "epoch": 0.5703604007776283,
+      "grad_norm": 1.1589677612121907,
+      "learning_rate": 4.108713394137766e-06,
+      "loss": 0.1784,
+      "step": 3814
+    },
+    {
+      "epoch": 0.5705099446687603,
+      "grad_norm": 1.8625931258889772,
+      "learning_rate": 4.106330461001299e-06,
+      "loss": 0.2186,
+      "step": 3815
+    },
+    {
+      "epoch": 0.5706594885598923,
+      "grad_norm": 1.8225390074850283,
+      "learning_rate": 4.103947737528321e-06,
+      "loss": 0.2044,
+      "step": 3816
+    },
+    {
+      "epoch": 0.5708090324510243,
+      "grad_norm": 1.2260264786541455,
+      "learning_rate": 4.101565224277837e-06,
+      "loss": 0.2178,
+      "step": 3817
+    },
+    {
+      "epoch": 0.5709585763421564,
+      "grad_norm": 2.25246042259685,
+      "learning_rate": 4.0991829218088104e-06,
+      "loss": 0.1861,
+      "step": 3818
+    },
+    {
+      "epoch": 0.5711081202332885,
+      "grad_norm": 1.4584755271889351,
+      "learning_rate": 4.096800830680151e-06,
+      "loss": 0.2435,
+      "step": 3819
+    },
+    {
+      "epoch": 0.5712576641244205,
+      "grad_norm": 1.3471748072584406,
+      "learning_rate": 4.094418951450721e-06,
+      "loss": 0.2657,
+      "step": 3820
+    },
+    {
+      "epoch": 0.5714072080155526,
+      "grad_norm": 1.4912123602040317,
+      "learning_rate": 4.092037284679335e-06,
+      "loss": 0.1924,
+      "step": 3821
+    },
+    {
+      "epoch": 0.5715567519066846,
+      "grad_norm": 2.3296405416523798,
+      "learning_rate": 4.08965583092475e-06,
+      "loss": 0.3366,
+      "step": 3822
+    },
+    {
+      "epoch": 0.5717062957978166,
+      "grad_norm": 1.2984835174705707,
+      "learning_rate": 4.087274590745681e-06,
+      "loss": 0.1936,
+      "step": 3823
+    },
+    {
+      "epoch": 0.5718558396889487,
+      "grad_norm": 1.824810422234852,
+      "learning_rate": 4.084893564700789e-06,
+      "loss": 0.2123,
+      "step": 3824
+    },
+    {
+      "epoch": 0.5720053835800808,
+      "grad_norm": 1.5552749721927202,
+      "learning_rate": 4.0825127533486865e-06,
+      "loss": 0.21,
+      "step": 3825
+    },
+    {
+      "epoch": 0.5721549274712128,
+      "grad_norm": 2.3462920390606934,
+      "learning_rate": 4.080132157247935e-06,
+      "loss": 0.5941,
+      "step": 3826
+    },
+    {
+      "epoch": 0.5723044713623449,
+      "grad_norm": 1.406630107053223,
+      "learning_rate": 4.077751776957043e-06,
+      "loss": 0.1833,
+      "step": 3827
+    },
+    {
+      "epoch": 0.5724540152534769,
+      "grad_norm": 1.6381181183648261,
+      "learning_rate": 4.075371613034472e-06,
+      "loss": 0.1843,
+      "step": 3828
+    },
+    {
+      "epoch": 0.5726035591446089,
+      "grad_norm": 1.7160296586520287,
+      "learning_rate": 4.072991666038633e-06,
+      "loss": 0.3517,
+      "step": 3829
+    },
+    {
+      "epoch": 0.572753103035741,
+      "grad_norm": 1.625254929701419,
+      "learning_rate": 4.070611936527882e-06,
+      "loss": 0.3434,
+      "step": 3830
+    },
+    {
+      "epoch": 0.572902646926873,
+      "grad_norm": 1.1799818297023703,
+      "learning_rate": 4.068232425060527e-06,
+      "loss": 0.1565,
+      "step": 3831
+    },
+    {
+      "epoch": 0.5730521908180051,
+      "grad_norm": 2.2431824234579283,
+      "learning_rate": 4.065853132194827e-06,
+      "loss": 0.2081,
+      "step": 3832
+    },
+    {
+      "epoch": 0.5732017347091372,
+      "grad_norm": 1.3731533378251932,
+      "learning_rate": 4.0634740584889835e-06,
+      "loss": 0.1461,
+      "step": 3833
+    },
+    {
+      "epoch": 0.5733512786002691,
+      "grad_norm": 1.8919618152249174,
+      "learning_rate": 4.061095204501151e-06,
+      "loss": 0.4743,
+      "step": 3834
+    },
+    {
+      "epoch": 0.5735008224914012,
+      "grad_norm": 1.5440897864945764,
+      "learning_rate": 4.0587165707894326e-06,
+      "loss": 0.2421,
+      "step": 3835
+    },
+    {
+      "epoch": 0.5736503663825333,
+      "grad_norm": 2.102210076263177,
+      "learning_rate": 4.056338157911877e-06,
+      "loss": 0.216,
+      "step": 3836
+    },
+    {
+      "epoch": 0.5737999102736653,
+      "grad_norm": 1.480929733210923,
+      "learning_rate": 4.053959966426488e-06,
+      "loss": 0.3114,
+      "step": 3837
+    },
+    {
+      "epoch": 0.5739494541647974,
+      "grad_norm": 1.675352548569913,
+      "learning_rate": 4.0515819968912055e-06,
+      "loss": 0.3535,
+      "step": 3838
+    },
+    {
+      "epoch": 0.5740989980559295,
+      "grad_norm": 1.2596206818226765,
+      "learning_rate": 4.049204249863926e-06,
+      "loss": 0.1904,
+      "step": 3839
+    },
+    {
+      "epoch": 0.5742485419470614,
+      "grad_norm": 2.5280885615982314,
+      "learning_rate": 4.0468267259024935e-06,
+      "loss": 0.4747,
+      "step": 3840
+    },
+    {
+      "epoch": 0.5743980858381935,
+      "grad_norm": 1.4594935661035582,
+      "learning_rate": 4.044449425564696e-06,
+      "loss": 0.3341,
+      "step": 3841
+    },
+    {
+      "epoch": 0.5745476297293256,
+      "grad_norm": 1.7090098221110461,
+      "learning_rate": 4.042072349408275e-06,
+      "loss": 0.3826,
+      "step": 3842
+    },
+    {
+      "epoch": 0.5746971736204576,
+      "grad_norm": 2.0106514775480755,
+      "learning_rate": 4.0396954979909105e-06,
+      "loss": 0.2264,
+      "step": 3843
+    },
+    {
+      "epoch": 0.5748467175115897,
+      "grad_norm": 1.9951662121993903,
+      "learning_rate": 4.037318871870238e-06,
+      "loss": 0.2587,
+      "step": 3844
+    },
+    {
+      "epoch": 0.5749962614027218,
+      "grad_norm": 2.3735689067814882,
+      "learning_rate": 4.034942471603835e-06,
+      "loss": 0.6212,
+      "step": 3845
+    },
+    {
+      "epoch": 0.5751458052938537,
+      "grad_norm": 1.3313805957100329,
+      "learning_rate": 4.032566297749229e-06,
+      "loss": 0.1866,
+      "step": 3846
+    },
+    {
+      "epoch": 0.5752953491849858,
+      "grad_norm": 1.7289313985492631,
+      "learning_rate": 4.030190350863894e-06,
+      "loss": 0.3472,
+      "step": 3847
+    },
+    {
+      "epoch": 0.5754448930761178,
+      "grad_norm": 1.493639708680967,
+      "learning_rate": 4.027814631505248e-06,
+      "loss": 0.1904,
+      "step": 3848
+    },
+    {
+      "epoch": 0.5755944369672499,
+      "grad_norm": 1.822043958390137,
+      "learning_rate": 4.0254391402306595e-06,
+      "loss": 0.3634,
+      "step": 3849
+    },
+    {
+      "epoch": 0.575743980858382,
+      "grad_norm": 1.713070502237203,
+      "learning_rate": 4.023063877597439e-06,
+      "loss": 0.4581,
+      "step": 3850
+    },
+    {
+      "epoch": 0.5758935247495139,
+      "grad_norm": 1.33095048873002,
+      "learning_rate": 4.02068884416285e-06,
+      "loss": 0.2208,
+      "step": 3851
+    },
+    {
+      "epoch": 0.576043068640646,
+      "grad_norm": 1.7910675740169966,
+      "learning_rate": 4.018314040484096e-06,
+      "loss": 0.2775,
+      "step": 3852
+    },
+    {
+      "epoch": 0.5761926125317781,
+      "grad_norm": 1.9328597024089476,
+      "learning_rate": 4.015939467118328e-06,
+      "loss": 0.1634,
+      "step": 3853
+    },
+    {
+      "epoch": 0.5763421564229101,
+      "grad_norm": 2.0376615827356233,
+      "learning_rate": 4.0135651246226456e-06,
+      "loss": 0.2157,
+      "step": 3854
+    },
+    {
+      "epoch": 0.5764917003140422,
+      "grad_norm": 1.680601957815068,
+      "learning_rate": 4.011191013554091e-06,
+      "loss": 0.3631,
+      "step": 3855
+    },
+    {
+      "epoch": 0.5766412442051743,
+      "grad_norm": 2.1132908114007214,
+      "learning_rate": 4.0088171344696545e-06,
+      "loss": 0.4247,
+      "step": 3856
+    },
+    {
+      "epoch": 0.5767907880963062,
+      "grad_norm": 1.5935575986177055,
+      "learning_rate": 4.006443487926272e-06,
+      "loss": 0.2774,
+      "step": 3857
+    },
+    {
+      "epoch": 0.5769403319874383,
+      "grad_norm": 9.754017882344208,
+      "learning_rate": 4.004070074480821e-06,
+      "loss": 0.2705,
+      "step": 3858
+    },
+    {
+      "epoch": 0.5770898758785704,
+      "grad_norm": 1.55564169140264,
+      "learning_rate": 4.001696894690128e-06,
+      "loss": 0.3352,
+      "step": 3859
+    },
+    {
+      "epoch": 0.5772394197697024,
+      "grad_norm": 1.798973114929995,
+      "learning_rate": 3.999323949110965e-06,
+      "loss": 0.3104,
+      "step": 3860
+    },
+    {
+      "epoch": 0.5773889636608345,
+      "grad_norm": 1.2999830259764698,
+      "learning_rate": 3.9969512383000465e-06,
+      "loss": 0.1917,
+      "step": 3861
+    },
+    {
+      "epoch": 0.5775385075519665,
+      "grad_norm": 1.6701086999443087,
+      "learning_rate": 3.994578762814037e-06,
+      "loss": 0.2341,
+      "step": 3862
+    },
+    {
+      "epoch": 0.5776880514430985,
+      "grad_norm": 2.007518390533276,
+      "learning_rate": 3.992206523209538e-06,
+      "loss": 0.4862,
+      "step": 3863
+    },
+    {
+      "epoch": 0.5778375953342306,
+      "grad_norm": 1.6758189519451339,
+      "learning_rate": 3.989834520043101e-06,
+      "loss": 0.3451,
+      "step": 3864
+    },
+    {
+      "epoch": 0.5779871392253626,
+      "grad_norm": 1.4063121728928503,
+      "learning_rate": 3.98746275387122e-06,
+      "loss": 0.1555,
+      "step": 3865
+    },
+    {
+      "epoch": 0.5781366831164947,
+      "grad_norm": 1.4879389391741649,
+      "learning_rate": 3.985091225250336e-06,
+      "loss": 0.1553,
+      "step": 3866
+    },
+    {
+      "epoch": 0.5782862270076268,
+      "grad_norm": 4.347154093050415,
+      "learning_rate": 3.982719934736832e-06,
+      "loss": 0.4053,
+      "step": 3867
+    },
+    {
+      "epoch": 0.5784357708987587,
+      "grad_norm": 1.5443034001101248,
+      "learning_rate": 3.980348882887035e-06,
+      "loss": 0.3198,
+      "step": 3868
+    },
+    {
+      "epoch": 0.5785853147898908,
+      "grad_norm": 1.7950442773636421,
+      "learning_rate": 3.977978070257216e-06,
+      "loss": 0.4694,
+      "step": 3869
+    },
+    {
+      "epoch": 0.5787348586810229,
+      "grad_norm": 2.1973684295340363,
+      "learning_rate": 3.975607497403592e-06,
+      "loss": 0.4131,
+      "step": 3870
+    },
+    {
+      "epoch": 0.5788844025721549,
+      "grad_norm": 1.7267711780843302,
+      "learning_rate": 3.973237164882321e-06,
+      "loss": 0.2991,
+      "step": 3871
+    },
+    {
+      "epoch": 0.579033946463287,
+      "grad_norm": 1.2991893137030361,
+      "learning_rate": 3.970867073249508e-06,
+      "loss": 0.2453,
+      "step": 3872
+    },
+    {
+      "epoch": 0.5791834903544191,
+      "grad_norm": 1.5602478676658988,
+      "learning_rate": 3.9684972230611956e-06,
+      "loss": 0.2093,
+      "step": 3873
+    },
+    {
+      "epoch": 0.579333034245551,
+      "grad_norm": 1.6934982295673262,
+      "learning_rate": 3.966127614873376e-06,
+      "loss": 0.2215,
+      "step": 3874
+    },
+    {
+      "epoch": 0.5794825781366831,
+      "grad_norm": 2.110410563267741,
+      "learning_rate": 3.963758249241983e-06,
+      "loss": 0.5277,
+      "step": 3875
+    },
+    {
+      "epoch": 0.5796321220278151,
+      "grad_norm": 1.8247166919937388,
+      "learning_rate": 3.96138912672289e-06,
+      "loss": 0.3023,
+      "step": 3876
+    },
+    {
+      "epoch": 0.5797816659189472,
+      "grad_norm": 1.3368960784081465,
+      "learning_rate": 3.9590202478719205e-06,
+      "loss": 0.2167,
+      "step": 3877
+    },
+    {
+      "epoch": 0.5799312098100793,
+      "grad_norm": 1.5584675563310413,
+      "learning_rate": 3.95665161324483e-06,
+      "loss": 0.3788,
+      "step": 3878
+    },
+    {
+      "epoch": 0.5800807537012113,
+      "grad_norm": 1.8874134439396686,
+      "learning_rate": 3.954283223397327e-06,
+      "loss": 0.2273,
+      "step": 3879
+    },
+    {
+      "epoch": 0.5802302975923433,
+      "grad_norm": 1.6256776139655844,
+      "learning_rate": 3.951915078885057e-06,
+      "loss": 0.3477,
+      "step": 3880
+    },
+    {
+      "epoch": 0.5803798414834754,
+      "grad_norm": 2.1424342396636025,
+      "learning_rate": 3.94954718026361e-06,
+      "loss": 0.2369,
+      "step": 3881
+    },
+    {
+      "epoch": 0.5805293853746074,
+      "grad_norm": 1.5525493295636965,
+      "learning_rate": 3.94717952808852e-06,
+      "loss": 0.2494,
+      "step": 3882
+    },
+    {
+      "epoch": 0.5806789292657395,
+      "grad_norm": 1.2525644074657651,
+      "learning_rate": 3.9448121229152585e-06,
+      "loss": 0.3307,
+      "step": 3883
+    },
+    {
+      "epoch": 0.5808284731568716,
+      "grad_norm": 1.3489399531915276,
+      "learning_rate": 3.942444965299242e-06,
+      "loss": 0.3343,
+      "step": 3884
+    },
+    {
+      "epoch": 0.5809780170480036,
+      "grad_norm": 1.6971348754950633,
+      "learning_rate": 3.9400780557958276e-06,
+      "loss": 0.358,
+      "step": 3885
+    },
+    {
+      "epoch": 0.5811275609391356,
+      "grad_norm": 1.0598261581438866,
+      "learning_rate": 3.937711394960318e-06,
+      "loss": 0.2013,
+      "step": 3886
+    },
+    {
+      "epoch": 0.5812771048302677,
+      "grad_norm": 1.7799188090141735,
+      "learning_rate": 3.9353449833479515e-06,
+      "loss": 0.3384,
+      "step": 3887
+    },
+    {
+      "epoch": 0.5814266487213997,
+      "grad_norm": 2.17623015999994,
+      "learning_rate": 3.932978821513913e-06,
+      "loss": 0.628,
+      "step": 3888
+    },
+    {
+      "epoch": 0.5815761926125318,
+      "grad_norm": 1.6370721585688572,
+      "learning_rate": 3.930612910013326e-06,
+      "loss": 0.4118,
+      "step": 3889
+    },
+    {
+      "epoch": 0.5817257365036638,
+      "grad_norm": 1.9972246926476178,
+      "learning_rate": 3.928247249401256e-06,
+      "loss": 0.3762,
+      "step": 3890
+    },
+    {
+      "epoch": 0.5818752803947959,
+      "grad_norm": 1.568923848614493,
+      "learning_rate": 3.925881840232711e-06,
+      "loss": 0.19,
+      "step": 3891
+    },
+    {
+      "epoch": 0.5820248242859279,
+      "grad_norm": 1.989967171024415,
+      "learning_rate": 3.9235166830626375e-06,
+      "loss": 0.6291,
+      "step": 3892
+    },
+    {
+      "epoch": 0.5821743681770599,
+      "grad_norm": 1.6320609023822952,
+      "learning_rate": 3.921151778445925e-06,
+      "loss": 0.2308,
+      "step": 3893
+    },
+    {
+      "epoch": 0.582323912068192,
+      "grad_norm": 1.6089921792274722,
+      "learning_rate": 3.918787126937401e-06,
+      "loss": 0.3225,
+      "step": 3894
+    },
+    {
+      "epoch": 0.5824734559593241,
+      "grad_norm": 1.7543181167298438,
+      "learning_rate": 3.916422729091838e-06,
+      "loss": 0.2871,
+      "step": 3895
+    },
+    {
+      "epoch": 0.5826229998504561,
+      "grad_norm": 1.7660217430179217,
+      "learning_rate": 3.914058585463946e-06,
+      "loss": 0.3934,
+      "step": 3896
+    },
+    {
+      "epoch": 0.5827725437415882,
+      "grad_norm": 1.6926823971057354,
+      "learning_rate": 3.911694696608377e-06,
+      "loss": 0.2323,
+      "step": 3897
+    },
+    {
+      "epoch": 0.5829220876327202,
+      "grad_norm": 1.428661667605896,
+      "learning_rate": 3.909331063079719e-06,
+      "loss": 0.2164,
+      "step": 3898
+    },
+    {
+      "epoch": 0.5830716315238522,
+      "grad_norm": 2.2116367730485087,
+      "learning_rate": 3.906967685432504e-06,
+      "loss": 0.6146,
+      "step": 3899
+    },
+    {
+      "epoch": 0.5832211754149843,
+      "grad_norm": 1.4672573042525427,
+      "learning_rate": 3.904604564221204e-06,
+      "loss": 0.359,
+      "step": 3900
+    },
+    {
+      "epoch": 0.5833707193061164,
+      "grad_norm": 1.4136745331637248,
+      "learning_rate": 3.90224170000023e-06,
+      "loss": 0.3082,
+      "step": 3901
+    },
+    {
+      "epoch": 0.5835202631972484,
+      "grad_norm": 1.5604421702220446,
+      "learning_rate": 3.899879093323935e-06,
+      "loss": 0.3606,
+      "step": 3902
+    },
+    {
+      "epoch": 0.5836698070883805,
+      "grad_norm": 1.4826709528011215,
+      "learning_rate": 3.897516744746605e-06,
+      "loss": 0.3643,
+      "step": 3903
+    },
+    {
+      "epoch": 0.5838193509795125,
+      "grad_norm": 1.436442769039814,
+      "learning_rate": 3.895154654822471e-06,
+      "loss": 0.3506,
+      "step": 3904
+    },
+    {
+      "epoch": 0.5839688948706445,
+      "grad_norm": 1.007210202220155,
+      "learning_rate": 3.892792824105703e-06,
+      "loss": 0.1583,
+      "step": 3905
+    },
+    {
+      "epoch": 0.5841184387617766,
+      "grad_norm": 1.6057856448875152,
+      "learning_rate": 3.89043125315041e-06,
+      "loss": 0.367,
+      "step": 3906
+    },
+    {
+      "epoch": 0.5842679826529086,
+      "grad_norm": 1.1873872540306754,
+      "learning_rate": 3.888069942510639e-06,
+      "loss": 0.1812,
+      "step": 3907
+    },
+    {
+      "epoch": 0.5844175265440407,
+      "grad_norm": 1.531148138356171,
+      "learning_rate": 3.885708892740374e-06,
+      "loss": 0.2584,
+      "step": 3908
+    },
+    {
+      "epoch": 0.5845670704351728,
+      "grad_norm": 1.53531282902828,
+      "learning_rate": 3.883348104393542e-06,
+      "loss": 0.2089,
+      "step": 3909
+    },
+    {
+      "epoch": 0.5847166143263047,
+      "grad_norm": 1.2571052242037701,
+      "learning_rate": 3.880987578024006e-06,
+      "loss": 0.1618,
+      "step": 3910
+    },
+    {
+      "epoch": 0.5848661582174368,
+      "grad_norm": 1.685780465923781,
+      "learning_rate": 3.878627314185569e-06,
+      "loss": 0.4313,
+      "step": 3911
+    },
+    {
+      "epoch": 0.5850157021085689,
+      "grad_norm": 1.7418918273189634,
+      "learning_rate": 3.876267313431973e-06,
+      "loss": 0.475,
+      "step": 3912
+    },
+    {
+      "epoch": 0.5851652459997009,
+      "grad_norm": 1.734807587708549,
+      "learning_rate": 3.873907576316892e-06,
+      "loss": 0.4154,
+      "step": 3913
+    },
+    {
+      "epoch": 0.585314789890833,
+      "grad_norm": 1.6406377655622573,
+      "learning_rate": 3.871548103393947e-06,
+      "loss": 0.3568,
+      "step": 3914
+    },
+    {
+      "epoch": 0.585464333781965,
+      "grad_norm": 1.1517941732296657,
+      "learning_rate": 3.869188895216692e-06,
+      "loss": 0.1842,
+      "step": 3915
+    },
+    {
+      "epoch": 0.585613877673097,
+      "grad_norm": 1.4377802648166382,
+      "learning_rate": 3.866829952338621e-06,
+      "loss": 0.3479,
+      "step": 3916
+    },
+    {
+      "epoch": 0.5857634215642291,
+      "grad_norm": 1.114349518595328,
+      "learning_rate": 3.864471275313164e-06,
+      "loss": 0.1237,
+      "step": 3917
+    },
+    {
+      "epoch": 0.5859129654553612,
+      "grad_norm": 1.4127129455946879,
+      "learning_rate": 3.862112864693691e-06,
+      "loss": 0.1984,
+      "step": 3918
+    },
+    {
+      "epoch": 0.5860625093464932,
+      "grad_norm": 1.4417924544505127,
+      "learning_rate": 3.859754721033504e-06,
+      "loss": 0.3395,
+      "step": 3919
+    },
+    {
+      "epoch": 0.5862120532376253,
+      "grad_norm": 1.7189752041055817,
+      "learning_rate": 3.857396844885849e-06,
+      "loss": 0.4097,
+      "step": 3920
+    },
+    {
+      "epoch": 0.5863615971287572,
+      "grad_norm": 1.3073724826843582,
+      "learning_rate": 3.8550392368039045e-06,
+      "loss": 0.2553,
+      "step": 3921
+    },
+    {
+      "epoch": 0.5865111410198893,
+      "grad_norm": 1.4564439495009185,
+      "learning_rate": 3.85268189734079e-06,
+      "loss": 0.2909,
+      "step": 3922
+    },
+    {
+      "epoch": 0.5866606849110214,
+      "grad_norm": 1.5841823741056198,
+      "learning_rate": 3.850324827049561e-06,
+      "loss": 0.2065,
+      "step": 3923
+    },
+    {
+      "epoch": 0.5868102288021534,
+      "grad_norm": 1.0916215110890355,
+      "learning_rate": 3.847968026483206e-06,
+      "loss": 0.1892,
+      "step": 3924
+    },
+    {
+      "epoch": 0.5869597726932855,
+      "grad_norm": 1.9627347229493133,
+      "learning_rate": 3.845611496194655e-06,
+      "loss": 0.2363,
+      "step": 3925
+    },
+    {
+      "epoch": 0.5871093165844176,
+      "grad_norm": 1.4671089891749798,
+      "learning_rate": 3.843255236736773e-06,
+      "loss": 0.1938,
+      "step": 3926
+    },
+    {
+      "epoch": 0.5872588604755495,
+      "grad_norm": 1.5337050885765506,
+      "learning_rate": 3.840899248662358e-06,
+      "loss": 0.3077,
+      "step": 3927
+    },
+    {
+      "epoch": 0.5874084043666816,
+      "grad_norm": 1.1406387988000504,
+      "learning_rate": 3.838543532524153e-06,
+      "loss": 0.1899,
+      "step": 3928
+    },
+    {
+      "epoch": 0.5875579482578137,
+      "grad_norm": 1.1589973791618806,
+      "learning_rate": 3.836188088874827e-06,
+      "loss": 0.1882,
+      "step": 3929
+    },
+    {
+      "epoch": 0.5877074921489457,
+      "grad_norm": 1.8799012061866471,
+      "learning_rate": 3.833832918266992e-06,
+      "loss": 0.2755,
+      "step": 3930
+    },
+    {
+      "epoch": 0.5878570360400778,
+      "grad_norm": 1.8985733996127807,
+      "learning_rate": 3.831478021253192e-06,
+      "loss": 0.5369,
+      "step": 3931
+    },
+    {
+      "epoch": 0.5880065799312099,
+      "grad_norm": 1.1384325773858892,
+      "learning_rate": 3.829123398385911e-06,
+      "loss": 0.1673,
+      "step": 3932
+    },
+    {
+      "epoch": 0.5881561238223418,
+      "grad_norm": 1.2281350484830058,
+      "learning_rate": 3.826769050217564e-06,
+      "loss": 0.1814,
+      "step": 3933
+    },
+    {
+      "epoch": 0.5883056677134739,
+      "grad_norm": 1.6894736624996396,
+      "learning_rate": 3.824414977300506e-06,
+      "loss": 0.3602,
+      "step": 3934
+    },
+    {
+      "epoch": 0.5884552116046059,
+      "grad_norm": 1.4936989174970754,
+      "learning_rate": 3.822061180187022e-06,
+      "loss": 0.3309,
+      "step": 3935
+    },
+    {
+      "epoch": 0.588604755495738,
+      "grad_norm": 1.2978536320218477,
+      "learning_rate": 3.819707659429339e-06,
+      "loss": 0.2046,
+      "step": 3936
+    },
+    {
+      "epoch": 0.5887542993868701,
+      "grad_norm": 1.8284004362578155,
+      "learning_rate": 3.817354415579612e-06,
+      "loss": 0.5456,
+      "step": 3937
+    },
+    {
+      "epoch": 0.588903843278002,
+      "grad_norm": 1.7662708358791046,
+      "learning_rate": 3.815001449189941e-06,
+      "loss": 0.3615,
+      "step": 3938
+    },
+    {
+      "epoch": 0.5890533871691341,
+      "grad_norm": 1.2321195352474157,
+      "learning_rate": 3.812648760812347e-06,
+      "loss": 0.1731,
+      "step": 3939
+    },
+    {
+      "epoch": 0.5892029310602662,
+      "grad_norm": 1.6990341437020333,
+      "learning_rate": 3.8102963509987965e-06,
+      "loss": 0.4168,
+      "step": 3940
+    },
+    {
+      "epoch": 0.5893524749513982,
+      "grad_norm": 1.6729207645316984,
+      "learning_rate": 3.807944220301188e-06,
+      "loss": 0.4699,
+      "step": 3941
+    },
+    {
+      "epoch": 0.5895020188425303,
+      "grad_norm": 1.2458633939386743,
+      "learning_rate": 3.8055923692713527e-06,
+      "loss": 0.3579,
+      "step": 3942
+    },
+    {
+      "epoch": 0.5896515627336624,
+      "grad_norm": 1.043880257216526,
+      "learning_rate": 3.803240798461061e-06,
+      "loss": 0.1603,
+      "step": 3943
+    },
+    {
+      "epoch": 0.5898011066247943,
+      "grad_norm": 0.9729214387193956,
+      "learning_rate": 3.80088950842201e-06,
+      "loss": 0.2006,
+      "step": 3944
+    },
+    {
+      "epoch": 0.5899506505159264,
+      "grad_norm": 1.6414730241820703,
+      "learning_rate": 3.7985384997058365e-06,
+      "loss": 0.2049,
+      "step": 3945
+    },
+    {
+      "epoch": 0.5901001944070585,
+      "grad_norm": 1.6657086768253946,
+      "learning_rate": 3.79618777286411e-06,
+      "loss": 0.2827,
+      "step": 3946
+    },
+    {
+      "epoch": 0.5902497382981905,
+      "grad_norm": 1.6956367490229918,
+      "learning_rate": 3.7938373284483336e-06,
+      "loss": 0.4608,
+      "step": 3947
+    },
+    {
+      "epoch": 0.5903992821893226,
+      "grad_norm": 1.4757058595390375,
+      "learning_rate": 3.791487167009945e-06,
+      "loss": 0.3557,
+      "step": 3948
+    },
+    {
+      "epoch": 0.5905488260804546,
+      "grad_norm": 1.5729666823262434,
+      "learning_rate": 3.789137289100313e-06,
+      "loss": 0.184,
+      "step": 3949
+    },
+    {
+      "epoch": 0.5906983699715866,
+      "grad_norm": 1.633027833675113,
+      "learning_rate": 3.786787695270743e-06,
+      "loss": 0.4198,
+      "step": 3950
+    },
+    {
+      "epoch": 0.5908479138627187,
+      "grad_norm": 1.438486968065632,
+      "learning_rate": 3.784438386072471e-06,
+      "loss": 0.2484,
+      "step": 3951
+    },
+    {
+      "epoch": 0.5909974577538507,
+      "grad_norm": 1.7582663564747014,
+      "learning_rate": 3.782089362056669e-06,
+      "loss": 0.2605,
+      "step": 3952
+    },
+    {
+      "epoch": 0.5911470016449828,
+      "grad_norm": 1.8491750366304591,
+      "learning_rate": 3.7797406237744406e-06,
+      "loss": 0.5609,
+      "step": 3953
+    },
+    {
+      "epoch": 0.5912965455361149,
+      "grad_norm": 2.1940854570062256,
+      "learning_rate": 3.7773921717768212e-06,
+      "loss": 0.3145,
+      "step": 3954
+    },
+    {
+      "epoch": 0.5914460894272469,
+      "grad_norm": 1.758778183907136,
+      "learning_rate": 3.7750440066147816e-06,
+      "loss": 0.4234,
+      "step": 3955
+    },
+    {
+      "epoch": 0.5915956333183789,
+      "grad_norm": 1.0902355030071746,
+      "learning_rate": 3.7726961288392227e-06,
+      "loss": 0.158,
+      "step": 3956
+    },
+    {
+      "epoch": 0.591745177209511,
+      "grad_norm": 1.3875658617841062,
+      "learning_rate": 3.7703485390009797e-06,
+      "loss": 0.2265,
+      "step": 3957
+    },
+    {
+      "epoch": 0.591894721100643,
+      "grad_norm": 1.6204448366336377,
+      "learning_rate": 3.7680012376508227e-06,
+      "loss": 0.3166,
+      "step": 3958
+    },
+    {
+      "epoch": 0.5920442649917751,
+      "grad_norm": 1.3645834401060262,
+      "learning_rate": 3.765654225339447e-06,
+      "loss": 0.4263,
+      "step": 3959
+    },
+    {
+      "epoch": 0.5921938088829072,
+      "grad_norm": 1.3231211927212307,
+      "learning_rate": 3.763307502617485e-06,
+      "loss": 0.1809,
+      "step": 3960
+    },
+    {
+      "epoch": 0.5923433527740392,
+      "grad_norm": 1.247672175483129,
+      "learning_rate": 3.7609610700355014e-06,
+      "loss": 0.1774,
+      "step": 3961
+    },
+    {
+      "epoch": 0.5924928966651712,
+      "grad_norm": 1.1720146064252939,
+      "learning_rate": 3.7586149281439908e-06,
+      "loss": 0.1641,
+      "step": 3962
+    },
+    {
+      "epoch": 0.5926424405563033,
+      "grad_norm": 1.5718680027072247,
+      "learning_rate": 3.756269077493385e-06,
+      "loss": 0.3326,
+      "step": 3963
+    },
+    {
+      "epoch": 0.5927919844474353,
+      "grad_norm": 1.7772158209903743,
+      "learning_rate": 3.7539235186340378e-06,
+      "loss": 0.2532,
+      "step": 3964
+    },
+    {
+      "epoch": 0.5929415283385674,
+      "grad_norm": 1.3316474232901556,
+      "learning_rate": 3.751578252116242e-06,
+      "loss": 0.2147,
+      "step": 3965
+    },
+    {
+      "epoch": 0.5930910722296994,
+      "grad_norm": 1.4220331937170747,
+      "learning_rate": 3.7492332784902197e-06,
+      "loss": 0.1649,
+      "step": 3966
+    },
+    {
+      "epoch": 0.5932406161208315,
+      "grad_norm": 1.4819153735880548,
+      "learning_rate": 3.7468885983061245e-06,
+      "loss": 0.2828,
+      "step": 3967
+    },
+    {
+      "epoch": 0.5933901600119635,
+      "grad_norm": 1.533582987774297,
+      "learning_rate": 3.7445442121140417e-06,
+      "loss": 0.3549,
+      "step": 3968
+    },
+    {
+      "epoch": 0.5935397039030955,
+      "grad_norm": 2.0084478805318846,
+      "learning_rate": 3.7422001204639855e-06,
+      "loss": 0.5592,
+      "step": 3969
+    },
+    {
+      "epoch": 0.5936892477942276,
+      "grad_norm": 1.4582643481241435,
+      "learning_rate": 3.7398563239059026e-06,
+      "loss": 0.3025,
+      "step": 3970
+    },
+    {
+      "epoch": 0.5938387916853597,
+      "grad_norm": 1.4895919525012717,
+      "learning_rate": 3.7375128229896707e-06,
+      "loss": 0.1886,
+      "step": 3971
+    },
+    {
+      "epoch": 0.5939883355764917,
+      "grad_norm": 1.551469791731259,
+      "learning_rate": 3.735169618265097e-06,
+      "loss": 0.2539,
+      "step": 3972
+    },
+    {
+      "epoch": 0.5941378794676238,
+      "grad_norm": 1.3051779904330787,
+      "learning_rate": 3.732826710281923e-06,
+      "loss": 0.3184,
+      "step": 3973
+    },
+    {
+      "epoch": 0.5942874233587558,
+      "grad_norm": 1.2392389402763688,
+      "learning_rate": 3.7304840995898116e-06,
+      "loss": 0.1694,
+      "step": 3974
+    },
+    {
+      "epoch": 0.5944369672498878,
+      "grad_norm": 1.374809357583305,
+      "learning_rate": 3.7281417867383663e-06,
+      "loss": 0.2051,
+      "step": 3975
+    },
+    {
+      "epoch": 0.5945865111410199,
+      "grad_norm": 1.6294109216768207,
+      "learning_rate": 3.7257997722771157e-06,
+      "loss": 0.3044,
+      "step": 3976
+    },
+    {
+      "epoch": 0.594736055032152,
+      "grad_norm": 1.3224369428308862,
+      "learning_rate": 3.7234580567555177e-06,
+      "loss": 0.2726,
+      "step": 3977
+    },
+    {
+      "epoch": 0.594885598923284,
+      "grad_norm": 1.7787635958789316,
+      "learning_rate": 3.7211166407229647e-06,
+      "loss": 0.2811,
+      "step": 3978
+    },
+    {
+      "epoch": 0.595035142814416,
+      "grad_norm": 1.4233934814396993,
+      "learning_rate": 3.718775524728771e-06,
+      "loss": 0.2117,
+      "step": 3979
+    },
+    {
+      "epoch": 0.595184686705548,
+      "grad_norm": 1.329758067841591,
+      "learning_rate": 3.7164347093221865e-06,
+      "loss": 0.1894,
+      "step": 3980
+    },
+    {
+      "epoch": 0.5953342305966801,
+      "grad_norm": 1.8256996890692794,
+      "learning_rate": 3.714094195052389e-06,
+      "loss": 0.5129,
+      "step": 3981
+    },
+    {
+      "epoch": 0.5954837744878122,
+      "grad_norm": 1.1205085695878227,
+      "learning_rate": 3.7117539824684866e-06,
+      "loss": 0.1646,
+      "step": 3982
+    },
+    {
+      "epoch": 0.5956333183789442,
+      "grad_norm": 1.7874489300108056,
+      "learning_rate": 3.709414072119516e-06,
+      "loss": 0.2973,
+      "step": 3983
+    },
+    {
+      "epoch": 0.5957828622700763,
+      "grad_norm": 1.1884738103980468,
+      "learning_rate": 3.707074464554441e-06,
+      "loss": 0.2175,
+      "step": 3984
+    },
+    {
+      "epoch": 0.5959324061612084,
+      "grad_norm": 1.623711079409265,
+      "learning_rate": 3.704735160322157e-06,
+      "loss": 0.3532,
+      "step": 3985
+    },
+    {
+      "epoch": 0.5960819500523403,
+      "grad_norm": 1.4825541420935746,
+      "learning_rate": 3.7023961599714876e-06,
+      "loss": 0.3893,
+      "step": 3986
+    },
+    {
+      "epoch": 0.5962314939434724,
+      "grad_norm": 1.6573795029984477,
+      "learning_rate": 3.7000574640511843e-06,
+      "loss": 0.349,
+      "step": 3987
+    },
+    {
+      "epoch": 0.5963810378346045,
+      "grad_norm": 1.5325568908598715,
+      "learning_rate": 3.697719073109929e-06,
+      "loss": 0.217,
+      "step": 3988
+    },
+    {
+      "epoch": 0.5965305817257365,
+      "grad_norm": 1.1985716673788906,
+      "learning_rate": 3.695380987696329e-06,
+      "loss": 0.208,
+      "step": 3989
+    },
+    {
+      "epoch": 0.5966801256168686,
+      "grad_norm": 1.511940247427596,
+      "learning_rate": 3.693043208358922e-06,
+      "loss": 0.3116,
+      "step": 3990
+    },
+    {
+      "epoch": 0.5968296695080006,
+      "grad_norm": 1.382101797672102,
+      "learning_rate": 3.6907057356461733e-06,
+      "loss": 0.4606,
+      "step": 3991
+    },
+    {
+      "epoch": 0.5969792133991326,
+      "grad_norm": 1.2466801874804314,
+      "learning_rate": 3.6883685701064774e-06,
+      "loss": 0.2164,
+      "step": 3992
+    },
+    {
+      "epoch": 0.5971287572902647,
+      "grad_norm": 1.2514060754927105,
+      "learning_rate": 3.6860317122881583e-06,
+      "loss": 0.2169,
+      "step": 3993
+    },
+    {
+      "epoch": 0.5972783011813967,
+      "grad_norm": 1.633454160231319,
+      "learning_rate": 3.6836951627394598e-06,
+      "loss": 0.2053,
+      "step": 3994
+    },
+    {
+      "epoch": 0.5974278450725288,
+      "grad_norm": 1.5000653948039326,
+      "learning_rate": 3.681358922008562e-06,
+      "loss": 0.3564,
+      "step": 3995
+    },
+    {
+      "epoch": 0.5975773889636609,
+      "grad_norm": 1.7042876877013595,
+      "learning_rate": 3.6790229906435706e-06,
+      "loss": 0.2217,
+      "step": 3996
+    },
+    {
+      "epoch": 0.5977269328547928,
+      "grad_norm": 1.1660551674685846,
+      "learning_rate": 3.6766873691925155e-06,
+      "loss": 0.2042,
+      "step": 3997
+    },
+    {
+      "epoch": 0.5978764767459249,
+      "grad_norm": 1.0016866110140052,
+      "learning_rate": 3.674352058203359e-06,
+      "loss": 0.1465,
+      "step": 3998
+    },
+    {
+      "epoch": 0.598026020637057,
+      "grad_norm": 1.358478565718614,
+      "learning_rate": 3.672017058223982e-06,
+      "loss": 0.234,
+      "step": 3999
+    },
+    {
+      "epoch": 0.598175564528189,
+      "grad_norm": 1.5541836099323216,
+      "learning_rate": 3.6696823698022034e-06,
+      "loss": 0.3531,
+      "step": 4000
+    },
+    {
+      "epoch": 0.5983251084193211,
+      "grad_norm": 1.4965638649664035,
+      "learning_rate": 3.6673479934857602e-06,
+      "loss": 0.1734,
+      "step": 4001
+    },
+    {
+      "epoch": 0.5984746523104532,
+      "grad_norm": 1.2798780853635923,
+      "learning_rate": 3.665013929822322e-06,
+      "loss": 0.2227,
+      "step": 4002
+    },
+    {
+      "epoch": 0.5986241962015851,
+      "grad_norm": 1.5631428969516747,
+      "learning_rate": 3.6626801793594802e-06,
+      "loss": 0.3341,
+      "step": 4003
+    },
+    {
+      "epoch": 0.5987737400927172,
+      "grad_norm": 1.1547611258266666,
+      "learning_rate": 3.6603467426447596e-06,
+      "loss": 0.2265,
+      "step": 4004
+    },
+    {
+      "epoch": 0.5989232839838493,
+      "grad_norm": 1.7661237138941523,
+      "learning_rate": 3.6580136202256024e-06,
+      "loss": 0.2831,
+      "step": 4005
+    },
+    {
+      "epoch": 0.5990728278749813,
+      "grad_norm": 1.608871850044881,
+      "learning_rate": 3.655680812649382e-06,
+      "loss": 0.1863,
+      "step": 4006
+    },
+    {
+      "epoch": 0.5992223717661134,
+      "grad_norm": 1.2120032911627765,
+      "learning_rate": 3.653348320463399e-06,
+      "loss": 0.2105,
+      "step": 4007
+    },
+    {
+      "epoch": 0.5993719156572453,
+      "grad_norm": 1.7713505485144023,
+      "learning_rate": 3.6510161442148783e-06,
+      "loss": 0.4625,
+      "step": 4008
+    },
+    {
+      "epoch": 0.5995214595483774,
+      "grad_norm": 1.5406088685048114,
+      "learning_rate": 3.6486842844509727e-06,
+      "loss": 0.2302,
+      "step": 4009
+    },
+    {
+      "epoch": 0.5996710034395095,
+      "grad_norm": 1.7566777699356078,
+      "learning_rate": 3.646352741718755e-06,
+      "loss": 0.4958,
+      "step": 4010
+    },
+    {
+      "epoch": 0.5998205473306415,
+      "grad_norm": 1.773701442285048,
+      "learning_rate": 3.6440215165652305e-06,
+      "loss": 0.1975,
+      "step": 4011
+    },
+    {
+      "epoch": 0.5999700912217736,
+      "grad_norm": 1.3009283223219021,
+      "learning_rate": 3.641690609537325e-06,
+      "loss": 0.292,
+      "step": 4012
+    },
+    {
+      "epoch": 0.6001196351129057,
+      "grad_norm": 1.4931178313458866,
+      "learning_rate": 3.639360021181894e-06,
+      "loss": 0.3095,
+      "step": 4013
+    },
+    {
+      "epoch": 0.6002691790040376,
+      "grad_norm": 1.5793921746413464,
+      "learning_rate": 3.637029752045716e-06,
+      "loss": 0.339,
+      "step": 4014
+    },
+    {
+      "epoch": 0.6004187228951697,
+      "grad_norm": 2.1241976260965463,
+      "learning_rate": 3.6346998026754927e-06,
+      "loss": 0.3532,
+      "step": 4015
+    },
+    {
+      "epoch": 0.6005682667863018,
+      "grad_norm": 1.0579231943140326,
+      "learning_rate": 3.6323701736178528e-06,
+      "loss": 0.1443,
+      "step": 4016
+    },
+    {
+      "epoch": 0.6007178106774338,
+      "grad_norm": 1.6519105898111532,
+      "learning_rate": 3.6300408654193515e-06,
+      "loss": 0.3343,
+      "step": 4017
+    },
+    {
+      "epoch": 0.6008673545685659,
+      "grad_norm": 1.1836835886912143,
+      "learning_rate": 3.6277118786264653e-06,
+      "loss": 0.2084,
+      "step": 4018
+    },
+    {
+      "epoch": 0.601016898459698,
+      "grad_norm": 1.6124984883515716,
+      "learning_rate": 3.6253832137856e-06,
+      "loss": 0.3631,
+      "step": 4019
+    },
+    {
+      "epoch": 0.6011664423508299,
+      "grad_norm": 1.8682133876991003,
+      "learning_rate": 3.623054871443078e-06,
+      "loss": 0.4665,
+      "step": 4020
+    },
+    {
+      "epoch": 0.601315986241962,
+      "grad_norm": 2.062835003276884,
+      "learning_rate": 3.620726852145154e-06,
+      "loss": 0.2746,
+      "step": 4021
+    },
+    {
+      "epoch": 0.6014655301330941,
+      "grad_norm": 1.583633628610332,
+      "learning_rate": 3.6183991564380023e-06,
+      "loss": 0.3516,
+      "step": 4022
+    },
+    {
+      "epoch": 0.6016150740242261,
+      "grad_norm": 1.8495282333605947,
+      "learning_rate": 3.616071784867723e-06,
+      "loss": 0.2359,
+      "step": 4023
+    },
+    {
+      "epoch": 0.6017646179153582,
+      "grad_norm": 1.577516249529156,
+      "learning_rate": 3.613744737980343e-06,
+      "loss": 0.2088,
+      "step": 4024
+    },
+    {
+      "epoch": 0.6019141618064902,
+      "grad_norm": 1.550353493297501,
+      "learning_rate": 3.6114180163218056e-06,
+      "loss": 0.1928,
+      "step": 4025
+    },
+    {
+      "epoch": 0.6020637056976222,
+      "grad_norm": 1.1352226203791365,
+      "learning_rate": 3.6090916204379834e-06,
+      "loss": 0.1816,
+      "step": 4026
+    },
+    {
+      "epoch": 0.6022132495887543,
+      "grad_norm": 1.5030785772526614,
+      "learning_rate": 3.6067655508746726e-06,
+      "loss": 0.2309,
+      "step": 4027
+    },
+    {
+      "epoch": 0.6023627934798863,
+      "grad_norm": 1.319464371625546,
+      "learning_rate": 3.60443980817759e-06,
+      "loss": 0.1861,
+      "step": 4028
+    },
+    {
+      "epoch": 0.6025123373710184,
+      "grad_norm": 1.3852086589255572,
+      "learning_rate": 3.6021143928923783e-06,
+      "loss": 0.1944,
+      "step": 4029
+    },
+    {
+      "epoch": 0.6026618812621505,
+      "grad_norm": 2.243854952724501,
+      "learning_rate": 3.599789305564602e-06,
+      "loss": 0.51,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6028114251532825,
+      "grad_norm": 1.419391997587051,
+      "learning_rate": 3.597464546739749e-06,
+      "loss": 0.3178,
+      "step": 4031
+    },
+    {
+      "epoch": 0.6029609690444145,
+      "grad_norm": 1.2894199829326205,
+      "learning_rate": 3.5951401169632293e-06,
+      "loss": 0.1947,
+      "step": 4032
+    },
+    {
+      "epoch": 0.6031105129355466,
+      "grad_norm": 1.530880304394482,
+      "learning_rate": 3.5928160167803784e-06,
+      "loss": 0.1948,
+      "step": 4033
+    },
+    {
+      "epoch": 0.6032600568266786,
+      "grad_norm": 1.1347517974588832,
+      "learning_rate": 3.5904922467364526e-06,
+      "loss": 0.2253,
+      "step": 4034
+    },
+    {
+      "epoch": 0.6034096007178107,
+      "grad_norm": 1.232257517693222,
+      "learning_rate": 3.5881688073766267e-06,
+      "loss": 0.205,
+      "step": 4035
+    },
+    {
+      "epoch": 0.6035591446089428,
+      "grad_norm": 1.6963988945541442,
+      "learning_rate": 3.5858456992460066e-06,
+      "loss": 0.2107,
+      "step": 4036
+    },
+    {
+      "epoch": 0.6037086885000748,
+      "grad_norm": 1.1849360298492893,
+      "learning_rate": 3.5835229228896133e-06,
+      "loss": 0.1807,
+      "step": 4037
+    },
+    {
+      "epoch": 0.6038582323912068,
+      "grad_norm": 1.8006880398870282,
+      "learning_rate": 3.5812004788523933e-06,
+      "loss": 0.3419,
+      "step": 4038
+    },
+    {
+      "epoch": 0.6040077762823388,
+      "grad_norm": 1.1218152100667766,
+      "learning_rate": 3.5788783676792176e-06,
+      "loss": 0.1963,
+      "step": 4039
+    },
+    {
+      "epoch": 0.6041573201734709,
+      "grad_norm": 1.4412889174387291,
+      "learning_rate": 3.5765565899148703e-06,
+      "loss": 0.1818,
+      "step": 4040
+    },
+    {
+      "epoch": 0.604306864064603,
+      "grad_norm": 1.6406744631622459,
+      "learning_rate": 3.5742351461040646e-06,
+      "loss": 0.1984,
+      "step": 4041
+    },
+    {
+      "epoch": 0.604456407955735,
+      "grad_norm": 1.7230960345125375,
+      "learning_rate": 3.571914036791435e-06,
+      "loss": 0.235,
+      "step": 4042
+    },
+    {
+      "epoch": 0.604605951846867,
+      "grad_norm": 1.4708145309645435,
+      "learning_rate": 3.5695932625215347e-06,
+      "loss": 0.254,
+      "step": 4043
+    },
+    {
+      "epoch": 0.6047554957379991,
+      "grad_norm": 1.8016752353343,
+      "learning_rate": 3.567272823838842e-06,
+      "loss": 0.2005,
+      "step": 4044
+    },
+    {
+      "epoch": 0.6049050396291311,
+      "grad_norm": 1.1493815123325322,
+      "learning_rate": 3.564952721287751e-06,
+      "loss": 0.1713,
+      "step": 4045
+    },
+    {
+      "epoch": 0.6050545835202632,
+      "grad_norm": 1.5119225059128631,
+      "learning_rate": 3.5626329554125823e-06,
+      "loss": 0.182,
+      "step": 4046
+    },
+    {
+      "epoch": 0.6052041274113953,
+      "grad_norm": 1.8759224212157393,
+      "learning_rate": 3.5603135267575737e-06,
+      "loss": 0.5205,
+      "step": 4047
+    },
+    {
+      "epoch": 0.6053536713025273,
+      "grad_norm": 1.565295827071973,
+      "learning_rate": 3.5579944358668873e-06,
+      "loss": 0.3772,
+      "step": 4048
+    },
+    {
+      "epoch": 0.6055032151936593,
+      "grad_norm": 2.122632657433492,
+      "learning_rate": 3.555675683284604e-06,
+      "loss": 0.4978,
+      "step": 4049
+    },
+    {
+      "epoch": 0.6056527590847914,
+      "grad_norm": 1.9409099996151258,
+      "learning_rate": 3.553357269554724e-06,
+      "loss": 0.3612,
+      "step": 4050
+    },
+    {
+      "epoch": 0.6058023029759234,
+      "grad_norm": 1.8573438786477485,
+      "learning_rate": 3.551039195221171e-06,
+      "loss": 0.2926,
+      "step": 4051
+    },
+    {
+      "epoch": 0.6059518468670555,
+      "grad_norm": 1.6830272916472488,
+      "learning_rate": 3.5487214608277857e-06,
+      "loss": 0.344,
+      "step": 4052
+    },
+    {
+      "epoch": 0.6061013907581875,
+      "grad_norm": 2.544342408333431,
+      "learning_rate": 3.546404066918333e-06,
+      "loss": 0.242,
+      "step": 4053
+    },
+    {
+      "epoch": 0.6062509346493196,
+      "grad_norm": 1.8351824486146582,
+      "learning_rate": 3.5440870140364967e-06,
+      "loss": 0.3456,
+      "step": 4054
+    },
+    {
+      "epoch": 0.6064004785404516,
+      "grad_norm": 1.9733766052564323,
+      "learning_rate": 3.5417703027258752e-06,
+      "loss": 0.2553,
+      "step": 4055
+    },
+    {
+      "epoch": 0.6065500224315836,
+      "grad_norm": 1.4266711606260554,
+      "learning_rate": 3.539453933529996e-06,
+      "loss": 0.3237,
+      "step": 4056
+    },
+    {
+      "epoch": 0.6066995663227157,
+      "grad_norm": 1.6718168939315017,
+      "learning_rate": 3.5371379069922983e-06,
+      "loss": 0.3281,
+      "step": 4057
+    },
+    {
+      "epoch": 0.6068491102138478,
+      "grad_norm": 2.082935916824144,
+      "learning_rate": 3.5348222236561467e-06,
+      "loss": 0.5821,
+      "step": 4058
+    },
+    {
+      "epoch": 0.6069986541049798,
+      "grad_norm": 1.5696132013810573,
+      "learning_rate": 3.5325068840648243e-06,
+      "loss": 0.2177,
+      "step": 4059
+    },
+    {
+      "epoch": 0.6071481979961119,
+      "grad_norm": 1.4087839185461357,
+      "learning_rate": 3.530191888761527e-06,
+      "loss": 0.3296,
+      "step": 4060
+    },
+    {
+      "epoch": 0.607297741887244,
+      "grad_norm": 2.099550365239757,
+      "learning_rate": 3.5278772382893777e-06,
+      "loss": 0.3616,
+      "step": 4061
+    },
+    {
+      "epoch": 0.6074472857783759,
+      "grad_norm": 1.6842257514961676,
+      "learning_rate": 3.5255629331914153e-06,
+      "loss": 0.331,
+      "step": 4062
+    },
+    {
+      "epoch": 0.607596829669508,
+      "grad_norm": 1.8703972469815064,
+      "learning_rate": 3.523248974010599e-06,
+      "loss": 0.1758,
+      "step": 4063
+    },
+    {
+      "epoch": 0.6077463735606401,
+      "grad_norm": 1.524647589128137,
+      "learning_rate": 3.5209353612898067e-06,
+      "loss": 0.3307,
+      "step": 4064
+    },
+    {
+      "epoch": 0.6078959174517721,
+      "grad_norm": 1.3067348714233367,
+      "learning_rate": 3.518622095571831e-06,
+      "loss": 0.1462,
+      "step": 4065
+    },
+    {
+      "epoch": 0.6080454613429042,
+      "grad_norm": 1.7759379098365817,
+      "learning_rate": 3.516309177399389e-06,
+      "loss": 0.4481,
+      "step": 4066
+    },
+    {
+      "epoch": 0.6081950052340361,
+      "grad_norm": 1.5543708215165737,
+      "learning_rate": 3.513996607315112e-06,
+      "loss": 0.2023,
+      "step": 4067
+    },
+    {
+      "epoch": 0.6083445491251682,
+      "grad_norm": 1.911842884824045,
+      "learning_rate": 3.5116843858615534e-06,
+      "loss": 0.4787,
+      "step": 4068
+    },
+    {
+      "epoch": 0.6084940930163003,
+      "grad_norm": 1.9298854332543098,
+      "learning_rate": 3.509372513581182e-06,
+      "loss": 0.3521,
+      "step": 4069
+    },
+    {
+      "epoch": 0.6086436369074323,
+      "grad_norm": 1.5428623983694816,
+      "learning_rate": 3.5070609910163824e-06,
+      "loss": 0.3474,
+      "step": 4070
+    },
+    {
+      "epoch": 0.6087931807985644,
+      "grad_norm": 0.9977909289978434,
+      "learning_rate": 3.504749818709463e-06,
+      "loss": 0.174,
+      "step": 4071
+    },
+    {
+      "epoch": 0.6089427246896965,
+      "grad_norm": 1.1421293763115168,
+      "learning_rate": 3.5024389972026467e-06,
+      "loss": 0.2034,
+      "step": 4072
+    },
+    {
+      "epoch": 0.6090922685808284,
+      "grad_norm": 1.3644097038150713,
+      "learning_rate": 3.500128527038074e-06,
+      "loss": 0.1842,
+      "step": 4073
+    },
+    {
+      "epoch": 0.6092418124719605,
+      "grad_norm": 1.332440548321116,
+      "learning_rate": 3.497818408757806e-06,
+      "loss": 0.2728,
+      "step": 4074
+    },
+    {
+      "epoch": 0.6093913563630926,
+      "grad_norm": 1.9608087269420822,
+      "learning_rate": 3.495508642903813e-06,
+      "loss": 0.2712,
+      "step": 4075
+    },
+    {
+      "epoch": 0.6095409002542246,
+      "grad_norm": 1.1647404978024332,
+      "learning_rate": 3.493199230017993e-06,
+      "loss": 0.1906,
+      "step": 4076
+    },
+    {
+      "epoch": 0.6096904441453567,
+      "grad_norm": 1.7451274126518581,
+      "learning_rate": 3.4908901706421548e-06,
+      "loss": 0.5245,
+      "step": 4077
+    },
+    {
+      "epoch": 0.6098399880364888,
+      "grad_norm": 1.4276907218803567,
+      "learning_rate": 3.488581465318026e-06,
+      "loss": 0.2377,
+      "step": 4078
+    },
+    {
+      "epoch": 0.6099895319276207,
+      "grad_norm": 1.7169738066445221,
+      "learning_rate": 3.4862731145872548e-06,
+      "loss": 0.3399,
+      "step": 4079
+    },
+    {
+      "epoch": 0.6101390758187528,
+      "grad_norm": 1.688093650559266,
+      "learning_rate": 3.4839651189913965e-06,
+      "loss": 0.387,
+      "step": 4080
+    },
+    {
+      "epoch": 0.6102886197098849,
+      "grad_norm": 1.9535517951044004,
+      "learning_rate": 3.4816574790719322e-06,
+      "loss": 0.4879,
+      "step": 4081
+    },
+    {
+      "epoch": 0.6104381636010169,
+      "grad_norm": 1.6770987277973084,
+      "learning_rate": 3.479350195370256e-06,
+      "loss": 0.2773,
+      "step": 4082
+    },
+    {
+      "epoch": 0.610587707492149,
+      "grad_norm": 1.5959756774847607,
+      "learning_rate": 3.477043268427679e-06,
+      "loss": 0.2258,
+      "step": 4083
+    },
+    {
+      "epoch": 0.6107372513832809,
+      "grad_norm": 1.3440527422443669,
+      "learning_rate": 3.4747366987854294e-06,
+      "loss": 0.1886,
+      "step": 4084
+    },
+    {
+      "epoch": 0.610886795274413,
+      "grad_norm": 1.1452652380873023,
+      "learning_rate": 3.472430486984648e-06,
+      "loss": 0.1815,
+      "step": 4085
+    },
+    {
+      "epoch": 0.6110363391655451,
+      "grad_norm": 1.8862381396037677,
+      "learning_rate": 3.4701246335663973e-06,
+      "loss": 0.28,
+      "step": 4086
+    },
+    {
+      "epoch": 0.6111858830566771,
+      "grad_norm": 1.3462017669113007,
+      "learning_rate": 3.467819139071651e-06,
+      "loss": 0.2976,
+      "step": 4087
+    },
+    {
+      "epoch": 0.6113354269478092,
+      "grad_norm": 0.9872082098092432,
+      "learning_rate": 3.465514004041301e-06,
+      "loss": 0.1924,
+      "step": 4088
+    },
+    {
+      "epoch": 0.6114849708389413,
+      "grad_norm": 1.258797674393771,
+      "learning_rate": 3.4632092290161547e-06,
+      "loss": 0.139,
+      "step": 4089
+    },
+    {
+      "epoch": 0.6116345147300732,
+      "grad_norm": 1.5951256051278744,
+      "learning_rate": 3.460904814536934e-06,
+      "loss": 0.2237,
+      "step": 4090
+    },
+    {
+      "epoch": 0.6117840586212053,
+      "grad_norm": 1.3552720149728599,
+      "learning_rate": 3.458600761144276e-06,
+      "loss": 0.205,
+      "step": 4091
+    },
+    {
+      "epoch": 0.6119336025123374,
+      "grad_norm": 1.2855312617700432,
+      "learning_rate": 3.456297069378734e-06,
+      "loss": 0.2011,
+      "step": 4092
+    },
+    {
+      "epoch": 0.6120831464034694,
+      "grad_norm": 1.131526768203718,
+      "learning_rate": 3.4539937397807765e-06,
+      "loss": 0.179,
+      "step": 4093
+    },
+    {
+      "epoch": 0.6122326902946015,
+      "grad_norm": 1.012638619251334,
+      "learning_rate": 3.4516907728907876e-06,
+      "loss": 0.1881,
+      "step": 4094
+    },
+    {
+      "epoch": 0.6123822341857336,
+      "grad_norm": 2.3057997137329433,
+      "learning_rate": 3.4493881692490676e-06,
+      "loss": 0.6228,
+      "step": 4095
+    },
+    {
+      "epoch": 0.6125317780768655,
+      "grad_norm": 1.6590161930605178,
+      "learning_rate": 3.4470859293958237e-06,
+      "loss": 0.2417,
+      "step": 4096
+    },
+    {
+      "epoch": 0.6126813219679976,
+      "grad_norm": 1.8637272203103565,
+      "learning_rate": 3.444784053871187e-06,
+      "loss": 0.276,
+      "step": 4097
+    },
+    {
+      "epoch": 0.6128308658591296,
+      "grad_norm": 1.7540355711160616,
+      "learning_rate": 3.4424825432151997e-06,
+      "loss": 0.3656,
+      "step": 4098
+    },
+    {
+      "epoch": 0.6129804097502617,
+      "grad_norm": 1.2210296377287981,
+      "learning_rate": 3.4401813979678184e-06,
+      "loss": 0.1902,
+      "step": 4099
+    },
+    {
+      "epoch": 0.6131299536413938,
+      "grad_norm": 1.3676262505744117,
+      "learning_rate": 3.437880618668916e-06,
+      "loss": 0.1401,
+      "step": 4100
+    },
+    {
+      "epoch": 0.6132794975325258,
+      "grad_norm": 1.7617762190417448,
+      "learning_rate": 3.435580205858272e-06,
+      "loss": 0.3606,
+      "step": 4101
+    },
+    {
+      "epoch": 0.6134290414236578,
+      "grad_norm": 1.1968515809658866,
+      "learning_rate": 3.4332801600755895e-06,
+      "loss": 0.1833,
+      "step": 4102
+    },
+    {
+      "epoch": 0.6135785853147899,
+      "grad_norm": 1.560412557500599,
+      "learning_rate": 3.430980481860481e-06,
+      "loss": 0.1793,
+      "step": 4103
+    },
+    {
+      "epoch": 0.6137281292059219,
+      "grad_norm": 2.0021038501103727,
+      "learning_rate": 3.4286811717524713e-06,
+      "loss": 0.3963,
+      "step": 4104
+    },
+    {
+      "epoch": 0.613877673097054,
+      "grad_norm": 1.0542430630039104,
+      "learning_rate": 3.4263822302910046e-06,
+      "loss": 0.1729,
+      "step": 4105
+    },
+    {
+      "epoch": 0.6140272169881861,
+      "grad_norm": 1.5902196547544478,
+      "learning_rate": 3.42408365801543e-06,
+      "loss": 0.2775,
+      "step": 4106
+    },
+    {
+      "epoch": 0.614176760879318,
+      "grad_norm": 1.3721418830423344,
+      "learning_rate": 3.421785455465017e-06,
+      "loss": 0.2971,
+      "step": 4107
+    },
+    {
+      "epoch": 0.6143263047704501,
+      "grad_norm": 1.729094375014569,
+      "learning_rate": 3.4194876231789464e-06,
+      "loss": 0.3713,
+      "step": 4108
+    },
+    {
+      "epoch": 0.6144758486615822,
+      "grad_norm": 2.4117639518236174,
+      "learning_rate": 3.417190161696311e-06,
+      "loss": 0.3593,
+      "step": 4109
+    },
+    {
+      "epoch": 0.6146253925527142,
+      "grad_norm": 2.086422185272039,
+      "learning_rate": 3.414893071556119e-06,
+      "loss": 0.3511,
+      "step": 4110
+    },
+    {
+      "epoch": 0.6147749364438463,
+      "grad_norm": 1.5735352312513073,
+      "learning_rate": 3.4125963532972878e-06,
+      "loss": 0.2897,
+      "step": 4111
+    },
+    {
+      "epoch": 0.6149244803349783,
+      "grad_norm": 1.8232983507476865,
+      "learning_rate": 3.41030000745865e-06,
+      "loss": 0.2121,
+      "step": 4112
+    },
+    {
+      "epoch": 0.6150740242261103,
+      "grad_norm": 1.542088897552113,
+      "learning_rate": 3.4080040345789515e-06,
+      "loss": 0.2548,
+      "step": 4113
+    },
+    {
+      "epoch": 0.6152235681172424,
+      "grad_norm": 2.055704862139885,
+      "learning_rate": 3.4057084351968497e-06,
+      "loss": 0.2074,
+      "step": 4114
+    },
+    {
+      "epoch": 0.6153731120083744,
+      "grad_norm": 1.5475598177649545,
+      "learning_rate": 3.4034132098509143e-06,
+      "loss": 0.3923,
+      "step": 4115
+    },
+    {
+      "epoch": 0.6155226558995065,
+      "grad_norm": 1.4512060477133912,
+      "learning_rate": 3.401118359079625e-06,
+      "loss": 0.1771,
+      "step": 4116
+    },
+    {
+      "epoch": 0.6156721997906386,
+      "grad_norm": 1.3040117809635243,
+      "learning_rate": 3.3988238834213785e-06,
+      "loss": 0.3118,
+      "step": 4117
+    },
+    {
+      "epoch": 0.6158217436817706,
+      "grad_norm": 1.193831767033035,
+      "learning_rate": 3.3965297834144796e-06,
+      "loss": 0.1846,
+      "step": 4118
+    },
+    {
+      "epoch": 0.6159712875729026,
+      "grad_norm": 1.2936870504351123,
+      "learning_rate": 3.394236059597147e-06,
+      "loss": 0.3206,
+      "step": 4119
+    },
+    {
+      "epoch": 0.6161208314640347,
+      "grad_norm": 1.9358809574481608,
+      "learning_rate": 3.3919427125075117e-06,
+      "loss": 0.4591,
+      "step": 4120
+    },
+    {
+      "epoch": 0.6162703753551667,
+      "grad_norm": 1.5626339711928905,
+      "learning_rate": 3.389649742683612e-06,
+      "loss": 0.3604,
+      "step": 4121
+    },
+    {
+      "epoch": 0.6164199192462988,
+      "grad_norm": 1.8684296406735175,
+      "learning_rate": 3.387357150663402e-06,
+      "loss": 0.3298,
+      "step": 4122
+    },
+    {
+      "epoch": 0.6165694631374309,
+      "grad_norm": 1.3357412406024458,
+      "learning_rate": 3.3850649369847455e-06,
+      "loss": 0.1642,
+      "step": 4123
+    },
+    {
+      "epoch": 0.6167190070285629,
+      "grad_norm": 2.19771279087013,
+      "learning_rate": 3.3827731021854194e-06,
+      "loss": 0.2888,
+      "step": 4124
+    },
+    {
+      "epoch": 0.616868550919695,
+      "grad_norm": 1.548714911422848,
+      "learning_rate": 3.38048164680311e-06,
+      "loss": 0.1871,
+      "step": 4125
+    },
+    {
+      "epoch": 0.6170180948108269,
+      "grad_norm": 1.159508819902876,
+      "learning_rate": 3.3781905713754134e-06,
+      "loss": 0.1897,
+      "step": 4126
+    },
+    {
+      "epoch": 0.617167638701959,
+      "grad_norm": 1.5115811682102942,
+      "learning_rate": 3.375899876439838e-06,
+      "loss": 0.3759,
+      "step": 4127
+    },
+    {
+      "epoch": 0.6173171825930911,
+      "grad_norm": 1.790849642617947,
+      "learning_rate": 3.3736095625338043e-06,
+      "loss": 0.4749,
+      "step": 4128
+    },
+    {
+      "epoch": 0.6174667264842231,
+      "grad_norm": 1.905917740616368,
+      "learning_rate": 3.3713196301946415e-06,
+      "loss": 0.285,
+      "step": 4129
+    },
+    {
+      "epoch": 0.6176162703753552,
+      "grad_norm": 2.218642923968202,
+      "learning_rate": 3.36903007995959e-06,
+      "loss": 0.52,
+      "step": 4130
+    },
+    {
+      "epoch": 0.6177658142664872,
+      "grad_norm": 2.304237580429699,
+      "learning_rate": 3.366740912365799e-06,
+      "loss": 0.5589,
+      "step": 4131
+    },
+    {
+      "epoch": 0.6179153581576192,
+      "grad_norm": 1.790679696297901,
+      "learning_rate": 3.3644521279503305e-06,
+      "loss": 0.3308,
+      "step": 4132
+    },
+    {
+      "epoch": 0.6180649020487513,
+      "grad_norm": 1.3187679561865975,
+      "learning_rate": 3.3621637272501555e-06,
+      "loss": 0.2061,
+      "step": 4133
+    },
+    {
+      "epoch": 0.6182144459398834,
+      "grad_norm": 1.8718764354464674,
+      "learning_rate": 3.3598757108021546e-06,
+      "loss": 0.2996,
+      "step": 4134
+    },
+    {
+      "epoch": 0.6183639898310154,
+      "grad_norm": 1.784502907359159,
+      "learning_rate": 3.3575880791431205e-06,
+      "loss": 0.3875,
+      "step": 4135
+    },
+    {
+      "epoch": 0.6185135337221475,
+      "grad_norm": 2.0257075474723276,
+      "learning_rate": 3.3553008328097473e-06,
+      "loss": 0.3315,
+      "step": 4136
+    },
+    {
+      "epoch": 0.6186630776132795,
+      "grad_norm": 1.7363579300166394,
+      "learning_rate": 3.353013972338651e-06,
+      "loss": 0.1923,
+      "step": 4137
+    },
+    {
+      "epoch": 0.6188126215044115,
+      "grad_norm": 1.645989344743434,
+      "learning_rate": 3.3507274982663495e-06,
+      "loss": 0.404,
+      "step": 4138
+    },
+    {
+      "epoch": 0.6189621653955436,
+      "grad_norm": 1.6006600046854644,
+      "learning_rate": 3.3484414111292716e-06,
+      "loss": 0.2298,
+      "step": 4139
+    },
+    {
+      "epoch": 0.6191117092866757,
+      "grad_norm": 1.7303842670320941,
+      "learning_rate": 3.3461557114637575e-06,
+      "loss": 0.1871,
+      "step": 4140
+    },
+    {
+      "epoch": 0.6192612531778077,
+      "grad_norm": 1.2683245690225082,
+      "learning_rate": 3.3438703998060503e-06,
+      "loss": 0.1994,
+      "step": 4141
+    },
+    {
+      "epoch": 0.6194107970689398,
+      "grad_norm": 1.6461137510586357,
+      "learning_rate": 3.3415854766923073e-06,
+      "loss": 0.1957,
+      "step": 4142
+    },
+    {
+      "epoch": 0.6195603409600717,
+      "grad_norm": 1.581644802224599,
+      "learning_rate": 3.339300942658595e-06,
+      "loss": 0.2005,
+      "step": 4143
+    },
+    {
+      "epoch": 0.6197098848512038,
+      "grad_norm": 1.821750932071071,
+      "learning_rate": 3.3370167982408857e-06,
+      "loss": 0.404,
+      "step": 4144
+    },
+    {
+      "epoch": 0.6198594287423359,
+      "grad_norm": 1.010226375154919,
+      "learning_rate": 3.3347330439750634e-06,
+      "loss": 0.1895,
+      "step": 4145
+    },
+    {
+      "epoch": 0.6200089726334679,
+      "grad_norm": 1.7213955718892866,
+      "learning_rate": 3.332449680396917e-06,
+      "loss": 0.3164,
+      "step": 4146
+    },
+    {
+      "epoch": 0.6201585165246,
+      "grad_norm": 1.4718996679664604,
+      "learning_rate": 3.330166708042146e-06,
+      "loss": 0.3313,
+      "step": 4147
+    },
+    {
+      "epoch": 0.620308060415732,
+      "grad_norm": 1.1698035624548326,
+      "learning_rate": 3.3278841274463585e-06,
+      "loss": 0.1969,
+      "step": 4148
+    },
+    {
+      "epoch": 0.620457604306864,
+      "grad_norm": 1.5316613724929764,
+      "learning_rate": 3.3256019391450696e-06,
+      "loss": 0.3458,
+      "step": 4149
+    },
+    {
+      "epoch": 0.6206071481979961,
+      "grad_norm": 1.67069813493684,
+      "learning_rate": 3.3233201436737032e-06,
+      "loss": 0.2107,
+      "step": 4150
+    },
+    {
+      "epoch": 0.6207566920891282,
+      "grad_norm": 1.8523375864622236,
+      "learning_rate": 3.3210387415675894e-06,
+      "loss": 0.5053,
+      "step": 4151
+    },
+    {
+      "epoch": 0.6209062359802602,
+      "grad_norm": 1.268125401219981,
+      "learning_rate": 3.318757733361967e-06,
+      "loss": 0.178,
+      "step": 4152
+    },
+    {
+      "epoch": 0.6210557798713923,
+      "grad_norm": 1.2720176937747762,
+      "learning_rate": 3.3164771195919833e-06,
+      "loss": 0.2731,
+      "step": 4153
+    },
+    {
+      "epoch": 0.6212053237625244,
+      "grad_norm": 1.4082416391595145,
+      "learning_rate": 3.3141969007926917e-06,
+      "loss": 0.1942,
+      "step": 4154
+    },
+    {
+      "epoch": 0.6213548676536563,
+      "grad_norm": 0.8960930795176492,
+      "learning_rate": 3.311917077499056e-06,
+      "loss": 0.2027,
+      "step": 4155
+    },
+    {
+      "epoch": 0.6215044115447884,
+      "grad_norm": 1.5829596350349728,
+      "learning_rate": 3.309637650245941e-06,
+      "loss": 0.1931,
+      "step": 4156
+    },
+    {
+      "epoch": 0.6216539554359204,
+      "grad_norm": 1.0024243338014172,
+      "learning_rate": 3.307358619568123e-06,
+      "loss": 0.1921,
+      "step": 4157
+    },
+    {
+      "epoch": 0.6218034993270525,
+      "grad_norm": 1.4575451361421503,
+      "learning_rate": 3.305079986000286e-06,
+      "loss": 0.196,
+      "step": 4158
+    },
+    {
+      "epoch": 0.6219530432181846,
+      "grad_norm": 1.5847817071066232,
+      "learning_rate": 3.3028017500770188e-06,
+      "loss": 0.2149,
+      "step": 4159
+    },
+    {
+      "epoch": 0.6221025871093165,
+      "grad_norm": 0.9644862006701003,
+      "learning_rate": 3.30052391233282e-06,
+      "loss": 0.1878,
+      "step": 4160
+    },
+    {
+      "epoch": 0.6222521310004486,
+      "grad_norm": 1.8148496541499093,
+      "learning_rate": 3.298246473302087e-06,
+      "loss": 0.3232,
+      "step": 4161
+    },
+    {
+      "epoch": 0.6224016748915807,
+      "grad_norm": 1.8222373927506255,
+      "learning_rate": 3.2959694335191328e-06,
+      "loss": 0.3742,
+      "step": 4162
+    },
+    {
+      "epoch": 0.6225512187827127,
+      "grad_norm": 1.5806770368482173,
+      "learning_rate": 3.293692793518171e-06,
+      "loss": 0.5245,
+      "step": 4163
+    },
+    {
+      "epoch": 0.6227007626738448,
+      "grad_norm": 1.9535273557118913,
+      "learning_rate": 3.2914165538333247e-06,
+      "loss": 0.6479,
+      "step": 4164
+    },
+    {
+      "epoch": 0.6228503065649769,
+      "grad_norm": 1.2749728150835529,
+      "learning_rate": 3.2891407149986223e-06,
+      "loss": 0.1906,
+      "step": 4165
+    },
+    {
+      "epoch": 0.6229998504561088,
+      "grad_norm": 2.1068364656942262,
+      "learning_rate": 3.2868652775479947e-06,
+      "loss": 0.563,
+      "step": 4166
+    },
+    {
+      "epoch": 0.6231493943472409,
+      "grad_norm": 1.6644405150545742,
+      "learning_rate": 3.2845902420152833e-06,
+      "loss": 0.242,
+      "step": 4167
+    },
+    {
+      "epoch": 0.623298938238373,
+      "grad_norm": 1.421746094855021,
+      "learning_rate": 3.2823156089342335e-06,
+      "loss": 0.2305,
+      "step": 4168
+    },
+    {
+      "epoch": 0.623448482129505,
+      "grad_norm": 1.7648579864130391,
+      "learning_rate": 3.2800413788384956e-06,
+      "loss": 0.4275,
+      "step": 4169
+    },
+    {
+      "epoch": 0.6235980260206371,
+      "grad_norm": 1.5957084104714234,
+      "learning_rate": 3.277767552261626e-06,
+      "loss": 0.2297,
+      "step": 4170
+    },
+    {
+      "epoch": 0.623747569911769,
+      "grad_norm": 1.61035540032609,
+      "learning_rate": 3.2754941297370872e-06,
+      "loss": 0.1548,
+      "step": 4171
+    },
+    {
+      "epoch": 0.6238971138029011,
+      "grad_norm": 1.717346159952656,
+      "learning_rate": 3.273221111798245e-06,
+      "loss": 0.4091,
+      "step": 4172
+    },
+    {
+      "epoch": 0.6240466576940332,
+      "grad_norm": 1.188823809082235,
+      "learning_rate": 3.2709484989783708e-06,
+      "loss": 0.2987,
+      "step": 4173
+    },
+    {
+      "epoch": 0.6241962015851652,
+      "grad_norm": 2.099313220051725,
+      "learning_rate": 3.2686762918106425e-06,
+      "loss": 0.6985,
+      "step": 4174
+    },
+    {
+      "epoch": 0.6243457454762973,
+      "grad_norm": 1.6757776086941085,
+      "learning_rate": 3.2664044908281413e-06,
+      "loss": 0.2769,
+      "step": 4175
+    },
+    {
+      "epoch": 0.6244952893674294,
+      "grad_norm": 1.6877573797191467,
+      "learning_rate": 3.2641330965638563e-06,
+      "loss": 0.2132,
+      "step": 4176
+    },
+    {
+      "epoch": 0.6246448332585613,
+      "grad_norm": 1.5982412325900388,
+      "learning_rate": 3.261862109550673e-06,
+      "loss": 0.437,
+      "step": 4177
+    },
+    {
+      "epoch": 0.6247943771496934,
+      "grad_norm": 1.3427594465047028,
+      "learning_rate": 3.2595915303213902e-06,
+      "loss": 0.3653,
+      "step": 4178
+    },
+    {
+      "epoch": 0.6249439210408255,
+      "grad_norm": 1.262764121214261,
+      "learning_rate": 3.2573213594087084e-06,
+      "loss": 0.179,
+      "step": 4179
+    },
+    {
+      "epoch": 0.6250934649319575,
+      "grad_norm": 1.1739874666573993,
+      "learning_rate": 3.2550515973452295e-06,
+      "loss": 0.2201,
+      "step": 4180
+    },
+    {
+      "epoch": 0.6252430088230896,
+      "grad_norm": 1.3535281039016394,
+      "learning_rate": 3.252782244663465e-06,
+      "loss": 0.1931,
+      "step": 4181
+    },
+    {
+      "epoch": 0.6253925527142217,
+      "grad_norm": 1.4264461743578645,
+      "learning_rate": 3.2505133018958224e-06,
+      "loss": 0.251,
+      "step": 4182
+    },
+    {
+      "epoch": 0.6255420966053536,
+      "grad_norm": 1.4547018012543345,
+      "learning_rate": 3.24824476957462e-06,
+      "loss": 0.2176,
+      "step": 4183
+    },
+    {
+      "epoch": 0.6256916404964857,
+      "grad_norm": 1.93315832456521,
+      "learning_rate": 3.2459766482320764e-06,
+      "loss": 0.5806,
+      "step": 4184
+    },
+    {
+      "epoch": 0.6258411843876177,
+      "grad_norm": 1.456983023451318,
+      "learning_rate": 3.243708938400316e-06,
+      "loss": 0.3683,
+      "step": 4185
+    },
+    {
+      "epoch": 0.6259907282787498,
+      "grad_norm": 1.2318359907860033,
+      "learning_rate": 3.2414416406113645e-06,
+      "loss": 0.176,
+      "step": 4186
+    },
+    {
+      "epoch": 0.6261402721698819,
+      "grad_norm": 1.8015103632850937,
+      "learning_rate": 3.239174755397152e-06,
+      "loss": 0.3032,
+      "step": 4187
+    },
+    {
+      "epoch": 0.6262898160610139,
+      "grad_norm": 1.3196497543247543,
+      "learning_rate": 3.2369082832895116e-06,
+      "loss": 0.2055,
+      "step": 4188
+    },
+    {
+      "epoch": 0.626439359952146,
+      "grad_norm": 1.4298858348667032,
+      "learning_rate": 3.2346422248201792e-06,
+      "loss": 0.3308,
+      "step": 4189
+    },
+    {
+      "epoch": 0.626588903843278,
+      "grad_norm": 1.3708295866539182,
+      "learning_rate": 3.232376580520794e-06,
+      "loss": 0.215,
+      "step": 4190
+    },
+    {
+      "epoch": 0.62673844773441,
+      "grad_norm": 1.4861225381998133,
+      "learning_rate": 3.2301113509228992e-06,
+      "loss": 0.3438,
+      "step": 4191
+    },
+    {
+      "epoch": 0.6268879916255421,
+      "grad_norm": 1.4461704439901406,
+      "learning_rate": 3.227846536557938e-06,
+      "loss": 0.328,
+      "step": 4192
+    },
+    {
+      "epoch": 0.6270375355166742,
+      "grad_norm": 1.811748443366764,
+      "learning_rate": 3.225582137957258e-06,
+      "loss": 0.3549,
+      "step": 4193
+    },
+    {
+      "epoch": 0.6271870794078062,
+      "grad_norm": 1.205193555796813,
+      "learning_rate": 3.223318155652109e-06,
+      "loss": 0.1965,
+      "step": 4194
+    },
+    {
+      "epoch": 0.6273366232989382,
+      "grad_norm": 1.2317204179453682,
+      "learning_rate": 3.2210545901736432e-06,
+      "loss": 0.2889,
+      "step": 4195
+    },
+    {
+      "epoch": 0.6274861671900703,
+      "grad_norm": 1.4296358807943457,
+      "learning_rate": 3.2187914420529176e-06,
+      "loss": 0.3053,
+      "step": 4196
+    },
+    {
+      "epoch": 0.6276357110812023,
+      "grad_norm": 1.832702074683805,
+      "learning_rate": 3.216528711820882e-06,
+      "loss": 0.4491,
+      "step": 4197
+    },
+    {
+      "epoch": 0.6277852549723344,
+      "grad_norm": 1.3247323137479683,
+      "learning_rate": 3.2142664000084e-06,
+      "loss": 0.1753,
+      "step": 4198
+    },
+    {
+      "epoch": 0.6279347988634665,
+      "grad_norm": 1.4737325221279896,
+      "learning_rate": 3.2120045071462303e-06,
+      "loss": 0.1908,
+      "step": 4199
+    },
+    {
+      "epoch": 0.6280843427545985,
+      "grad_norm": 2.128037805773552,
+      "learning_rate": 3.2097430337650347e-06,
+      "loss": 0.4037,
+      "step": 4200
+    },
+    {
+      "epoch": 0.6282338866457305,
+      "grad_norm": 1.9359432049498655,
+      "learning_rate": 3.207481980395379e-06,
+      "loss": 0.3577,
+      "step": 4201
+    },
+    {
+      "epoch": 0.6283834305368625,
+      "grad_norm": 1.4914189826217994,
+      "learning_rate": 3.205221347567723e-06,
+      "loss": 0.3296,
+      "step": 4202
+    },
+    {
+      "epoch": 0.6285329744279946,
+      "grad_norm": 1.548420456754965,
+      "learning_rate": 3.202961135812437e-06,
+      "loss": 0.3748,
+      "step": 4203
+    },
+    {
+      "epoch": 0.6286825183191267,
+      "grad_norm": 1.9692451823118604,
+      "learning_rate": 3.2007013456597864e-06,
+      "loss": 0.1725,
+      "step": 4204
+    },
+    {
+      "epoch": 0.6288320622102587,
+      "grad_norm": 1.420481431208107,
+      "learning_rate": 3.1984419776399413e-06,
+      "loss": 0.2428,
+      "step": 4205
+    },
+    {
+      "epoch": 0.6289816061013908,
+      "grad_norm": 1.3526682134449342,
+      "learning_rate": 3.1961830322829707e-06,
+      "loss": 0.1709,
+      "step": 4206
+    },
+    {
+      "epoch": 0.6291311499925228,
+      "grad_norm": 1.5002481533459355,
+      "learning_rate": 3.1939245101188444e-06,
+      "loss": 0.3167,
+      "step": 4207
+    },
+    {
+      "epoch": 0.6292806938836548,
+      "grad_norm": 2.3464477589549126,
+      "learning_rate": 3.1916664116774333e-06,
+      "loss": 0.3406,
+      "step": 4208
+    },
+    {
+      "epoch": 0.6294302377747869,
+      "grad_norm": 1.2641045597562264,
+      "learning_rate": 3.1894087374885095e-06,
+      "loss": 0.2122,
+      "step": 4209
+    },
+    {
+      "epoch": 0.629579781665919,
+      "grad_norm": 1.3787624835316412,
+      "learning_rate": 3.1871514880817454e-06,
+      "loss": 0.2141,
+      "step": 4210
+    },
+    {
+      "epoch": 0.629729325557051,
+      "grad_norm": 1.2143198490479026,
+      "learning_rate": 3.184894663986715e-06,
+      "loss": 0.2066,
+      "step": 4211
+    },
+    {
+      "epoch": 0.629878869448183,
+      "grad_norm": 1.5367748497524958,
+      "learning_rate": 3.1826382657328877e-06,
+      "loss": 0.2403,
+      "step": 4212
+    },
+    {
+      "epoch": 0.6300284133393151,
+      "grad_norm": 1.2749341203227855,
+      "learning_rate": 3.1803822938496377e-06,
+      "loss": 0.2034,
+      "step": 4213
+    },
+    {
+      "epoch": 0.6301779572304471,
+      "grad_norm": 1.8350648672932401,
+      "learning_rate": 3.1781267488662383e-06,
+      "loss": 0.1805,
+      "step": 4214
+    },
+    {
+      "epoch": 0.6303275011215792,
+      "grad_norm": 2.0058052776288537,
+      "learning_rate": 3.175871631311861e-06,
+      "loss": 0.2047,
+      "step": 4215
+    },
+    {
+      "epoch": 0.6304770450127112,
+      "grad_norm": 1.7220131761526412,
+      "learning_rate": 3.1736169417155814e-06,
+      "loss": 0.4788,
+      "step": 4216
+    },
+    {
+      "epoch": 0.6306265889038433,
+      "grad_norm": 1.40884208997203,
+      "learning_rate": 3.171362680606366e-06,
+      "loss": 0.3919,
+      "step": 4217
+    },
+    {
+      "epoch": 0.6307761327949754,
+      "grad_norm": 1.4379399445316037,
+      "learning_rate": 3.1691088485130896e-06,
+      "loss": 0.2367,
+      "step": 4218
+    },
+    {
+      "epoch": 0.6309256766861073,
+      "grad_norm": 1.046787066111561,
+      "learning_rate": 3.1668554459645217e-06,
+      "loss": 0.1705,
+      "step": 4219
+    },
+    {
+      "epoch": 0.6310752205772394,
+      "grad_norm": 1.8102251238144302,
+      "learning_rate": 3.164602473489333e-06,
+      "loss": 0.2386,
+      "step": 4220
+    },
+    {
+      "epoch": 0.6312247644683715,
+      "grad_norm": 1.6606923415837127,
+      "learning_rate": 3.1623499316160955e-06,
+      "loss": 0.3065,
+      "step": 4221
+    },
+    {
+      "epoch": 0.6313743083595035,
+      "grad_norm": 1.6277864893202596,
+      "learning_rate": 3.1600978208732712e-06,
+      "loss": 0.1967,
+      "step": 4222
+    },
+    {
+      "epoch": 0.6315238522506356,
+      "grad_norm": 1.5787248038141106,
+      "learning_rate": 3.15784614178923e-06,
+      "loss": 0.2251,
+      "step": 4223
+    },
+    {
+      "epoch": 0.6316733961417677,
+      "grad_norm": 1.2158381326740617,
+      "learning_rate": 3.1555948948922376e-06,
+      "loss": 0.2048,
+      "step": 4224
+    },
+    {
+      "epoch": 0.6318229400328996,
+      "grad_norm": 1.3127440129245087,
+      "learning_rate": 3.153344080710459e-06,
+      "loss": 0.2573,
+      "step": 4225
+    },
+    {
+      "epoch": 0.6319724839240317,
+      "grad_norm": 1.7939040127871422,
+      "learning_rate": 3.1510936997719557e-06,
+      "loss": 0.2294,
+      "step": 4226
+    },
+    {
+      "epoch": 0.6321220278151638,
+      "grad_norm": 1.3974871896214638,
+      "learning_rate": 3.1488437526046876e-06,
+      "loss": 0.2024,
+      "step": 4227
+    },
+    {
+      "epoch": 0.6322715717062958,
+      "grad_norm": 1.6249650116968457,
+      "learning_rate": 3.1465942397365168e-06,
+      "loss": 0.2019,
+      "step": 4228
+    },
+    {
+      "epoch": 0.6324211155974279,
+      "grad_norm": 1.3398096425022026,
+      "learning_rate": 3.144345161695199e-06,
+      "loss": 0.1755,
+      "step": 4229
+    },
+    {
+      "epoch": 0.6325706594885598,
+      "grad_norm": 1.476733344149943,
+      "learning_rate": 3.142096519008389e-06,
+      "loss": 0.1771,
+      "step": 4230
+    },
+    {
+      "epoch": 0.6327202033796919,
+      "grad_norm": 1.471079632513585,
+      "learning_rate": 3.1398483122036427e-06,
+      "loss": 0.4238,
+      "step": 4231
+    },
+    {
+      "epoch": 0.632869747270824,
+      "grad_norm": 1.6381073298660154,
+      "learning_rate": 3.1376005418084082e-06,
+      "loss": 0.3599,
+      "step": 4232
+    },
+    {
+      "epoch": 0.633019291161956,
+      "grad_norm": 1.7979103671376482,
+      "learning_rate": 3.135353208350035e-06,
+      "loss": 0.2413,
+      "step": 4233
+    },
+    {
+      "epoch": 0.6331688350530881,
+      "grad_norm": 1.9747639038112241,
+      "learning_rate": 3.1331063123557692e-06,
+      "loss": 0.3426,
+      "step": 4234
+    },
+    {
+      "epoch": 0.6333183789442202,
+      "grad_norm": 1.6627981863210517,
+      "learning_rate": 3.130859854352755e-06,
+      "loss": 0.228,
+      "step": 4235
+    },
+    {
+      "epoch": 0.6334679228353521,
+      "grad_norm": 1.584328384767029,
+      "learning_rate": 3.128613834868035e-06,
+      "loss": 0.1844,
+      "step": 4236
+    },
+    {
+      "epoch": 0.6336174667264842,
+      "grad_norm": 2.0955618164544694,
+      "learning_rate": 3.126368254428541e-06,
+      "loss": 0.5052,
+      "step": 4237
+    },
+    {
+      "epoch": 0.6337670106176163,
+      "grad_norm": 1.6176100653355505,
+      "learning_rate": 3.1241231135611116e-06,
+      "loss": 0.2819,
+      "step": 4238
+    },
+    {
+      "epoch": 0.6339165545087483,
+      "grad_norm": 2.9551331489114685,
+      "learning_rate": 3.1218784127924795e-06,
+      "loss": 0.5355,
+      "step": 4239
+    },
+    {
+      "epoch": 0.6340660983998804,
+      "grad_norm": 1.8688017618703836,
+      "learning_rate": 3.1196341526492715e-06,
+      "loss": 0.2341,
+      "step": 4240
+    },
+    {
+      "epoch": 0.6342156422910125,
+      "grad_norm": 1.9962730625781477,
+      "learning_rate": 3.1173903336580146e-06,
+      "loss": 0.3511,
+      "step": 4241
+    },
+    {
+      "epoch": 0.6343651861821444,
+      "grad_norm": 1.6150488602773196,
+      "learning_rate": 3.1151469563451275e-06,
+      "loss": 0.221,
+      "step": 4242
+    },
+    {
+      "epoch": 0.6345147300732765,
+      "grad_norm": 2.3530875232138797,
+      "learning_rate": 3.1129040212369286e-06,
+      "loss": 0.3598,
+      "step": 4243
+    },
+    {
+      "epoch": 0.6346642739644085,
+      "grad_norm": 1.2420835811404023,
+      "learning_rate": 3.1106615288596337e-06,
+      "loss": 0.2156,
+      "step": 4244
+    },
+    {
+      "epoch": 0.6348138178555406,
+      "grad_norm": 1.401385423495105,
+      "learning_rate": 3.108419479739352e-06,
+      "loss": 0.2142,
+      "step": 4245
+    },
+    {
+      "epoch": 0.6349633617466727,
+      "grad_norm": 2.1568206778278993,
+      "learning_rate": 3.1061778744020916e-06,
+      "loss": 0.2041,
+      "step": 4246
+    },
+    {
+      "epoch": 0.6351129056378046,
+      "grad_norm": 1.2507736523913935,
+      "learning_rate": 3.1039367133737517e-06,
+      "loss": 0.1835,
+      "step": 4247
+    },
+    {
+      "epoch": 0.6352624495289367,
+      "grad_norm": 1.7213048240141702,
+      "learning_rate": 3.1016959971801326e-06,
+      "loss": 0.4563,
+      "step": 4248
+    },
+    {
+      "epoch": 0.6354119934200688,
+      "grad_norm": 1.2148837576653622,
+      "learning_rate": 3.0994557263469267e-06,
+      "loss": 0.1998,
+      "step": 4249
+    },
+    {
+      "epoch": 0.6355615373112008,
+      "grad_norm": 1.848918443348012,
+      "learning_rate": 3.097215901399724e-06,
+      "loss": 0.4183,
+      "step": 4250
+    },
+    {
+      "epoch": 0.6357110812023329,
+      "grad_norm": 1.447495899079333,
+      "learning_rate": 3.0949765228640094e-06,
+      "loss": 0.2844,
+      "step": 4251
+    },
+    {
+      "epoch": 0.635860625093465,
+      "grad_norm": 1.9257511306790576,
+      "learning_rate": 3.092737591265161e-06,
+      "loss": 0.1671,
+      "step": 4252
+    },
+    {
+      "epoch": 0.636010168984597,
+      "grad_norm": 1.81538166148831,
+      "learning_rate": 3.0904991071284544e-06,
+      "loss": 0.3598,
+      "step": 4253
+    },
+    {
+      "epoch": 0.636159712875729,
+      "grad_norm": 1.840220273117095,
+      "learning_rate": 3.0882610709790606e-06,
+      "loss": 0.2887,
+      "step": 4254
+    },
+    {
+      "epoch": 0.6363092567668611,
+      "grad_norm": 2.161029128850978,
+      "learning_rate": 3.086023483342043e-06,
+      "loss": 0.6133,
+      "step": 4255
+    },
+    {
+      "epoch": 0.6364588006579931,
+      "grad_norm": 1.6267242774433444,
+      "learning_rate": 3.083786344742362e-06,
+      "loss": 0.217,
+      "step": 4256
+    },
+    {
+      "epoch": 0.6366083445491252,
+      "grad_norm": 1.5140885204209047,
+      "learning_rate": 3.081549655704874e-06,
+      "loss": 0.3897,
+      "step": 4257
+    },
+    {
+      "epoch": 0.6367578884402573,
+      "grad_norm": 1.5106568808657759,
+      "learning_rate": 3.079313416754322e-06,
+      "loss": 0.1983,
+      "step": 4258
+    },
+    {
+      "epoch": 0.6369074323313892,
+      "grad_norm": 1.6011885974138298,
+      "learning_rate": 3.0770776284153544e-06,
+      "loss": 0.2696,
+      "step": 4259
+    },
+    {
+      "epoch": 0.6370569762225213,
+      "grad_norm": 1.7578971545709878,
+      "learning_rate": 3.0748422912125066e-06,
+      "loss": 0.3563,
+      "step": 4260
+    },
+    {
+      "epoch": 0.6372065201136533,
+      "grad_norm": 1.4522858152587212,
+      "learning_rate": 3.072607405670211e-06,
+      "loss": 0.2182,
+      "step": 4261
+    },
+    {
+      "epoch": 0.6373560640047854,
+      "grad_norm": 1.601429394098701,
+      "learning_rate": 3.070372972312795e-06,
+      "loss": 0.3268,
+      "step": 4262
+    },
+    {
+      "epoch": 0.6375056078959175,
+      "grad_norm": 1.586472172174774,
+      "learning_rate": 3.0681389916644745e-06,
+      "loss": 0.1793,
+      "step": 4263
+    },
+    {
+      "epoch": 0.6376551517870495,
+      "grad_norm": 1.6949417042764672,
+      "learning_rate": 3.065905464249364e-06,
+      "loss": 0.2796,
+      "step": 4264
+    },
+    {
+      "epoch": 0.6378046956781815,
+      "grad_norm": 1.3534270198383676,
+      "learning_rate": 3.0636723905914717e-06,
+      "loss": 0.3281,
+      "step": 4265
+    },
+    {
+      "epoch": 0.6379542395693136,
+      "grad_norm": 1.3866514082125108,
+      "learning_rate": 3.0614397712146974e-06,
+      "loss": 0.2263,
+      "step": 4266
+    },
+    {
+      "epoch": 0.6381037834604456,
+      "grad_norm": 1.8462458786372813,
+      "learning_rate": 3.0592076066428367e-06,
+      "loss": 0.5081,
+      "step": 4267
+    },
+    {
+      "epoch": 0.6382533273515777,
+      "grad_norm": 1.5215387444826622,
+      "learning_rate": 3.0569758973995745e-06,
+      "loss": 0.3697,
+      "step": 4268
+    },
+    {
+      "epoch": 0.6384028712427098,
+      "grad_norm": 1.572814999435716,
+      "learning_rate": 3.0547446440084914e-06,
+      "loss": 0.3098,
+      "step": 4269
+    },
+    {
+      "epoch": 0.6385524151338418,
+      "grad_norm": 1.3241548494658333,
+      "learning_rate": 3.052513846993063e-06,
+      "loss": 0.3048,
+      "step": 4270
+    },
+    {
+      "epoch": 0.6387019590249738,
+      "grad_norm": 1.5147109741247766,
+      "learning_rate": 3.0502835068766545e-06,
+      "loss": 0.3428,
+      "step": 4271
+    },
+    {
+      "epoch": 0.6388515029161059,
+      "grad_norm": 1.7322061612980246,
+      "learning_rate": 3.0480536241825263e-06,
+      "loss": 0.3386,
+      "step": 4272
+    },
+    {
+      "epoch": 0.6390010468072379,
+      "grad_norm": 1.5645781715890368,
+      "learning_rate": 3.045824199433829e-06,
+      "loss": 0.2606,
+      "step": 4273
+    },
+    {
+      "epoch": 0.63915059069837,
+      "grad_norm": 2.3293893184482797,
+      "learning_rate": 3.0435952331536072e-06,
+      "loss": 0.5555,
+      "step": 4274
+    },
+    {
+      "epoch": 0.639300134589502,
+      "grad_norm": 1.0615012965836783,
+      "learning_rate": 3.0413667258647983e-06,
+      "loss": 0.1921,
+      "step": 4275
+    },
+    {
+      "epoch": 0.639449678480634,
+      "grad_norm": 1.6734758493782524,
+      "learning_rate": 3.039138678090232e-06,
+      "loss": 0.2011,
+      "step": 4276
+    },
+    {
+      "epoch": 0.6395992223717661,
+      "grad_norm": 2.1157051612171163,
+      "learning_rate": 3.036911090352631e-06,
+      "loss": 0.4092,
+      "step": 4277
+    },
+    {
+      "epoch": 0.6397487662628981,
+      "grad_norm": 1.3554481867568882,
+      "learning_rate": 3.034683963174604e-06,
+      "loss": 0.1613,
+      "step": 4278
+    },
+    {
+      "epoch": 0.6398983101540302,
+      "grad_norm": 1.3257952142585847,
+      "learning_rate": 3.0324572970786607e-06,
+      "loss": 0.2839,
+      "step": 4279
+    },
+    {
+      "epoch": 0.6400478540451623,
+      "grad_norm": 1.691758447323851,
+      "learning_rate": 3.030231092587198e-06,
+      "loss": 0.4357,
+      "step": 4280
+    },
+    {
+      "epoch": 0.6401973979362943,
+      "grad_norm": 1.403610467612681,
+      "learning_rate": 3.028005350222504e-06,
+      "loss": 0.3382,
+      "step": 4281
+    },
+    {
+      "epoch": 0.6403469418274264,
+      "grad_norm": 1.2279332452901197,
+      "learning_rate": 3.0257800705067614e-06,
+      "loss": 0.2253,
+      "step": 4282
+    },
+    {
+      "epoch": 0.6404964857185584,
+      "grad_norm": 1.3097407879338883,
+      "learning_rate": 3.0235552539620384e-06,
+      "loss": 0.2013,
+      "step": 4283
+    },
+    {
+      "epoch": 0.6406460296096904,
+      "grad_norm": 1.7716307700503333,
+      "learning_rate": 3.021330901110301e-06,
+      "loss": 0.3922,
+      "step": 4284
+    },
+    {
+      "epoch": 0.6407955735008225,
+      "grad_norm": 1.4417814009504493,
+      "learning_rate": 3.0191070124734034e-06,
+      "loss": 0.3302,
+      "step": 4285
+    },
+    {
+      "epoch": 0.6409451173919546,
+      "grad_norm": 1.3554969696713202,
+      "learning_rate": 3.016883588573091e-06,
+      "loss": 0.1897,
+      "step": 4286
+    },
+    {
+      "epoch": 0.6410946612830866,
+      "grad_norm": 1.6342674280144482,
+      "learning_rate": 3.014660629931002e-06,
+      "loss": 0.2236,
+      "step": 4287
+    },
+    {
+      "epoch": 0.6412442051742187,
+      "grad_norm": 1.2027823026684163,
+      "learning_rate": 3.0124381370686617e-06,
+      "loss": 0.188,
+      "step": 4288
+    },
+    {
+      "epoch": 0.6413937490653506,
+      "grad_norm": 1.0825612297165725,
+      "learning_rate": 3.010216110507489e-06,
+      "loss": 0.1893,
+      "step": 4289
+    },
+    {
+      "epoch": 0.6415432929564827,
+      "grad_norm": 1.8647778656423657,
+      "learning_rate": 3.007994550768793e-06,
+      "loss": 0.4889,
+      "step": 4290
+    },
+    {
+      "epoch": 0.6416928368476148,
+      "grad_norm": 1.3882812293916185,
+      "learning_rate": 3.005773458373773e-06,
+      "loss": 0.2079,
+      "step": 4291
+    },
+    {
+      "epoch": 0.6418423807387468,
+      "grad_norm": 1.3562041151287785,
+      "learning_rate": 3.0035528338435205e-06,
+      "loss": 0.2236,
+      "step": 4292
+    },
+    {
+      "epoch": 0.6419919246298789,
+      "grad_norm": 1.3686846791536245,
+      "learning_rate": 3.001332677699012e-06,
+      "loss": 0.2996,
+      "step": 4293
+    },
+    {
+      "epoch": 0.642141468521011,
+      "grad_norm": 1.8880382962351019,
+      "learning_rate": 2.9991129904611193e-06,
+      "loss": 0.3409,
+      "step": 4294
+    },
+    {
+      "epoch": 0.6422910124121429,
+      "grad_norm": 1.112992761276077,
+      "learning_rate": 2.996893772650602e-06,
+      "loss": 0.1865,
+      "step": 4295
+    },
+    {
+      "epoch": 0.642440556303275,
+      "grad_norm": 2.0473388917881783,
+      "learning_rate": 2.9946750247881107e-06,
+      "loss": 0.4307,
+      "step": 4296
+    },
+    {
+      "epoch": 0.6425901001944071,
+      "grad_norm": 1.3086798312214234,
+      "learning_rate": 2.9924567473941867e-06,
+      "loss": 0.1451,
+      "step": 4297
+    },
+    {
+      "epoch": 0.6427396440855391,
+      "grad_norm": 1.329348124352134,
+      "learning_rate": 2.9902389409892553e-06,
+      "loss": 0.3251,
+      "step": 4298
+    },
+    {
+      "epoch": 0.6428891879766712,
+      "grad_norm": 1.348801188692966,
+      "learning_rate": 2.9880216060936364e-06,
+      "loss": 0.1869,
+      "step": 4299
+    },
+    {
+      "epoch": 0.6430387318678032,
+      "grad_norm": 1.7721779505445987,
+      "learning_rate": 2.9858047432275387e-06,
+      "loss": 0.3939,
+      "step": 4300
+    },
+    {
+      "epoch": 0.6431882757589352,
+      "grad_norm": 2.270411296146692,
+      "learning_rate": 2.983588352911061e-06,
+      "loss": 0.1998,
+      "step": 4301
+    },
+    {
+      "epoch": 0.6433378196500673,
+      "grad_norm": 1.3542833009033532,
+      "learning_rate": 2.98137243566419e-06,
+      "loss": 0.1676,
+      "step": 4302
+    },
+    {
+      "epoch": 0.6434873635411994,
+      "grad_norm": 1.7866702598193438,
+      "learning_rate": 2.9791569920067975e-06,
+      "loss": 0.2745,
+      "step": 4303
+    },
+    {
+      "epoch": 0.6436369074323314,
+      "grad_norm": 1.7819432699718787,
+      "learning_rate": 2.976942022458651e-06,
+      "loss": 0.3678,
+      "step": 4304
+    },
+    {
+      "epoch": 0.6437864513234635,
+      "grad_norm": 1.9658303788785358,
+      "learning_rate": 2.974727527539403e-06,
+      "loss": 0.4969,
+      "step": 4305
+    },
+    {
+      "epoch": 0.6439359952145954,
+      "grad_norm": 1.607317627948274,
+      "learning_rate": 2.9725135077685946e-06,
+      "loss": 0.3529,
+      "step": 4306
+    },
+    {
+      "epoch": 0.6440855391057275,
+      "grad_norm": 1.475033336602126,
+      "learning_rate": 2.9702999636656584e-06,
+      "loss": 0.2222,
+      "step": 4307
+    },
+    {
+      "epoch": 0.6442350829968596,
+      "grad_norm": 1.973299582548237,
+      "learning_rate": 2.968086895749911e-06,
+      "loss": 0.2234,
+      "step": 4308
+    },
+    {
+      "epoch": 0.6443846268879916,
+      "grad_norm": 1.7058778528147858,
+      "learning_rate": 2.9658743045405593e-06,
+      "loss": 0.2131,
+      "step": 4309
+    },
+    {
+      "epoch": 0.6445341707791237,
+      "grad_norm": 0.9302716463605547,
+      "learning_rate": 2.9636621905566997e-06,
+      "loss": 0.1485,
+      "step": 4310
+    },
+    {
+      "epoch": 0.6446837146702558,
+      "grad_norm": 1.1167565995299549,
+      "learning_rate": 2.9614505543173154e-06,
+      "loss": 0.2204,
+      "step": 4311
+    },
+    {
+      "epoch": 0.6448332585613877,
+      "grad_norm": 1.5708854066715467,
+      "learning_rate": 2.959239396341278e-06,
+      "loss": 0.2758,
+      "step": 4312
+    },
+    {
+      "epoch": 0.6449828024525198,
+      "grad_norm": 2.089312733718862,
+      "learning_rate": 2.957028717147345e-06,
+      "loss": 0.201,
+      "step": 4313
+    },
+    {
+      "epoch": 0.6451323463436519,
+      "grad_norm": 2.1281298525302588,
+      "learning_rate": 2.954818517254164e-06,
+      "loss": 0.4695,
+      "step": 4314
+    },
+    {
+      "epoch": 0.6452818902347839,
+      "grad_norm": 1.5959774304509975,
+      "learning_rate": 2.9526087971802685e-06,
+      "loss": 0.3108,
+      "step": 4315
+    },
+    {
+      "epoch": 0.645431434125916,
+      "grad_norm": 1.5718195549000646,
+      "learning_rate": 2.950399557444081e-06,
+      "loss": 0.3204,
+      "step": 4316
+    },
+    {
+      "epoch": 0.6455809780170481,
+      "grad_norm": 1.4213056419991819,
+      "learning_rate": 2.948190798563912e-06,
+      "loss": 0.3553,
+      "step": 4317
+    },
+    {
+      "epoch": 0.64573052190818,
+      "grad_norm": 1.8761753689830492,
+      "learning_rate": 2.9459825210579534e-06,
+      "loss": 0.361,
+      "step": 4318
+    },
+    {
+      "epoch": 0.6458800657993121,
+      "grad_norm": 1.9599100361948998,
+      "learning_rate": 2.94377472544429e-06,
+      "loss": 0.5,
+      "step": 4319
+    },
+    {
+      "epoch": 0.6460296096904441,
+      "grad_norm": 1.452596811980564,
+      "learning_rate": 2.941567412240893e-06,
+      "loss": 0.3076,
+      "step": 4320
+    },
+    {
+      "epoch": 0.6461791535815762,
+      "grad_norm": 1.867725765581649,
+      "learning_rate": 2.9393605819656203e-06,
+      "loss": 0.4896,
+      "step": 4321
+    },
+    {
+      "epoch": 0.6463286974727083,
+      "grad_norm": 2.0028651025738604,
+      "learning_rate": 2.937154235136215e-06,
+      "loss": 0.4414,
+      "step": 4322
+    },
+    {
+      "epoch": 0.6464782413638402,
+      "grad_norm": 1.444525605000369,
+      "learning_rate": 2.934948372270305e-06,
+      "loss": 0.1703,
+      "step": 4323
+    },
+    {
+      "epoch": 0.6466277852549723,
+      "grad_norm": 1.6260674196352805,
+      "learning_rate": 2.932742993885408e-06,
+      "loss": 0.2936,
+      "step": 4324
+    },
+    {
+      "epoch": 0.6467773291461044,
+      "grad_norm": 1.6357422467132101,
+      "learning_rate": 2.930538100498928e-06,
+      "loss": 0.1485,
+      "step": 4325
+    },
+    {
+      "epoch": 0.6469268730372364,
+      "grad_norm": 1.4888503420859003,
+      "learning_rate": 2.9283336926281535e-06,
+      "loss": 0.49,
+      "step": 4326
+    },
+    {
+      "epoch": 0.6470764169283685,
+      "grad_norm": 1.0301201982118935,
+      "learning_rate": 2.9261297707902614e-06,
+      "loss": 0.205,
+      "step": 4327
+    },
+    {
+      "epoch": 0.6472259608195006,
+      "grad_norm": 1.5229938350109378,
+      "learning_rate": 2.9239263355023104e-06,
+      "loss": 0.3428,
+      "step": 4328
+    },
+    {
+      "epoch": 0.6473755047106325,
+      "grad_norm": 1.3838220241069505,
+      "learning_rate": 2.9217233872812502e-06,
+      "loss": 0.2269,
+      "step": 4329
+    },
+    {
+      "epoch": 0.6475250486017646,
+      "grad_norm": 2.00517100675116,
+      "learning_rate": 2.919520926643911e-06,
+      "loss": 0.5935,
+      "step": 4330
+    },
+    {
+      "epoch": 0.6476745924928967,
+      "grad_norm": 1.4602024554784119,
+      "learning_rate": 2.9173189541070124e-06,
+      "loss": 0.3243,
+      "step": 4331
+    },
+    {
+      "epoch": 0.6478241363840287,
+      "grad_norm": 1.160834929528934,
+      "learning_rate": 2.9151174701871616e-06,
+      "loss": 0.1662,
+      "step": 4332
+    },
+    {
+      "epoch": 0.6479736802751608,
+      "grad_norm": 1.2978941922617255,
+      "learning_rate": 2.9129164754008433e-06,
+      "loss": 0.2174,
+      "step": 4333
+    },
+    {
+      "epoch": 0.6481232241662928,
+      "grad_norm": 1.2391057330727242,
+      "learning_rate": 2.910715970264433e-06,
+      "loss": 0.182,
+      "step": 4334
+    },
+    {
+      "epoch": 0.6482727680574248,
+      "grad_norm": 1.1677889159191694,
+      "learning_rate": 2.908515955294192e-06,
+      "loss": 0.1465,
+      "step": 4335
+    },
+    {
+      "epoch": 0.6484223119485569,
+      "grad_norm": 1.8166755640689873,
+      "learning_rate": 2.9063164310062643e-06,
+      "loss": 0.2768,
+      "step": 4336
+    },
+    {
+      "epoch": 0.6485718558396889,
+      "grad_norm": 1.3259310299996228,
+      "learning_rate": 2.9041173979166813e-06,
+      "loss": 0.2222,
+      "step": 4337
+    },
+    {
+      "epoch": 0.648721399730821,
+      "grad_norm": 1.5337321468214244,
+      "learning_rate": 2.9019188565413535e-06,
+      "loss": 0.2905,
+      "step": 4338
+    },
+    {
+      "epoch": 0.6488709436219531,
+      "grad_norm": 1.5311289564286026,
+      "learning_rate": 2.899720807396082e-06,
+      "loss": 0.3367,
+      "step": 4339
+    },
+    {
+      "epoch": 0.649020487513085,
+      "grad_norm": 1.5629391529927086,
+      "learning_rate": 2.89752325099655e-06,
+      "loss": 0.2184,
+      "step": 4340
+    },
+    {
+      "epoch": 0.6491700314042171,
+      "grad_norm": 1.1937759596313804,
+      "learning_rate": 2.8953261878583263e-06,
+      "loss": 0.1794,
+      "step": 4341
+    },
+    {
+      "epoch": 0.6493195752953492,
+      "grad_norm": 1.8427953915887811,
+      "learning_rate": 2.8931296184968614e-06,
+      "loss": 0.2087,
+      "step": 4342
+    },
+    {
+      "epoch": 0.6494691191864812,
+      "grad_norm": 1.3918761481043547,
+      "learning_rate": 2.8909335434274932e-06,
+      "loss": 0.2514,
+      "step": 4343
+    },
+    {
+      "epoch": 0.6496186630776133,
+      "grad_norm": 2.3172764293559864,
+      "learning_rate": 2.888737963165442e-06,
+      "loss": 0.394,
+      "step": 4344
+    },
+    {
+      "epoch": 0.6497682069687454,
+      "grad_norm": 1.34481782125672,
+      "learning_rate": 2.886542878225811e-06,
+      "loss": 0.2146,
+      "step": 4345
+    },
+    {
+      "epoch": 0.6499177508598774,
+      "grad_norm": 1.3706664753686384,
+      "learning_rate": 2.8843482891235895e-06,
+      "loss": 0.1911,
+      "step": 4346
+    },
+    {
+      "epoch": 0.6500672947510094,
+      "grad_norm": 0.9839726564346342,
+      "learning_rate": 2.8821541963736482e-06,
+      "loss": 0.1612,
+      "step": 4347
+    },
+    {
+      "epoch": 0.6502168386421414,
+      "grad_norm": 2.3385887027986003,
+      "learning_rate": 2.8799606004907452e-06,
+      "loss": 0.3699,
+      "step": 4348
+    },
+    {
+      "epoch": 0.6503663825332735,
+      "grad_norm": 0.9434101059503864,
+      "learning_rate": 2.8777675019895144e-06,
+      "loss": 0.2116,
+      "step": 4349
+    },
+    {
+      "epoch": 0.6505159264244056,
+      "grad_norm": 1.4524046063305824,
+      "learning_rate": 2.8755749013844813e-06,
+      "loss": 0.2218,
+      "step": 4350
+    },
+    {
+      "epoch": 0.6506654703155376,
+      "grad_norm": 1.380876290138162,
+      "learning_rate": 2.873382799190049e-06,
+      "loss": 0.2275,
+      "step": 4351
+    },
+    {
+      "epoch": 0.6508150142066697,
+      "grad_norm": 1.2042028565506917,
+      "learning_rate": 2.871191195920508e-06,
+      "loss": 0.2236,
+      "step": 4352
+    },
+    {
+      "epoch": 0.6509645580978017,
+      "grad_norm": 1.561198882536686,
+      "learning_rate": 2.86900009209003e-06,
+      "loss": 0.4117,
+      "step": 4353
+    },
+    {
+      "epoch": 0.6511141019889337,
+      "grad_norm": 1.78117218182217,
+      "learning_rate": 2.8668094882126658e-06,
+      "loss": 0.363,
+      "step": 4354
+    },
+    {
+      "epoch": 0.6512636458800658,
+      "grad_norm": 1.4500223615377268,
+      "learning_rate": 2.864619384802354e-06,
+      "loss": 0.2251,
+      "step": 4355
+    },
+    {
+      "epoch": 0.6514131897711979,
+      "grad_norm": 1.8816120507444667,
+      "learning_rate": 2.862429782372914e-06,
+      "loss": 0.5379,
+      "step": 4356
+    },
+    {
+      "epoch": 0.6515627336623299,
+      "grad_norm": 1.2483361555174965,
+      "learning_rate": 2.860240681438048e-06,
+      "loss": 0.2536,
+      "step": 4357
+    },
+    {
+      "epoch": 0.651712277553462,
+      "grad_norm": 2.434529906670917,
+      "learning_rate": 2.858052082511339e-06,
+      "loss": 0.359,
+      "step": 4358
+    },
+    {
+      "epoch": 0.651861821444594,
+      "grad_norm": 1.3095152101580132,
+      "learning_rate": 2.8558639861062544e-06,
+      "loss": 0.3292,
+      "step": 4359
+    },
+    {
+      "epoch": 0.652011365335726,
+      "grad_norm": 1.0014669324387186,
+      "learning_rate": 2.853676392736142e-06,
+      "loss": 0.1774,
+      "step": 4360
+    },
+    {
+      "epoch": 0.6521609092268581,
+      "grad_norm": 1.107667196103754,
+      "learning_rate": 2.8514893029142337e-06,
+      "loss": 0.1987,
+      "step": 4361
+    },
+    {
+      "epoch": 0.6523104531179902,
+      "grad_norm": 0.8289718211895785,
+      "learning_rate": 2.8493027171536403e-06,
+      "loss": 0.1364,
+      "step": 4362
+    },
+    {
+      "epoch": 0.6524599970091222,
+      "grad_norm": 1.0509461340888442,
+      "learning_rate": 2.847116635967359e-06,
+      "loss": 0.1618,
+      "step": 4363
+    },
+    {
+      "epoch": 0.6526095409002542,
+      "grad_norm": 1.9624907660418935,
+      "learning_rate": 2.844931059868261e-06,
+      "loss": 0.5125,
+      "step": 4364
+    },
+    {
+      "epoch": 0.6527590847913862,
+      "grad_norm": 1.1521614286081878,
+      "learning_rate": 2.842745989369106e-06,
+      "loss": 0.2053,
+      "step": 4365
+    },
+    {
+      "epoch": 0.6529086286825183,
+      "grad_norm": 1.4513340162465773,
+      "learning_rate": 2.840561424982531e-06,
+      "loss": 0.3688,
+      "step": 4366
+    },
+    {
+      "epoch": 0.6530581725736504,
+      "grad_norm": 2.464838230471469,
+      "learning_rate": 2.838377367221057e-06,
+      "loss": 0.7219,
+      "step": 4367
+    },
+    {
+      "epoch": 0.6532077164647824,
+      "grad_norm": 1.297373135966818,
+      "learning_rate": 2.8361938165970876e-06,
+      "loss": 0.2473,
+      "step": 4368
+    },
+    {
+      "epoch": 0.6533572603559145,
+      "grad_norm": 1.5055358776731897,
+      "learning_rate": 2.8340107736229e-06,
+      "loss": 0.411,
+      "step": 4369
+    },
+    {
+      "epoch": 0.6535068042470465,
+      "grad_norm": 1.6739994793740698,
+      "learning_rate": 2.831828238810658e-06,
+      "loss": 0.4319,
+      "step": 4370
+    },
+    {
+      "epoch": 0.6536563481381785,
+      "grad_norm": 1.2825319924370413,
+      "learning_rate": 2.8296462126724077e-06,
+      "loss": 0.2329,
+      "step": 4371
+    },
+    {
+      "epoch": 0.6538058920293106,
+      "grad_norm": 1.5202140426315065,
+      "learning_rate": 2.8274646957200724e-06,
+      "loss": 0.2369,
+      "step": 4372
+    },
+    {
+      "epoch": 0.6539554359204427,
+      "grad_norm": 1.739325259152319,
+      "learning_rate": 2.8252836884654594e-06,
+      "loss": 0.3713,
+      "step": 4373
+    },
+    {
+      "epoch": 0.6541049798115747,
+      "grad_norm": 1.3433364131880747,
+      "learning_rate": 2.82310319142025e-06,
+      "loss": 0.2089,
+      "step": 4374
+    },
+    {
+      "epoch": 0.6542545237027068,
+      "grad_norm": 1.2350274365674798,
+      "learning_rate": 2.8209232050960105e-06,
+      "loss": 0.2776,
+      "step": 4375
+    },
+    {
+      "epoch": 0.6544040675938388,
+      "grad_norm": 1.3867166787790224,
+      "learning_rate": 2.818743730004188e-06,
+      "loss": 0.2008,
+      "step": 4376
+    },
+    {
+      "epoch": 0.6545536114849708,
+      "grad_norm": 0.941825782121814,
+      "learning_rate": 2.816564766656108e-06,
+      "loss": 0.1858,
+      "step": 4377
+    },
+    {
+      "epoch": 0.6547031553761029,
+      "grad_norm": 1.1183426227174296,
+      "learning_rate": 2.8143863155629793e-06,
+      "loss": 0.3019,
+      "step": 4378
+    },
+    {
+      "epoch": 0.6548526992672349,
+      "grad_norm": 1.9187011951738253,
+      "learning_rate": 2.812208377235881e-06,
+      "loss": 0.3987,
+      "step": 4379
+    },
+    {
+      "epoch": 0.655002243158367,
+      "grad_norm": 1.3958417841195268,
+      "learning_rate": 2.8100309521857825e-06,
+      "loss": 0.3578,
+      "step": 4380
+    },
+    {
+      "epoch": 0.6551517870494991,
+      "grad_norm": 1.1711576507342767,
+      "learning_rate": 2.807854040923529e-06,
+      "loss": 0.1963,
+      "step": 4381
+    },
+    {
+      "epoch": 0.655301330940631,
+      "grad_norm": 0.9913151578185128,
+      "learning_rate": 2.8056776439598433e-06,
+      "loss": 0.1655,
+      "step": 4382
+    },
+    {
+      "epoch": 0.6554508748317631,
+      "grad_norm": 1.5681331704171262,
+      "learning_rate": 2.8035017618053296e-06,
+      "loss": 0.2992,
+      "step": 4383
+    },
+    {
+      "epoch": 0.6556004187228952,
+      "grad_norm": 1.4790018156584064,
+      "learning_rate": 2.8013263949704706e-06,
+      "loss": 0.3136,
+      "step": 4384
+    },
+    {
+      "epoch": 0.6557499626140272,
+      "grad_norm": 0.9994189149888982,
+      "learning_rate": 2.7991515439656297e-06,
+      "loss": 0.1362,
+      "step": 4385
+    },
+    {
+      "epoch": 0.6558995065051593,
+      "grad_norm": 1.3602945298003222,
+      "learning_rate": 2.796977209301047e-06,
+      "loss": 0.1949,
+      "step": 4386
+    },
+    {
+      "epoch": 0.6560490503962914,
+      "grad_norm": 1.5519253212439887,
+      "learning_rate": 2.7948033914868415e-06,
+      "loss": 0.1664,
+      "step": 4387
+    },
+    {
+      "epoch": 0.6561985942874233,
+      "grad_norm": 1.3116382312883463,
+      "learning_rate": 2.792630091033015e-06,
+      "loss": 0.2667,
+      "step": 4388
+    },
+    {
+      "epoch": 0.6563481381785554,
+      "grad_norm": 1.5456359505487205,
+      "learning_rate": 2.7904573084494407e-06,
+      "loss": 0.3245,
+      "step": 4389
+    },
+    {
+      "epoch": 0.6564976820696875,
+      "grad_norm": 1.2913211368683502,
+      "learning_rate": 2.7882850442458755e-06,
+      "loss": 0.3066,
+      "step": 4390
+    },
+    {
+      "epoch": 0.6566472259608195,
+      "grad_norm": 1.869518459083425,
+      "learning_rate": 2.7861132989319544e-06,
+      "loss": 0.3495,
+      "step": 4391
+    },
+    {
+      "epoch": 0.6567967698519516,
+      "grad_norm": 1.373328927428082,
+      "learning_rate": 2.7839420730171895e-06,
+      "loss": 0.2125,
+      "step": 4392
+    },
+    {
+      "epoch": 0.6569463137430835,
+      "grad_norm": 1.8085123954255349,
+      "learning_rate": 2.781771367010973e-06,
+      "loss": 0.2415,
+      "step": 4393
+    },
+    {
+      "epoch": 0.6570958576342156,
+      "grad_norm": 1.0139033929676113,
+      "learning_rate": 2.77960118142257e-06,
+      "loss": 0.1742,
+      "step": 4394
+    },
+    {
+      "epoch": 0.6572454015253477,
+      "grad_norm": 1.685995811270678,
+      "learning_rate": 2.7774315167611276e-06,
+      "loss": 0.3223,
+      "step": 4395
+    },
+    {
+      "epoch": 0.6573949454164797,
+      "grad_norm": 1.7868388236948014,
+      "learning_rate": 2.7752623735356722e-06,
+      "loss": 0.3727,
+      "step": 4396
+    },
+    {
+      "epoch": 0.6575444893076118,
+      "grad_norm": 2.171369150655544,
+      "learning_rate": 2.7730937522551043e-06,
+      "loss": 0.3697,
+      "step": 4397
+    },
+    {
+      "epoch": 0.6576940331987439,
+      "grad_norm": 1.9125192931470543,
+      "learning_rate": 2.7709256534282037e-06,
+      "loss": 0.2111,
+      "step": 4398
+    },
+    {
+      "epoch": 0.6578435770898758,
+      "grad_norm": 1.113585534256139,
+      "learning_rate": 2.768758077563627e-06,
+      "loss": 0.1945,
+      "step": 4399
+    },
+    {
+      "epoch": 0.6579931209810079,
+      "grad_norm": 1.2761501807828817,
+      "learning_rate": 2.7665910251699083e-06,
+      "loss": 0.1971,
+      "step": 4400
+    },
+    {
+      "epoch": 0.65814266487214,
+      "grad_norm": 1.4329874019305284,
+      "learning_rate": 2.7644244967554595e-06,
+      "loss": 0.3153,
+      "step": 4401
+    },
+    {
+      "epoch": 0.658292208763272,
+      "grad_norm": 1.3299910395004908,
+      "learning_rate": 2.762258492828569e-06,
+      "loss": 0.2046,
+      "step": 4402
+    },
+    {
+      "epoch": 0.6584417526544041,
+      "grad_norm": 1.6253732014304119,
+      "learning_rate": 2.7600930138974035e-06,
+      "loss": 0.2748,
+      "step": 4403
+    },
+    {
+      "epoch": 0.6585912965455362,
+      "grad_norm": 1.3081383695661915,
+      "learning_rate": 2.7579280604700025e-06,
+      "loss": 0.2215,
+      "step": 4404
+    },
+    {
+      "epoch": 0.6587408404366681,
+      "grad_norm": 1.562999848046207,
+      "learning_rate": 2.755763633054286e-06,
+      "loss": 0.1758,
+      "step": 4405
+    },
+    {
+      "epoch": 0.6588903843278002,
+      "grad_norm": 1.6911611048983322,
+      "learning_rate": 2.75359973215805e-06,
+      "loss": 0.3255,
+      "step": 4406
+    },
+    {
+      "epoch": 0.6590399282189322,
+      "grad_norm": 1.4640913722306848,
+      "learning_rate": 2.7514363582889653e-06,
+      "loss": 0.2046,
+      "step": 4407
+    },
+    {
+      "epoch": 0.6591894721100643,
+      "grad_norm": 1.218210501349125,
+      "learning_rate": 2.749273511954583e-06,
+      "loss": 0.2054,
+      "step": 4408
+    },
+    {
+      "epoch": 0.6593390160011964,
+      "grad_norm": 1.3940094978870614,
+      "learning_rate": 2.7471111936623253e-06,
+      "loss": 0.2438,
+      "step": 4409
+    },
+    {
+      "epoch": 0.6594885598923284,
+      "grad_norm": 1.3026691632055731,
+      "learning_rate": 2.74494940391949e-06,
+      "loss": 0.2353,
+      "step": 4410
+    },
+    {
+      "epoch": 0.6596381037834604,
+      "grad_norm": 1.6382291329843632,
+      "learning_rate": 2.7427881432332603e-06,
+      "loss": 0.312,
+      "step": 4411
+    },
+    {
+      "epoch": 0.6597876476745925,
+      "grad_norm": 1.5603652939172554,
+      "learning_rate": 2.7406274121106856e-06,
+      "loss": 0.3006,
+      "step": 4412
+    },
+    {
+      "epoch": 0.6599371915657245,
+      "grad_norm": 1.090656104231796,
+      "learning_rate": 2.7384672110586963e-06,
+      "loss": 0.1889,
+      "step": 4413
+    },
+    {
+      "epoch": 0.6600867354568566,
+      "grad_norm": 1.436850245454595,
+      "learning_rate": 2.736307540584092e-06,
+      "loss": 0.2647,
+      "step": 4414
+    },
+    {
+      "epoch": 0.6602362793479887,
+      "grad_norm": 1.7597105163684004,
+      "learning_rate": 2.7341484011935554e-06,
+      "loss": 0.385,
+      "step": 4415
+    },
+    {
+      "epoch": 0.6603858232391207,
+      "grad_norm": 1.2090788451524748,
+      "learning_rate": 2.7319897933936408e-06,
+      "loss": 0.241,
+      "step": 4416
+    },
+    {
+      "epoch": 0.6605353671302527,
+      "grad_norm": 1.7319328710330286,
+      "learning_rate": 2.7298317176907772e-06,
+      "loss": 0.451,
+      "step": 4417
+    },
+    {
+      "epoch": 0.6606849110213848,
+      "grad_norm": 1.5724306339844993,
+      "learning_rate": 2.727674174591274e-06,
+      "loss": 0.2432,
+      "step": 4418
+    },
+    {
+      "epoch": 0.6608344549125168,
+      "grad_norm": 1.9853434639918945,
+      "learning_rate": 2.7255171646013066e-06,
+      "loss": 0.4608,
+      "step": 4419
+    },
+    {
+      "epoch": 0.6609839988036489,
+      "grad_norm": 1.7678752206900543,
+      "learning_rate": 2.7233606882269326e-06,
+      "loss": 0.3429,
+      "step": 4420
+    },
+    {
+      "epoch": 0.661133542694781,
+      "grad_norm": 1.9928878413352948,
+      "learning_rate": 2.721204745974082e-06,
+      "loss": 0.2974,
+      "step": 4421
+    },
+    {
+      "epoch": 0.661283086585913,
+      "grad_norm": 1.5540662511476908,
+      "learning_rate": 2.719049338348559e-06,
+      "loss": 0.2107,
+      "step": 4422
+    },
+    {
+      "epoch": 0.661432630477045,
+      "grad_norm": 1.250271025720372,
+      "learning_rate": 2.7168944658560436e-06,
+      "loss": 0.1951,
+      "step": 4423
+    },
+    {
+      "epoch": 0.661582174368177,
+      "grad_norm": 1.8632502883162096,
+      "learning_rate": 2.7147401290020903e-06,
+      "loss": 0.4555,
+      "step": 4424
+    },
+    {
+      "epoch": 0.6617317182593091,
+      "grad_norm": 1.7615241135628605,
+      "learning_rate": 2.712586328292126e-06,
+      "loss": 0.2952,
+      "step": 4425
+    },
+    {
+      "epoch": 0.6618812621504412,
+      "grad_norm": 1.538421076032742,
+      "learning_rate": 2.7104330642314534e-06,
+      "loss": 0.2333,
+      "step": 4426
+    },
+    {
+      "epoch": 0.6620308060415732,
+      "grad_norm": 1.2007569968926926,
+      "learning_rate": 2.70828033732525e-06,
+      "loss": 0.208,
+      "step": 4427
+    },
+    {
+      "epoch": 0.6621803499327052,
+      "grad_norm": 1.3593872319142528,
+      "learning_rate": 2.7061281480785655e-06,
+      "loss": 0.1821,
+      "step": 4428
+    },
+    {
+      "epoch": 0.6623298938238373,
+      "grad_norm": 1.232105994512284,
+      "learning_rate": 2.703976496996327e-06,
+      "loss": 0.1709,
+      "step": 4429
+    },
+    {
+      "epoch": 0.6624794377149693,
+      "grad_norm": 1.45991562993099,
+      "learning_rate": 2.7018253845833265e-06,
+      "loss": 0.3415,
+      "step": 4430
+    },
+    {
+      "epoch": 0.6626289816061014,
+      "grad_norm": 1.8114343832084139,
+      "learning_rate": 2.6996748113442397e-06,
+      "loss": 0.4451,
+      "step": 4431
+    },
+    {
+      "epoch": 0.6627785254972335,
+      "grad_norm": 1.2612949226382904,
+      "learning_rate": 2.697524777783611e-06,
+      "loss": 0.1998,
+      "step": 4432
+    },
+    {
+      "epoch": 0.6629280693883655,
+      "grad_norm": 1.65296698826448,
+      "learning_rate": 2.69537528440586e-06,
+      "loss": 0.187,
+      "step": 4433
+    },
+    {
+      "epoch": 0.6630776132794975,
+      "grad_norm": 1.8179891682696618,
+      "learning_rate": 2.6932263317152797e-06,
+      "loss": 0.2954,
+      "step": 4434
+    },
+    {
+      "epoch": 0.6632271571706296,
+      "grad_norm": 1.3918087600920022,
+      "learning_rate": 2.691077920216031e-06,
+      "loss": 0.2249,
+      "step": 4435
+    },
+    {
+      "epoch": 0.6633767010617616,
+      "grad_norm": 1.181894833041793,
+      "learning_rate": 2.6889300504121537e-06,
+      "loss": 0.3181,
+      "step": 4436
+    },
+    {
+      "epoch": 0.6635262449528937,
+      "grad_norm": 1.5083352946651556,
+      "learning_rate": 2.6867827228075596e-06,
+      "loss": 0.1869,
+      "step": 4437
+    },
+    {
+      "epoch": 0.6636757888440257,
+      "grad_norm": 1.4057723638218287,
+      "learning_rate": 2.6846359379060317e-06,
+      "loss": 0.2019,
+      "step": 4438
+    },
+    {
+      "epoch": 0.6638253327351578,
+      "grad_norm": 1.5144730730196767,
+      "learning_rate": 2.6824896962112268e-06,
+      "loss": 0.3392,
+      "step": 4439
+    },
+    {
+      "epoch": 0.6639748766262898,
+      "grad_norm": 1.8769019176472341,
+      "learning_rate": 2.680343998226674e-06,
+      "loss": 0.3426,
+      "step": 4440
+    },
+    {
+      "epoch": 0.6641244205174218,
+      "grad_norm": 1.5553252335483767,
+      "learning_rate": 2.678198844455775e-06,
+      "loss": 0.3165,
+      "step": 4441
+    },
+    {
+      "epoch": 0.6642739644085539,
+      "grad_norm": 2.146334878090752,
+      "learning_rate": 2.6760542354018026e-06,
+      "loss": 0.6555,
+      "step": 4442
+    },
+    {
+      "epoch": 0.664423508299686,
+      "grad_norm": 1.3461122126317742,
+      "learning_rate": 2.6739101715679035e-06,
+      "loss": 0.1645,
+      "step": 4443
+    },
+    {
+      "epoch": 0.664573052190818,
+      "grad_norm": 2.2230428081308347,
+      "learning_rate": 2.6717666534570976e-06,
+      "loss": 0.3223,
+      "step": 4444
+    },
+    {
+      "epoch": 0.6647225960819501,
+      "grad_norm": 1.786876248438967,
+      "learning_rate": 2.6696236815722703e-06,
+      "loss": 0.2194,
+      "step": 4445
+    },
+    {
+      "epoch": 0.6648721399730821,
+      "grad_norm": 1.563808365448449,
+      "learning_rate": 2.6674812564161867e-06,
+      "loss": 0.2264,
+      "step": 4446
+    },
+    {
+      "epoch": 0.6650216838642141,
+      "grad_norm": 1.665370120699706,
+      "learning_rate": 2.6653393784914795e-06,
+      "loss": 0.2036,
+      "step": 4447
+    },
+    {
+      "epoch": 0.6651712277553462,
+      "grad_norm": 1.5909818001796514,
+      "learning_rate": 2.663198048300654e-06,
+      "loss": 0.2692,
+      "step": 4448
+    },
+    {
+      "epoch": 0.6653207716464783,
+      "grad_norm": 1.3241563095855466,
+      "learning_rate": 2.6610572663460888e-06,
+      "loss": 0.2268,
+      "step": 4449
+    },
+    {
+      "epoch": 0.6654703155376103,
+      "grad_norm": 1.4147371163461582,
+      "learning_rate": 2.6589170331300284e-06,
+      "loss": 0.2954,
+      "step": 4450
+    },
+    {
+      "epoch": 0.6656198594287424,
+      "grad_norm": 1.290890230215134,
+      "learning_rate": 2.6567773491545916e-06,
+      "loss": 0.1932,
+      "step": 4451
+    },
+    {
+      "epoch": 0.6657694033198743,
+      "grad_norm": 1.328536295316399,
+      "learning_rate": 2.654638214921773e-06,
+      "loss": 0.2945,
+      "step": 4452
+    },
+    {
+      "epoch": 0.6659189472110064,
+      "grad_norm": 1.4691191878880958,
+      "learning_rate": 2.6524996309334326e-06,
+      "loss": 0.1881,
+      "step": 4453
+    },
+    {
+      "epoch": 0.6660684911021385,
+      "grad_norm": 1.9846992941067327,
+      "learning_rate": 2.650361597691305e-06,
+      "loss": 0.2611,
+      "step": 4454
+    },
+    {
+      "epoch": 0.6662180349932705,
+      "grad_norm": 1.1002284284226602,
+      "learning_rate": 2.6482241156969885e-06,
+      "loss": 0.143,
+      "step": 4455
+    },
+    {
+      "epoch": 0.6663675788844026,
+      "grad_norm": 1.6582016063094627,
+      "learning_rate": 2.6460871854519594e-06,
+      "loss": 0.4064,
+      "step": 4456
+    },
+    {
+      "epoch": 0.6665171227755347,
+      "grad_norm": 1.9724808152248507,
+      "learning_rate": 2.643950807457562e-06,
+      "loss": 0.2726,
+      "step": 4457
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 1.6258299157165212,
+      "learning_rate": 2.6418149822150115e-06,
+      "loss": 0.3193,
+      "step": 4458
+    },
+    {
+      "epoch": 0.6668162105577987,
+      "grad_norm": 1.4247436876423234,
+      "learning_rate": 2.6396797102253944e-06,
+      "loss": 0.1605,
+      "step": 4459
+    },
+    {
+      "epoch": 0.6669657544489308,
+      "grad_norm": 1.6721232488411581,
+      "learning_rate": 2.6375449919896627e-06,
+      "loss": 0.3464,
+      "step": 4460
+    },
+    {
+      "epoch": 0.6671152983400628,
+      "grad_norm": 2.9391451297239892,
+      "learning_rate": 2.6354108280086445e-06,
+      "loss": 0.2093,
+      "step": 4461
+    },
+    {
+      "epoch": 0.6672648422311949,
+      "grad_norm": 1.6599264142878403,
+      "learning_rate": 2.6332772187830337e-06,
+      "loss": 0.4938,
+      "step": 4462
+    },
+    {
+      "epoch": 0.667414386122327,
+      "grad_norm": 1.4217088492225147,
+      "learning_rate": 2.6311441648133962e-06,
+      "loss": 0.2503,
+      "step": 4463
+    },
+    {
+      "epoch": 0.6675639300134589,
+      "grad_norm": 1.1416005231658408,
+      "learning_rate": 2.6290116666001676e-06,
+      "loss": 0.1838,
+      "step": 4464
+    },
+    {
+      "epoch": 0.667713473904591,
+      "grad_norm": 1.2704339969898428,
+      "learning_rate": 2.626879724643652e-06,
+      "loss": 0.3511,
+      "step": 4465
+    },
+    {
+      "epoch": 0.667863017795723,
+      "grad_norm": 1.819552430700748,
+      "learning_rate": 2.6247483394440233e-06,
+      "loss": 0.2264,
+      "step": 4466
+    },
+    {
+      "epoch": 0.6680125616868551,
+      "grad_norm": 1.2833183620041992,
+      "learning_rate": 2.622617511501325e-06,
+      "loss": 0.3377,
+      "step": 4467
+    },
+    {
+      "epoch": 0.6681621055779872,
+      "grad_norm": 1.6275093986896139,
+      "learning_rate": 2.6204872413154714e-06,
+      "loss": 0.3522,
+      "step": 4468
+    },
+    {
+      "epoch": 0.6683116494691191,
+      "grad_norm": 1.253423027263163,
+      "learning_rate": 2.618357529386244e-06,
+      "loss": 0.2172,
+      "step": 4469
+    },
+    {
+      "epoch": 0.6684611933602512,
+      "grad_norm": 1.5980094854941072,
+      "learning_rate": 2.6162283762132923e-06,
+      "loss": 0.2653,
+      "step": 4470
+    },
+    {
+      "epoch": 0.6686107372513833,
+      "grad_norm": 1.403119879248096,
+      "learning_rate": 2.6140997822961358e-06,
+      "loss": 0.1532,
+      "step": 4471
+    },
+    {
+      "epoch": 0.6687602811425153,
+      "grad_norm": 1.4562304037028075,
+      "learning_rate": 2.6119717481341642e-06,
+      "loss": 0.2866,
+      "step": 4472
+    },
+    {
+      "epoch": 0.6689098250336474,
+      "grad_norm": 1.6249281060380736,
+      "learning_rate": 2.609844274226635e-06,
+      "loss": 0.3281,
+      "step": 4473
+    },
+    {
+      "epoch": 0.6690593689247795,
+      "grad_norm": 1.5803890747488134,
+      "learning_rate": 2.6077173610726757e-06,
+      "loss": 0.3632,
+      "step": 4474
+    },
+    {
+      "epoch": 0.6692089128159114,
+      "grad_norm": 1.1776060853740318,
+      "learning_rate": 2.6055910091712772e-06,
+      "loss": 0.1556,
+      "step": 4475
+    },
+    {
+      "epoch": 0.6693584567070435,
+      "grad_norm": 1.5261404365716826,
+      "learning_rate": 2.603465219021304e-06,
+      "loss": 0.2003,
+      "step": 4476
+    },
+    {
+      "epoch": 0.6695080005981756,
+      "grad_norm": 1.947084837453816,
+      "learning_rate": 2.601339991121486e-06,
+      "loss": 0.4741,
+      "step": 4477
+    },
+    {
+      "epoch": 0.6696575444893076,
+      "grad_norm": 1.4616592429774347,
+      "learning_rate": 2.599215325970423e-06,
+      "loss": 0.2207,
+      "step": 4478
+    },
+    {
+      "epoch": 0.6698070883804397,
+      "grad_norm": 1.5158163962493627,
+      "learning_rate": 2.5970912240665815e-06,
+      "loss": 0.1717,
+      "step": 4479
+    },
+    {
+      "epoch": 0.6699566322715718,
+      "grad_norm": 1.3272386482349343,
+      "learning_rate": 2.5949676859082973e-06,
+      "loss": 0.2796,
+      "step": 4480
+    },
+    {
+      "epoch": 0.6701061761627037,
+      "grad_norm": 1.1894607731615547,
+      "learning_rate": 2.5928447119937717e-06,
+      "loss": 0.1955,
+      "step": 4481
+    },
+    {
+      "epoch": 0.6702557200538358,
+      "grad_norm": 1.9252345719407216,
+      "learning_rate": 2.590722302821075e-06,
+      "loss": 0.4252,
+      "step": 4482
+    },
+    {
+      "epoch": 0.6704052639449678,
+      "grad_norm": 1.1239347712911045,
+      "learning_rate": 2.5886004588881452e-06,
+      "loss": 0.1884,
+      "step": 4483
+    },
+    {
+      "epoch": 0.6705548078360999,
+      "grad_norm": 1.6982575120283836,
+      "learning_rate": 2.5864791806927904e-06,
+      "loss": 0.3321,
+      "step": 4484
+    },
+    {
+      "epoch": 0.670704351727232,
+      "grad_norm": 1.5015928793800224,
+      "learning_rate": 2.584358468732676e-06,
+      "loss": 0.218,
+      "step": 4485
+    },
+    {
+      "epoch": 0.670853895618364,
+      "grad_norm": 1.4247217029985146,
+      "learning_rate": 2.5822383235053464e-06,
+      "loss": 0.3234,
+      "step": 4486
+    },
+    {
+      "epoch": 0.671003439509496,
+      "grad_norm": 1.3017682120157468,
+      "learning_rate": 2.5801187455082064e-06,
+      "loss": 0.2223,
+      "step": 4487
+    },
+    {
+      "epoch": 0.6711529834006281,
+      "grad_norm": 1.3499705028843545,
+      "learning_rate": 2.577999735238531e-06,
+      "loss": 0.2955,
+      "step": 4488
+    },
+    {
+      "epoch": 0.6713025272917601,
+      "grad_norm": 1.3330957656181952,
+      "learning_rate": 2.5758812931934606e-06,
+      "loss": 0.1952,
+      "step": 4489
+    },
+    {
+      "epoch": 0.6714520711828922,
+      "grad_norm": 1.6714305866356522,
+      "learning_rate": 2.573763419869999e-06,
+      "loss": 0.303,
+      "step": 4490
+    },
+    {
+      "epoch": 0.6716016150740243,
+      "grad_norm": 1.6228094558496275,
+      "learning_rate": 2.5716461157650217e-06,
+      "loss": 0.2104,
+      "step": 4491
+    },
+    {
+      "epoch": 0.6717511589651562,
+      "grad_norm": 1.5380762009085422,
+      "learning_rate": 2.5695293813752664e-06,
+      "loss": 0.1895,
+      "step": 4492
+    },
+    {
+      "epoch": 0.6719007028562883,
+      "grad_norm": 1.615717823601208,
+      "learning_rate": 2.5674132171973444e-06,
+      "loss": 0.449,
+      "step": 4493
+    },
+    {
+      "epoch": 0.6720502467474204,
+      "grad_norm": 1.7981929838198951,
+      "learning_rate": 2.565297623727726e-06,
+      "loss": 0.437,
+      "step": 4494
+    },
+    {
+      "epoch": 0.6721997906385524,
+      "grad_norm": 1.6339389985322836,
+      "learning_rate": 2.5631826014627483e-06,
+      "loss": 0.2037,
+      "step": 4495
+    },
+    {
+      "epoch": 0.6723493345296845,
+      "grad_norm": 1.5686916032689824,
+      "learning_rate": 2.561068150898616e-06,
+      "loss": 0.4212,
+      "step": 4496
+    },
+    {
+      "epoch": 0.6724988784208165,
+      "grad_norm": 1.4778132137047602,
+      "learning_rate": 2.558954272531401e-06,
+      "loss": 0.2207,
+      "step": 4497
+    },
+    {
+      "epoch": 0.6726484223119485,
+      "grad_norm": 2.4730464271902917,
+      "learning_rate": 2.556840966857038e-06,
+      "loss": 0.369,
+      "step": 4498
+    },
+    {
+      "epoch": 0.6727979662030806,
+      "grad_norm": 1.1173307437030382,
+      "learning_rate": 2.554728234371333e-06,
+      "loss": 0.1953,
+      "step": 4499
+    },
+    {
+      "epoch": 0.6729475100942126,
+      "grad_norm": 1.2102968474734468,
+      "learning_rate": 2.5526160755699476e-06,
+      "loss": 0.1725,
+      "step": 4500
+    },
+    {
+      "epoch": 0.6730970539853447,
+      "grad_norm": 1.841005221927744,
+      "learning_rate": 2.5505044909484172e-06,
+      "loss": 0.4732,
+      "step": 4501
+    },
+    {
+      "epoch": 0.6732465978764768,
+      "grad_norm": 1.6515349173791407,
+      "learning_rate": 2.54839348100214e-06,
+      "loss": 0.3197,
+      "step": 4502
+    },
+    {
+      "epoch": 0.6733961417676088,
+      "grad_norm": 1.5072986883603168,
+      "learning_rate": 2.5462830462263787e-06,
+      "loss": 0.4328,
+      "step": 4503
+    },
+    {
+      "epoch": 0.6735456856587408,
+      "grad_norm": 1.880239085336064,
+      "learning_rate": 2.5441731871162633e-06,
+      "loss": 0.476,
+      "step": 4504
+    },
+    {
+      "epoch": 0.6736952295498729,
+      "grad_norm": 1.7936220910991645,
+      "learning_rate": 2.542063904166785e-06,
+      "loss": 0.4094,
+      "step": 4505
+    },
+    {
+      "epoch": 0.6738447734410049,
+      "grad_norm": 1.4488223065722017,
+      "learning_rate": 2.5399551978728033e-06,
+      "loss": 0.2158,
+      "step": 4506
+    },
+    {
+      "epoch": 0.673994317332137,
+      "grad_norm": 1.5085310088771018,
+      "learning_rate": 2.5378470687290397e-06,
+      "loss": 0.1753,
+      "step": 4507
+    },
+    {
+      "epoch": 0.6741438612232691,
+      "grad_norm": 1.517874631059701,
+      "learning_rate": 2.535739517230083e-06,
+      "loss": 0.325,
+      "step": 4508
+    },
+    {
+      "epoch": 0.6742934051144011,
+      "grad_norm": 1.7195395780133773,
+      "learning_rate": 2.5336325438703863e-06,
+      "loss": 0.2393,
+      "step": 4509
+    },
+    {
+      "epoch": 0.6744429490055331,
+      "grad_norm": 1.3387517983694313,
+      "learning_rate": 2.531526149144262e-06,
+      "loss": 0.3497,
+      "step": 4510
+    },
+    {
+      "epoch": 0.6745924928966651,
+      "grad_norm": 1.95315712711614,
+      "learning_rate": 2.5294203335458935e-06,
+      "loss": 0.4593,
+      "step": 4511
+    },
+    {
+      "epoch": 0.6747420367877972,
+      "grad_norm": 0.9559347391135435,
+      "learning_rate": 2.527315097569324e-06,
+      "loss": 0.1639,
+      "step": 4512
+    },
+    {
+      "epoch": 0.6748915806789293,
+      "grad_norm": 1.414032582762148,
+      "learning_rate": 2.5252104417084626e-06,
+      "loss": 0.2025,
+      "step": 4513
+    },
+    {
+      "epoch": 0.6750411245700613,
+      "grad_norm": 1.566845839312385,
+      "learning_rate": 2.523106366457082e-06,
+      "loss": 0.1844,
+      "step": 4514
+    },
+    {
+      "epoch": 0.6751906684611934,
+      "grad_norm": 1.9898672991386208,
+      "learning_rate": 2.521002872308821e-06,
+      "loss": 0.4331,
+      "step": 4515
+    },
+    {
+      "epoch": 0.6753402123523254,
+      "grad_norm": 1.7100230390337947,
+      "learning_rate": 2.518899959757175e-06,
+      "loss": 0.1873,
+      "step": 4516
+    },
+    {
+      "epoch": 0.6754897562434574,
+      "grad_norm": 2.0525844021935935,
+      "learning_rate": 2.51679762929551e-06,
+      "loss": 0.2439,
+      "step": 4517
+    },
+    {
+      "epoch": 0.6756393001345895,
+      "grad_norm": 1.141829150628222,
+      "learning_rate": 2.514695881417052e-06,
+      "loss": 0.1951,
+      "step": 4518
+    },
+    {
+      "epoch": 0.6757888440257216,
+      "grad_norm": 1.8463419499999694,
+      "learning_rate": 2.5125947166148923e-06,
+      "loss": 0.3175,
+      "step": 4519
+    },
+    {
+      "epoch": 0.6759383879168536,
+      "grad_norm": 1.4468038571336173,
+      "learning_rate": 2.5104941353819847e-06,
+      "loss": 0.1468,
+      "step": 4520
+    },
+    {
+      "epoch": 0.6760879318079857,
+      "grad_norm": 1.840830955171049,
+      "learning_rate": 2.5083941382111442e-06,
+      "loss": 0.2451,
+      "step": 4521
+    },
+    {
+      "epoch": 0.6762374756991177,
+      "grad_norm": 1.2525028087295136,
+      "learning_rate": 2.506294725595052e-06,
+      "loss": 0.2028,
+      "step": 4522
+    },
+    {
+      "epoch": 0.6763870195902497,
+      "grad_norm": 1.6825655787031588,
+      "learning_rate": 2.50419589802625e-06,
+      "loss": 0.1967,
+      "step": 4523
+    },
+    {
+      "epoch": 0.6765365634813818,
+      "grad_norm": 2.0206736390111946,
+      "learning_rate": 2.502097655997143e-06,
+      "loss": 0.253,
+      "step": 4524
+    },
+    {
+      "epoch": 0.6766861073725138,
+      "grad_norm": 2.290957099706866,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.2258,
+      "step": 4525
+    },
+    {
+      "epoch": 0.6768356512636459,
+      "grad_norm": 1.687469367349712,
+      "learning_rate": 2.4979029305269507e-06,
+      "loss": 0.4734,
+      "step": 4526
+    },
+    {
+      "epoch": 0.676985195154778,
+      "grad_norm": 3.698097562618829,
+      "learning_rate": 2.495806448069987e-06,
+      "loss": 0.1845,
+      "step": 4527
+    },
+    {
+      "epoch": 0.6771347390459099,
+      "grad_norm": 1.4094663924466104,
+      "learning_rate": 2.493710553120963e-06,
+      "loss": 0.187,
+      "step": 4528
+    },
+    {
+      "epoch": 0.677284282937042,
+      "grad_norm": 1.5924439969538318,
+      "learning_rate": 2.4916152461715987e-06,
+      "loss": 0.3456,
+      "step": 4529
+    },
+    {
+      "epoch": 0.6774338268281741,
+      "grad_norm": 1.717909780950052,
+      "learning_rate": 2.4895205277134733e-06,
+      "loss": 0.1999,
+      "step": 4530
+    },
+    {
+      "epoch": 0.6775833707193061,
+      "grad_norm": 3.204725423929768,
+      "learning_rate": 2.4874263982380242e-06,
+      "loss": 0.2913,
+      "step": 4531
+    },
+    {
+      "epoch": 0.6777329146104382,
+      "grad_norm": 1.98394915749841,
+      "learning_rate": 2.485332858236555e-06,
+      "loss": 0.2699,
+      "step": 4532
+    },
+    {
+      "epoch": 0.6778824585015703,
+      "grad_norm": 2.1405017009782705,
+      "learning_rate": 2.483239908200234e-06,
+      "loss": 0.3382,
+      "step": 4533
+    },
+    {
+      "epoch": 0.6780320023927022,
+      "grad_norm": 1.733333956096687,
+      "learning_rate": 2.4811475486200852e-06,
+      "loss": 0.1584,
+      "step": 4534
+    },
+    {
+      "epoch": 0.6781815462838343,
+      "grad_norm": 1.2434392026777898,
+      "learning_rate": 2.479055779986999e-06,
+      "loss": 0.1823,
+      "step": 4535
+    },
+    {
+      "epoch": 0.6783310901749664,
+      "grad_norm": 1.7956490190481578,
+      "learning_rate": 2.47696460279172e-06,
+      "loss": 0.3989,
+      "step": 4536
+    },
+    {
+      "epoch": 0.6784806340660984,
+      "grad_norm": 2.310256294816532,
+      "learning_rate": 2.4748740175248605e-06,
+      "loss": 0.6357,
+      "step": 4537
+    },
+    {
+      "epoch": 0.6786301779572305,
+      "grad_norm": 1.2356897547091927,
+      "learning_rate": 2.472784024676892e-06,
+      "loss": 0.2037,
+      "step": 4538
+    },
+    {
+      "epoch": 0.6787797218483625,
+      "grad_norm": 1.4315797654938371,
+      "learning_rate": 2.4706946247381468e-06,
+      "loss": 0.2856,
+      "step": 4539
+    },
+    {
+      "epoch": 0.6789292657394945,
+      "grad_norm": 1.9722002314369849,
+      "learning_rate": 2.4686058181988208e-06,
+      "loss": 0.4924,
+      "step": 4540
+    },
+    {
+      "epoch": 0.6790788096306266,
+      "grad_norm": 1.4626799815033218,
+      "learning_rate": 2.4665176055489636e-06,
+      "loss": 0.286,
+      "step": 4541
+    },
+    {
+      "epoch": 0.6792283535217586,
+      "grad_norm": 1.865826625567713,
+      "learning_rate": 2.4644299872784927e-06,
+      "loss": 0.3587,
+      "step": 4542
+    },
+    {
+      "epoch": 0.6793778974128907,
+      "grad_norm": 1.8507709461020339,
+      "learning_rate": 2.4623429638771825e-06,
+      "loss": 0.5021,
+      "step": 4543
+    },
+    {
+      "epoch": 0.6795274413040228,
+      "grad_norm": 1.8053822274235412,
+      "learning_rate": 2.4602565358346696e-06,
+      "loss": 0.3932,
+      "step": 4544
+    },
+    {
+      "epoch": 0.6796769851951547,
+      "grad_norm": 1.96415152215224,
+      "learning_rate": 2.4581707036404494e-06,
+      "loss": 0.2108,
+      "step": 4545
+    },
+    {
+      "epoch": 0.6798265290862868,
+      "grad_norm": 2.1231248520162427,
+      "learning_rate": 2.456085467783879e-06,
+      "loss": 0.3537,
+      "step": 4546
+    },
+    {
+      "epoch": 0.6799760729774189,
+      "grad_norm": 1.658230950938692,
+      "learning_rate": 2.4540008287541746e-06,
+      "loss": 0.3203,
+      "step": 4547
+    },
+    {
+      "epoch": 0.6801256168685509,
+      "grad_norm": 2.1309611630975507,
+      "learning_rate": 2.4519167870404126e-06,
+      "loss": 0.3312,
+      "step": 4548
+    },
+    {
+      "epoch": 0.680275160759683,
+      "grad_norm": 1.670714314146571,
+      "learning_rate": 2.4498333431315287e-06,
+      "loss": 0.1994,
+      "step": 4549
+    },
+    {
+      "epoch": 0.6804247046508151,
+      "grad_norm": 1.5024268565249808,
+      "learning_rate": 2.447750497516321e-06,
+      "loss": 0.2307,
+      "step": 4550
+    },
+    {
+      "epoch": 0.680574248541947,
+      "grad_norm": 2.242283862233074,
+      "learning_rate": 2.445668250683443e-06,
+      "loss": 0.4919,
+      "step": 4551
+    },
+    {
+      "epoch": 0.6807237924330791,
+      "grad_norm": 1.421126923428181,
+      "learning_rate": 2.4435866031214095e-06,
+      "loss": 0.2254,
+      "step": 4552
+    },
+    {
+      "epoch": 0.6808733363242112,
+      "grad_norm": 1.4927463259013674,
+      "learning_rate": 2.4415055553185955e-06,
+      "loss": 0.3069,
+      "step": 4553
+    },
+    {
+      "epoch": 0.6810228802153432,
+      "grad_norm": 1.2998806696923215,
+      "learning_rate": 2.4394251077632364e-06,
+      "loss": 0.1615,
+      "step": 4554
+    },
+    {
+      "epoch": 0.6811724241064753,
+      "grad_norm": 1.146805911490797,
+      "learning_rate": 2.4373452609434255e-06,
+      "loss": 0.1684,
+      "step": 4555
+    },
+    {
+      "epoch": 0.6813219679976072,
+      "grad_norm": 1.7713961216520178,
+      "learning_rate": 2.435266015347112e-06,
+      "loss": 0.4814,
+      "step": 4556
+    },
+    {
+      "epoch": 0.6814715118887393,
+      "grad_norm": 1.4120882911216142,
+      "learning_rate": 2.4331873714621084e-06,
+      "loss": 0.2663,
+      "step": 4557
+    },
+    {
+      "epoch": 0.6816210557798714,
+      "grad_norm": 1.8178671099844292,
+      "learning_rate": 2.431109329776085e-06,
+      "loss": 0.3536,
+      "step": 4558
+    },
+    {
+      "epoch": 0.6817705996710034,
+      "grad_norm": 1.896015073812396,
+      "learning_rate": 2.4290318907765698e-06,
+      "loss": 0.2162,
+      "step": 4559
+    },
+    {
+      "epoch": 0.6819201435621355,
+      "grad_norm": 1.549976332160772,
+      "learning_rate": 2.4269550549509504e-06,
+      "loss": 0.2146,
+      "step": 4560
+    },
+    {
+      "epoch": 0.6820696874532676,
+      "grad_norm": 1.7333003400526625,
+      "learning_rate": 2.424878822786473e-06,
+      "loss": 0.3507,
+      "step": 4561
+    },
+    {
+      "epoch": 0.6822192313443995,
+      "grad_norm": 2.080118048199713,
+      "learning_rate": 2.4228031947702404e-06,
+      "loss": 0.4497,
+      "step": 4562
+    },
+    {
+      "epoch": 0.6823687752355316,
+      "grad_norm": 1.3144787761798393,
+      "learning_rate": 2.420728171389216e-06,
+      "loss": 0.1721,
+      "step": 4563
+    },
+    {
+      "epoch": 0.6825183191266637,
+      "grad_norm": 1.9989255760998235,
+      "learning_rate": 2.41865375313022e-06,
+      "loss": 0.3456,
+      "step": 4564
+    },
+    {
+      "epoch": 0.6826678630177957,
+      "grad_norm": 2.1096027039438416,
+      "learning_rate": 2.4165799404799326e-06,
+      "loss": 0.2227,
+      "step": 4565
+    },
+    {
+      "epoch": 0.6828174069089278,
+      "grad_norm": 1.7404558488234287,
+      "learning_rate": 2.4145067339248872e-06,
+      "loss": 0.1725,
+      "step": 4566
+    },
+    {
+      "epoch": 0.6829669508000599,
+      "grad_norm": 1.5383538100690357,
+      "learning_rate": 2.412434133951479e-06,
+      "loss": 0.1495,
+      "step": 4567
+    },
+    {
+      "epoch": 0.6831164946911918,
+      "grad_norm": 1.8764531226763614,
+      "learning_rate": 2.4103621410459605e-06,
+      "loss": 0.2587,
+      "step": 4568
+    },
+    {
+      "epoch": 0.6832660385823239,
+      "grad_norm": 1.4443894712316598,
+      "learning_rate": 2.4082907556944407e-06,
+      "loss": 0.1567,
+      "step": 4569
+    },
+    {
+      "epoch": 0.6834155824734559,
+      "grad_norm": 2.045835653356828,
+      "learning_rate": 2.4062199783828887e-06,
+      "loss": 0.5029,
+      "step": 4570
+    },
+    {
+      "epoch": 0.683565126364588,
+      "grad_norm": 1.7592933563003443,
+      "learning_rate": 2.4041498095971253e-06,
+      "loss": 0.2877,
+      "step": 4571
+    },
+    {
+      "epoch": 0.6837146702557201,
+      "grad_norm": 1.8238984261708266,
+      "learning_rate": 2.4020802498228333e-06,
+      "loss": 0.5675,
+      "step": 4572
+    },
+    {
+      "epoch": 0.6838642141468521,
+      "grad_norm": 1.2044384462662545,
+      "learning_rate": 2.4000112995455505e-06,
+      "loss": 0.1853,
+      "step": 4573
+    },
+    {
+      "epoch": 0.6840137580379841,
+      "grad_norm": 1.9777223930621288,
+      "learning_rate": 2.397942959250676e-06,
+      "loss": 0.1935,
+      "step": 4574
+    },
+    {
+      "epoch": 0.6841633019291162,
+      "grad_norm": 1.6845231378556524,
+      "learning_rate": 2.395875229423461e-06,
+      "loss": 0.3364,
+      "step": 4575
+    },
+    {
+      "epoch": 0.6843128458202482,
+      "grad_norm": 2.1285112381986346,
+      "learning_rate": 2.393808110549013e-06,
+      "loss": 0.6261,
+      "step": 4576
+    },
+    {
+      "epoch": 0.6844623897113803,
+      "grad_norm": 1.7401607339435936,
+      "learning_rate": 2.3917416031122994e-06,
+      "loss": 0.3315,
+      "step": 4577
+    },
+    {
+      "epoch": 0.6846119336025124,
+      "grad_norm": 1.6390547094896344,
+      "learning_rate": 2.3896757075981415e-06,
+      "loss": 0.3245,
+      "step": 4578
+    },
+    {
+      "epoch": 0.6847614774936444,
+      "grad_norm": 1.6121215254166528,
+      "learning_rate": 2.3876104244912197e-06,
+      "loss": 0.4402,
+      "step": 4579
+    },
+    {
+      "epoch": 0.6849110213847764,
+      "grad_norm": 1.5197600998932719,
+      "learning_rate": 2.3855457542760705e-06,
+      "loss": 0.3134,
+      "step": 4580
+    },
+    {
+      "epoch": 0.6850605652759085,
+      "grad_norm": 1.365469861766198,
+      "learning_rate": 2.3834816974370823e-06,
+      "loss": 0.2025,
+      "step": 4581
+    },
+    {
+      "epoch": 0.6852101091670405,
+      "grad_norm": 2.206570913193162,
+      "learning_rate": 2.381418254458504e-06,
+      "loss": 0.2094,
+      "step": 4582
+    },
+    {
+      "epoch": 0.6853596530581726,
+      "grad_norm": 2.179403006196204,
+      "learning_rate": 2.37935542582444e-06,
+      "loss": 0.2479,
+      "step": 4583
+    },
+    {
+      "epoch": 0.6855091969493046,
+      "grad_norm": 1.2681576531108862,
+      "learning_rate": 2.3772932120188486e-06,
+      "loss": 0.2015,
+      "step": 4584
+    },
+    {
+      "epoch": 0.6856587408404367,
+      "grad_norm": 1.6301619906028688,
+      "learning_rate": 2.3752316135255453e-06,
+      "loss": 0.3426,
+      "step": 4585
+    },
+    {
+      "epoch": 0.6858082847315687,
+      "grad_norm": 1.8368572690743332,
+      "learning_rate": 2.373170630828202e-06,
+      "loss": 0.3368,
+      "step": 4586
+    },
+    {
+      "epoch": 0.6859578286227007,
+      "grad_norm": 1.5036958968449223,
+      "learning_rate": 2.3711102644103447e-06,
+      "loss": 0.2787,
+      "step": 4587
+    },
+    {
+      "epoch": 0.6861073725138328,
+      "grad_norm": 1.576871903212716,
+      "learning_rate": 2.3690505147553537e-06,
+      "loss": 0.326,
+      "step": 4588
+    },
+    {
+      "epoch": 0.6862569164049649,
+      "grad_norm": 2.1317550217444183,
+      "learning_rate": 2.3669913823464685e-06,
+      "loss": 0.3305,
+      "step": 4589
+    },
+    {
+      "epoch": 0.6864064602960969,
+      "grad_norm": 2.0013226069777748,
+      "learning_rate": 2.3649328676667817e-06,
+      "loss": 0.2133,
+      "step": 4590
+    },
+    {
+      "epoch": 0.686556004187229,
+      "grad_norm": 3.0309540515640054,
+      "learning_rate": 2.362874971199237e-06,
+      "loss": 0.5016,
+      "step": 4591
+    },
+    {
+      "epoch": 0.686705548078361,
+      "grad_norm": 1.698566386079889,
+      "learning_rate": 2.3608176934266395e-06,
+      "loss": 0.3094,
+      "step": 4592
+    },
+    {
+      "epoch": 0.686855091969493,
+      "grad_norm": 1.1218626097364182,
+      "learning_rate": 2.358761034831646e-06,
+      "loss": 0.1704,
+      "step": 4593
+    },
+    {
+      "epoch": 0.6870046358606251,
+      "grad_norm": 1.7480987711613902,
+      "learning_rate": 2.356704995896768e-06,
+      "loss": 0.2841,
+      "step": 4594
+    },
+    {
+      "epoch": 0.6871541797517572,
+      "grad_norm": 1.9320402744451117,
+      "learning_rate": 2.3546495771043744e-06,
+      "loss": 0.2722,
+      "step": 4595
+    },
+    {
+      "epoch": 0.6873037236428892,
+      "grad_norm": 1.5760126092513453,
+      "learning_rate": 2.352594778936682e-06,
+      "loss": 0.2132,
+      "step": 4596
+    },
+    {
+      "epoch": 0.6874532675340213,
+      "grad_norm": 1.504536632845543,
+      "learning_rate": 2.3505406018757694e-06,
+      "loss": 0.3034,
+      "step": 4597
+    },
+    {
+      "epoch": 0.6876028114251533,
+      "grad_norm": 2.0882148205501943,
+      "learning_rate": 2.348487046403564e-06,
+      "loss": 0.4912,
+      "step": 4598
+    },
+    {
+      "epoch": 0.6877523553162853,
+      "grad_norm": 1.6574096421206816,
+      "learning_rate": 2.3464341130018525e-06,
+      "loss": 0.1946,
+      "step": 4599
+    },
+    {
+      "epoch": 0.6879018992074174,
+      "grad_norm": 1.625508448944616,
+      "learning_rate": 2.344381802152271e-06,
+      "loss": 0.2829,
+      "step": 4600
+    },
+    {
+      "epoch": 0.6880514430985494,
+      "grad_norm": 2.006732033965754,
+      "learning_rate": 2.3423301143363112e-06,
+      "loss": 0.4434,
+      "step": 4601
+    },
+    {
+      "epoch": 0.6882009869896815,
+      "grad_norm": 1.2939033760616547,
+      "learning_rate": 2.34027905003532e-06,
+      "loss": 0.1808,
+      "step": 4602
+    },
+    {
+      "epoch": 0.6883505308808135,
+      "grad_norm": 1.3243865635408918,
+      "learning_rate": 2.3382286097304963e-06,
+      "loss": 0.17,
+      "step": 4603
+    },
+    {
+      "epoch": 0.6885000747719455,
+      "grad_norm": 1.999281798990057,
+      "learning_rate": 2.3361787939028933e-06,
+      "loss": 0.471,
+      "step": 4604
+    },
+    {
+      "epoch": 0.6886496186630776,
+      "grad_norm": 1.7744976857477415,
+      "learning_rate": 2.3341296030334177e-06,
+      "loss": 0.3231,
+      "step": 4605
+    },
+    {
+      "epoch": 0.6887991625542097,
+      "grad_norm": 1.6893906333699582,
+      "learning_rate": 2.3320810376028306e-06,
+      "loss": 0.1839,
+      "step": 4606
+    },
+    {
+      "epoch": 0.6889487064453417,
+      "grad_norm": 1.7226376910950774,
+      "learning_rate": 2.3300330980917423e-06,
+      "loss": 0.3491,
+      "step": 4607
+    },
+    {
+      "epoch": 0.6890982503364738,
+      "grad_norm": 1.4118718729748658,
+      "learning_rate": 2.3279857849806203e-06,
+      "loss": 0.3047,
+      "step": 4608
+    },
+    {
+      "epoch": 0.6892477942276058,
+      "grad_norm": 1.6701895872877535,
+      "learning_rate": 2.325939098749785e-06,
+      "loss": 0.2104,
+      "step": 4609
+    },
+    {
+      "epoch": 0.6893973381187378,
+      "grad_norm": 1.1844959534454789,
+      "learning_rate": 2.3238930398794067e-06,
+      "loss": 0.1852,
+      "step": 4610
+    },
+    {
+      "epoch": 0.6895468820098699,
+      "grad_norm": 1.7090626464358218,
+      "learning_rate": 2.321847608849515e-06,
+      "loss": 0.2402,
+      "step": 4611
+    },
+    {
+      "epoch": 0.689696425901002,
+      "grad_norm": 1.7427682977617915,
+      "learning_rate": 2.3198028061399824e-06,
+      "loss": 0.3506,
+      "step": 4612
+    },
+    {
+      "epoch": 0.689845969792134,
+      "grad_norm": 1.8109575667927527,
+      "learning_rate": 2.317758632230541e-06,
+      "loss": 0.3346,
+      "step": 4613
+    },
+    {
+      "epoch": 0.6899955136832661,
+      "grad_norm": 2.2380666713150004,
+      "learning_rate": 2.315715087600773e-06,
+      "loss": 0.5128,
+      "step": 4614
+    },
+    {
+      "epoch": 0.690145057574398,
+      "grad_norm": 1.5943030258047348,
+      "learning_rate": 2.3136721727301164e-06,
+      "loss": 0.3733,
+      "step": 4615
+    },
+    {
+      "epoch": 0.6902946014655301,
+      "grad_norm": 1.609221048306236,
+      "learning_rate": 2.3116298880978594e-06,
+      "loss": 0.3455,
+      "step": 4616
+    },
+    {
+      "epoch": 0.6904441453566622,
+      "grad_norm": 1.6369844098138138,
+      "learning_rate": 2.309588234183137e-06,
+      "loss": 0.2209,
+      "step": 4617
+    },
+    {
+      "epoch": 0.6905936892477942,
+      "grad_norm": 1.2426055048948337,
+      "learning_rate": 2.3075472114649438e-06,
+      "loss": 0.1828,
+      "step": 4618
+    },
+    {
+      "epoch": 0.6907432331389263,
+      "grad_norm": 1.1641756793179936,
+      "learning_rate": 2.3055068204221226e-06,
+      "loss": 0.1906,
+      "step": 4619
+    },
+    {
+      "epoch": 0.6908927770300584,
+      "grad_norm": 1.5002027327407788,
+      "learning_rate": 2.3034670615333693e-06,
+      "loss": 0.2173,
+      "step": 4620
+    },
+    {
+      "epoch": 0.6910423209211903,
+      "grad_norm": 1.7566844796442385,
+      "learning_rate": 2.301427935277233e-06,
+      "loss": 0.3071,
+      "step": 4621
+    },
+    {
+      "epoch": 0.6911918648123224,
+      "grad_norm": 1.1780204836628183,
+      "learning_rate": 2.299389442132108e-06,
+      "loss": 0.217,
+      "step": 4622
+    },
+    {
+      "epoch": 0.6913414087034545,
+      "grad_norm": 1.1814865815819897,
+      "learning_rate": 2.2973515825762464e-06,
+      "loss": 0.2224,
+      "step": 4623
+    },
+    {
+      "epoch": 0.6914909525945865,
+      "grad_norm": 1.6353908218378568,
+      "learning_rate": 2.2953143570877507e-06,
+      "loss": 0.1858,
+      "step": 4624
+    },
+    {
+      "epoch": 0.6916404964857186,
+      "grad_norm": 1.8174800119127352,
+      "learning_rate": 2.2932777661445726e-06,
+      "loss": 0.2081,
+      "step": 4625
+    },
+    {
+      "epoch": 0.6917900403768507,
+      "grad_norm": 1.4736221292865608,
+      "learning_rate": 2.291241810224516e-06,
+      "loss": 0.2021,
+      "step": 4626
+    },
+    {
+      "epoch": 0.6919395842679826,
+      "grad_norm": 1.5436710352914425,
+      "learning_rate": 2.289206489805236e-06,
+      "loss": 0.3133,
+      "step": 4627
+    },
+    {
+      "epoch": 0.6920891281591147,
+      "grad_norm": 1.6060115936001564,
+      "learning_rate": 2.287171805364238e-06,
+      "loss": 0.2044,
+      "step": 4628
+    },
+    {
+      "epoch": 0.6922386720502467,
+      "grad_norm": 1.2741827188445998,
+      "learning_rate": 2.2851377573788795e-06,
+      "loss": 0.3343,
+      "step": 4629
+    },
+    {
+      "epoch": 0.6923882159413788,
+      "grad_norm": 1.2549708146286533,
+      "learning_rate": 2.2831043463263674e-06,
+      "loss": 0.1389,
+      "step": 4630
+    },
+    {
+      "epoch": 0.6925377598325109,
+      "grad_norm": 1.4200924583299743,
+      "learning_rate": 2.2810715726837613e-06,
+      "loss": 0.205,
+      "step": 4631
+    },
+    {
+      "epoch": 0.6926873037236428,
+      "grad_norm": 1.1966057167582587,
+      "learning_rate": 2.2790394369279657e-06,
+      "loss": 0.1924,
+      "step": 4632
+    },
+    {
+      "epoch": 0.6928368476147749,
+      "grad_norm": 2.0421745842756533,
+      "learning_rate": 2.2770079395357404e-06,
+      "loss": 0.4616,
+      "step": 4633
+    },
+    {
+      "epoch": 0.692986391505907,
+      "grad_norm": 1.3679535926488904,
+      "learning_rate": 2.274977080983695e-06,
+      "loss": 0.1438,
+      "step": 4634
+    },
+    {
+      "epoch": 0.693135935397039,
+      "grad_norm": 1.4377322216709112,
+      "learning_rate": 2.272946861748289e-06,
+      "loss": 0.369,
+      "step": 4635
+    },
+    {
+      "epoch": 0.6932854792881711,
+      "grad_norm": 1.7226883891734146,
+      "learning_rate": 2.270917282305833e-06,
+      "loss": 0.3433,
+      "step": 4636
+    },
+    {
+      "epoch": 0.6934350231793032,
+      "grad_norm": 1.2758073680864919,
+      "learning_rate": 2.2688883431324808e-06,
+      "loss": 0.229,
+      "step": 4637
+    },
+    {
+      "epoch": 0.6935845670704351,
+      "grad_norm": 1.2156104155599643,
+      "learning_rate": 2.266860044704245e-06,
+      "loss": 0.1971,
+      "step": 4638
+    },
+    {
+      "epoch": 0.6937341109615672,
+      "grad_norm": 1.7207331660199756,
+      "learning_rate": 2.2648323874969824e-06,
+      "loss": 0.423,
+      "step": 4639
+    },
+    {
+      "epoch": 0.6938836548526993,
+      "grad_norm": 1.3030113714977887,
+      "learning_rate": 2.262805371986402e-06,
+      "loss": 0.2408,
+      "step": 4640
+    },
+    {
+      "epoch": 0.6940331987438313,
+      "grad_norm": 1.4113844940770655,
+      "learning_rate": 2.26077899864806e-06,
+      "loss": 0.2106,
+      "step": 4641
+    },
+    {
+      "epoch": 0.6941827426349634,
+      "grad_norm": 2.0974485482758363,
+      "learning_rate": 2.2587532679573644e-06,
+      "loss": 0.4824,
+      "step": 4642
+    },
+    {
+      "epoch": 0.6943322865260954,
+      "grad_norm": 1.531885185651406,
+      "learning_rate": 2.2567281803895696e-06,
+      "loss": 0.3369,
+      "step": 4643
+    },
+    {
+      "epoch": 0.6944818304172274,
+      "grad_norm": 1.4252330509985323,
+      "learning_rate": 2.2547037364197825e-06,
+      "loss": 0.3347,
+      "step": 4644
+    },
+    {
+      "epoch": 0.6946313743083595,
+      "grad_norm": 1.702316971805034,
+      "learning_rate": 2.252679936522956e-06,
+      "loss": 0.3386,
+      "step": 4645
+    },
+    {
+      "epoch": 0.6947809181994915,
+      "grad_norm": 1.3719010462493337,
+      "learning_rate": 2.250656781173895e-06,
+      "loss": 0.3806,
+      "step": 4646
+    },
+    {
+      "epoch": 0.6949304620906236,
+      "grad_norm": 1.3372620268377342,
+      "learning_rate": 2.248634270847248e-06,
+      "loss": 0.3181,
+      "step": 4647
+    },
+    {
+      "epoch": 0.6950800059817557,
+      "grad_norm": 1.6909543197914825,
+      "learning_rate": 2.246612406017517e-06,
+      "loss": 0.3417,
+      "step": 4648
+    },
+    {
+      "epoch": 0.6952295498728877,
+      "grad_norm": 1.8885107671906933,
+      "learning_rate": 2.2445911871590507e-06,
+      "loss": 0.2214,
+      "step": 4649
+    },
+    {
+      "epoch": 0.6953790937640197,
+      "grad_norm": 1.9377077129756617,
+      "learning_rate": 2.242570614746047e-06,
+      "loss": 0.2576,
+      "step": 4650
+    },
+    {
+      "epoch": 0.6955286376551518,
+      "grad_norm": 1.2269989622068174,
+      "learning_rate": 2.240550689252553e-06,
+      "loss": 0.1895,
+      "step": 4651
+    },
+    {
+      "epoch": 0.6956781815462838,
+      "grad_norm": 1.7420057321674778,
+      "learning_rate": 2.238531411152459e-06,
+      "loss": 0.3922,
+      "step": 4652
+    },
+    {
+      "epoch": 0.6958277254374159,
+      "grad_norm": 1.2689943322625534,
+      "learning_rate": 2.2365127809195096e-06,
+      "loss": 0.2144,
+      "step": 4653
+    },
+    {
+      "epoch": 0.695977269328548,
+      "grad_norm": 1.4521893544900446,
+      "learning_rate": 2.234494799027293e-06,
+      "loss": 0.3224,
+      "step": 4654
+    },
+    {
+      "epoch": 0.69612681321968,
+      "grad_norm": 1.3816305243737939,
+      "learning_rate": 2.2324774659492505e-06,
+      "loss": 0.2161,
+      "step": 4655
+    },
+    {
+      "epoch": 0.696276357110812,
+      "grad_norm": 1.7389754977060294,
+      "learning_rate": 2.230460782158668e-06,
+      "loss": 0.3868,
+      "step": 4656
+    },
+    {
+      "epoch": 0.6964259010019441,
+      "grad_norm": 1.1403573388410169,
+      "learning_rate": 2.2284447481286747e-06,
+      "loss": 0.157,
+      "step": 4657
+    },
+    {
+      "epoch": 0.6965754448930761,
+      "grad_norm": 1.6258223910169618,
+      "learning_rate": 2.2264293643322544e-06,
+      "loss": 0.2441,
+      "step": 4658
+    },
+    {
+      "epoch": 0.6967249887842082,
+      "grad_norm": 1.4946074012223682,
+      "learning_rate": 2.224414631242235e-06,
+      "loss": 0.2575,
+      "step": 4659
+    },
+    {
+      "epoch": 0.6968745326753402,
+      "grad_norm": 1.9173994792778464,
+      "learning_rate": 2.2224005493312922e-06,
+      "loss": 0.181,
+      "step": 4660
+    },
+    {
+      "epoch": 0.6970240765664723,
+      "grad_norm": 1.3932682055651664,
+      "learning_rate": 2.220387119071951e-06,
+      "loss": 0.1464,
+      "step": 4661
+    },
+    {
+      "epoch": 0.6971736204576043,
+      "grad_norm": 1.779921479622985,
+      "learning_rate": 2.2183743409365786e-06,
+      "loss": 0.4004,
+      "step": 4662
+    },
+    {
+      "epoch": 0.6973231643487363,
+      "grad_norm": 1.6183054740781828,
+      "learning_rate": 2.216362215397393e-06,
+      "loss": 0.2814,
+      "step": 4663
+    },
+    {
+      "epoch": 0.6974727082398684,
+      "grad_norm": 1.8066594550723725,
+      "learning_rate": 2.214350742926459e-06,
+      "loss": 0.1958,
+      "step": 4664
+    },
+    {
+      "epoch": 0.6976222521310005,
+      "grad_norm": 1.2700224996999203,
+      "learning_rate": 2.2123399239956864e-06,
+      "loss": 0.1792,
+      "step": 4665
+    },
+    {
+      "epoch": 0.6977717960221325,
+      "grad_norm": 1.31888606535632,
+      "learning_rate": 2.2103297590768334e-06,
+      "loss": 0.2179,
+      "step": 4666
+    },
+    {
+      "epoch": 0.6979213399132645,
+      "grad_norm": 1.4998521334563557,
+      "learning_rate": 2.2083202486415045e-06,
+      "loss": 0.4194,
+      "step": 4667
+    },
+    {
+      "epoch": 0.6980708838043966,
+      "grad_norm": 1.3738896535935858,
+      "learning_rate": 2.206311393161149e-06,
+      "loss": 0.1909,
+      "step": 4668
+    },
+    {
+      "epoch": 0.6982204276955286,
+      "grad_norm": 1.6437393216232281,
+      "learning_rate": 2.2043031931070652e-06,
+      "loss": 0.1955,
+      "step": 4669
+    },
+    {
+      "epoch": 0.6983699715866607,
+      "grad_norm": 1.4681419448028947,
+      "learning_rate": 2.202295648950395e-06,
+      "loss": 0.2907,
+      "step": 4670
+    },
+    {
+      "epoch": 0.6985195154777928,
+      "grad_norm": 1.6154368318665644,
+      "learning_rate": 2.20028876116213e-06,
+      "loss": 0.3833,
+      "step": 4671
+    },
+    {
+      "epoch": 0.6986690593689248,
+      "grad_norm": 1.4243786745580567,
+      "learning_rate": 2.1982825302131018e-06,
+      "loss": 0.1742,
+      "step": 4672
+    },
+    {
+      "epoch": 0.6988186032600568,
+      "grad_norm": 1.6890524438669836,
+      "learning_rate": 2.1962769565739926e-06,
+      "loss": 0.2357,
+      "step": 4673
+    },
+    {
+      "epoch": 0.6989681471511888,
+      "grad_norm": 1.19650494243045,
+      "learning_rate": 2.1942720407153305e-06,
+      "loss": 0.1925,
+      "step": 4674
+    },
+    {
+      "epoch": 0.6991176910423209,
+      "grad_norm": 1.3446356466305722,
+      "learning_rate": 2.1922677831074875e-06,
+      "loss": 0.1953,
+      "step": 4675
+    },
+    {
+      "epoch": 0.699267234933453,
+      "grad_norm": 1.65371932269424,
+      "learning_rate": 2.1902641842206827e-06,
+      "loss": 0.1712,
+      "step": 4676
+    },
+    {
+      "epoch": 0.699416778824585,
+      "grad_norm": 1.6193435228911113,
+      "learning_rate": 2.1882612445249778e-06,
+      "loss": 0.1903,
+      "step": 4677
+    },
+    {
+      "epoch": 0.6995663227157171,
+      "grad_norm": 1.8691890475100623,
+      "learning_rate": 2.1862589644902825e-06,
+      "loss": 0.4669,
+      "step": 4678
+    },
+    {
+      "epoch": 0.6997158666068491,
+      "grad_norm": 3.966671376799878,
+      "learning_rate": 2.184257344586351e-06,
+      "loss": 0.3711,
+      "step": 4679
+    },
+    {
+      "epoch": 0.6998654104979811,
+      "grad_norm": 1.525153768127254,
+      "learning_rate": 2.1822563852827827e-06,
+      "loss": 0.2788,
+      "step": 4680
+    },
+    {
+      "epoch": 0.7000149543891132,
+      "grad_norm": 1.4895632771384457,
+      "learning_rate": 2.1802560870490226e-06,
+      "loss": 0.2301,
+      "step": 4681
+    },
+    {
+      "epoch": 0.7001644982802453,
+      "grad_norm": 1.4192521773579143,
+      "learning_rate": 2.178256450354359e-06,
+      "loss": 0.2622,
+      "step": 4682
+    },
+    {
+      "epoch": 0.7003140421713773,
+      "grad_norm": 1.9599235390448408,
+      "learning_rate": 2.176257475667926e-06,
+      "loss": 0.3301,
+      "step": 4683
+    },
+    {
+      "epoch": 0.7004635860625094,
+      "grad_norm": 1.7477260785398092,
+      "learning_rate": 2.174259163458703e-06,
+      "loss": 0.3082,
+      "step": 4684
+    },
+    {
+      "epoch": 0.7006131299536414,
+      "grad_norm": 1.5200763401354904,
+      "learning_rate": 2.172261514195513e-06,
+      "loss": 0.194,
+      "step": 4685
+    },
+    {
+      "epoch": 0.7007626738447734,
+      "grad_norm": 1.7098216420009527,
+      "learning_rate": 2.1702645283470238e-06,
+      "loss": 0.347,
+      "step": 4686
+    },
+    {
+      "epoch": 0.7009122177359055,
+      "grad_norm": 2.0044185139827784,
+      "learning_rate": 2.168268206381749e-06,
+      "loss": 0.4704,
+      "step": 4687
+    },
+    {
+      "epoch": 0.7010617616270375,
+      "grad_norm": 1.779151039553705,
+      "learning_rate": 2.1662725487680415e-06,
+      "loss": 0.1837,
+      "step": 4688
+    },
+    {
+      "epoch": 0.7012113055181696,
+      "grad_norm": 1.6643292477429161,
+      "learning_rate": 2.164277555974104e-06,
+      "loss": 0.1878,
+      "step": 4689
+    },
+    {
+      "epoch": 0.7013608494093017,
+      "grad_norm": 2.9446504668455784,
+      "learning_rate": 2.162283228467981e-06,
+      "loss": 0.2075,
+      "step": 4690
+    },
+    {
+      "epoch": 0.7015103933004336,
+      "grad_norm": 1.7449397256765202,
+      "learning_rate": 2.16028956671756e-06,
+      "loss": 0.2779,
+      "step": 4691
+    },
+    {
+      "epoch": 0.7016599371915657,
+      "grad_norm": 1.4132965394938541,
+      "learning_rate": 2.158296571190576e-06,
+      "loss": 0.2747,
+      "step": 4692
+    },
+    {
+      "epoch": 0.7018094810826978,
+      "grad_norm": 1.6241527925634025,
+      "learning_rate": 2.1563042423546014e-06,
+      "loss": 0.3598,
+      "step": 4693
+    },
+    {
+      "epoch": 0.7019590249738298,
+      "grad_norm": 1.9481740533296699,
+      "learning_rate": 2.154312580677057e-06,
+      "loss": 0.3221,
+      "step": 4694
+    },
+    {
+      "epoch": 0.7021085688649619,
+      "grad_norm": 2.0935699791166966,
+      "learning_rate": 2.1523215866252046e-06,
+      "loss": 0.3407,
+      "step": 4695
+    },
+    {
+      "epoch": 0.702258112756094,
+      "grad_norm": 1.6559635872881795,
+      "learning_rate": 2.1503312606661536e-06,
+      "loss": 0.319,
+      "step": 4696
+    },
+    {
+      "epoch": 0.7024076566472259,
+      "grad_norm": 1.179897182347415,
+      "learning_rate": 2.148341603266854e-06,
+      "loss": 0.1896,
+      "step": 4697
+    },
+    {
+      "epoch": 0.702557200538358,
+      "grad_norm": 1.4572695161455325,
+      "learning_rate": 2.146352614894095e-06,
+      "loss": 0.3342,
+      "step": 4698
+    },
+    {
+      "epoch": 0.7027067444294901,
+      "grad_norm": 1.5720481959154646,
+      "learning_rate": 2.1443642960145146e-06,
+      "loss": 0.3083,
+      "step": 4699
+    },
+    {
+      "epoch": 0.7028562883206221,
+      "grad_norm": 1.1533343663921902,
+      "learning_rate": 2.1423766470945904e-06,
+      "loss": 0.1564,
+      "step": 4700
+    },
+    {
+      "epoch": 0.7030058322117542,
+      "grad_norm": 1.3098024593490292,
+      "learning_rate": 2.1403896686006455e-06,
+      "loss": 0.1676,
+      "step": 4701
+    },
+    {
+      "epoch": 0.7031553761028861,
+      "grad_norm": 1.6963395027211619,
+      "learning_rate": 2.1384033609988446e-06,
+      "loss": 0.3673,
+      "step": 4702
+    },
+    {
+      "epoch": 0.7033049199940182,
+      "grad_norm": 1.3604451165819607,
+      "learning_rate": 2.136417724755192e-06,
+      "loss": 0.1528,
+      "step": 4703
+    },
+    {
+      "epoch": 0.7034544638851503,
+      "grad_norm": 1.701125090664159,
+      "learning_rate": 2.1344327603355386e-06,
+      "loss": 0.4854,
+      "step": 4704
+    },
+    {
+      "epoch": 0.7036040077762823,
+      "grad_norm": 2.0597156312410765,
+      "learning_rate": 2.132448468205576e-06,
+      "loss": 0.3941,
+      "step": 4705
+    },
+    {
+      "epoch": 0.7037535516674144,
+      "grad_norm": 1.9368210495078715,
+      "learning_rate": 2.1304648488308383e-06,
+      "loss": 0.4007,
+      "step": 4706
+    },
+    {
+      "epoch": 0.7039030955585465,
+      "grad_norm": 1.6767915245759573,
+      "learning_rate": 2.1284819026767016e-06,
+      "loss": 0.2381,
+      "step": 4707
+    },
+    {
+      "epoch": 0.7040526394496784,
+      "grad_norm": 1.3144913337675048,
+      "learning_rate": 2.126499630208385e-06,
+      "loss": 0.2306,
+      "step": 4708
+    },
+    {
+      "epoch": 0.7042021833408105,
+      "grad_norm": 1.5820996540492964,
+      "learning_rate": 2.1245180318909482e-06,
+      "loss": 0.2811,
+      "step": 4709
+    },
+    {
+      "epoch": 0.7043517272319426,
+      "grad_norm": 2.050004263326817,
+      "learning_rate": 2.1225371081892927e-06,
+      "loss": 0.5782,
+      "step": 4710
+    },
+    {
+      "epoch": 0.7045012711230746,
+      "grad_norm": 1.8213512788429163,
+      "learning_rate": 2.120556859568163e-06,
+      "loss": 0.5186,
+      "step": 4711
+    },
+    {
+      "epoch": 0.7046508150142067,
+      "grad_norm": 1.4587002417536927,
+      "learning_rate": 2.118577286492146e-06,
+      "loss": 0.2573,
+      "step": 4712
+    },
+    {
+      "epoch": 0.7048003589053388,
+      "grad_norm": 1.8259926937548316,
+      "learning_rate": 2.1165983894256647e-06,
+      "loss": 0.3384,
+      "step": 4713
+    },
+    {
+      "epoch": 0.7049499027964707,
+      "grad_norm": 1.5961581880480131,
+      "learning_rate": 2.1146201688329904e-06,
+      "loss": 0.2016,
+      "step": 4714
+    },
+    {
+      "epoch": 0.7050994466876028,
+      "grad_norm": 1.23548395103365,
+      "learning_rate": 2.1126426251782317e-06,
+      "loss": 0.2576,
+      "step": 4715
+    },
+    {
+      "epoch": 0.7052489905787349,
+      "grad_norm": 1.5061958421966601,
+      "learning_rate": 2.1106657589253395e-06,
+      "loss": 0.185,
+      "step": 4716
+    },
+    {
+      "epoch": 0.7053985344698669,
+      "grad_norm": 1.8046523829754955,
+      "learning_rate": 2.1086895705381076e-06,
+      "loss": 0.5278,
+      "step": 4717
+    },
+    {
+      "epoch": 0.705548078360999,
+      "grad_norm": 1.4351344755345183,
+      "learning_rate": 2.106714060480165e-06,
+      "loss": 0.2583,
+      "step": 4718
+    },
+    {
+      "epoch": 0.705697622252131,
+      "grad_norm": 1.5501057188835787,
+      "learning_rate": 2.1047392292149882e-06,
+      "loss": 0.2195,
+      "step": 4719
+    },
+    {
+      "epoch": 0.705847166143263,
+      "grad_norm": 1.5774802245738384,
+      "learning_rate": 2.1027650772058907e-06,
+      "loss": 0.1787,
+      "step": 4720
+    },
+    {
+      "epoch": 0.7059967100343951,
+      "grad_norm": 1.8231536892793683,
+      "learning_rate": 2.100791604916027e-06,
+      "loss": 0.3842,
+      "step": 4721
+    },
+    {
+      "epoch": 0.7061462539255271,
+      "grad_norm": 1.6854332833675343,
+      "learning_rate": 2.098818812808394e-06,
+      "loss": 0.1943,
+      "step": 4722
+    },
+    {
+      "epoch": 0.7062957978166592,
+      "grad_norm": 1.6958455069696248,
+      "learning_rate": 2.0968467013458278e-06,
+      "loss": 0.3186,
+      "step": 4723
+    },
+    {
+      "epoch": 0.7064453417077913,
+      "grad_norm": 1.4747719150923775,
+      "learning_rate": 2.0948752709910035e-06,
+      "loss": 0.2034,
+      "step": 4724
+    },
+    {
+      "epoch": 0.7065948855989233,
+      "grad_norm": 1.9794588768390642,
+      "learning_rate": 2.092904522206439e-06,
+      "loss": 0.324,
+      "step": 4725
+    },
+    {
+      "epoch": 0.7067444294900553,
+      "grad_norm": 1.6232444055400037,
+      "learning_rate": 2.09093445545449e-06,
+      "loss": 0.3518,
+      "step": 4726
+    },
+    {
+      "epoch": 0.7068939733811874,
+      "grad_norm": 2.2386200267330305,
+      "learning_rate": 2.088965071197355e-06,
+      "loss": 0.2243,
+      "step": 4727
+    },
+    {
+      "epoch": 0.7070435172723194,
+      "grad_norm": 1.5936315997405437,
+      "learning_rate": 2.0869963698970676e-06,
+      "loss": 0.1983,
+      "step": 4728
+    },
+    {
+      "epoch": 0.7071930611634515,
+      "grad_norm": 1.4775550208581971,
+      "learning_rate": 2.0850283520155052e-06,
+      "loss": 0.1618,
+      "step": 4729
+    },
+    {
+      "epoch": 0.7073426050545836,
+      "grad_norm": 1.9080994069269488,
+      "learning_rate": 2.0830610180143846e-06,
+      "loss": 0.4301,
+      "step": 4730
+    },
+    {
+      "epoch": 0.7074921489457155,
+      "grad_norm": 1.4333880898039384,
+      "learning_rate": 2.08109436835526e-06,
+      "loss": 0.3574,
+      "step": 4731
+    },
+    {
+      "epoch": 0.7076416928368476,
+      "grad_norm": 1.851506015060867,
+      "learning_rate": 2.0791284034995296e-06,
+      "loss": 0.418,
+      "step": 4732
+    },
+    {
+      "epoch": 0.7077912367279796,
+      "grad_norm": 1.9375351779574483,
+      "learning_rate": 2.0771631239084233e-06,
+      "loss": 0.4139,
+      "step": 4733
+    },
+    {
+      "epoch": 0.7079407806191117,
+      "grad_norm": 1.554772005707828,
+      "learning_rate": 2.075198530043016e-06,
+      "loss": 0.1786,
+      "step": 4734
+    },
+    {
+      "epoch": 0.7080903245102438,
+      "grad_norm": 1.75390264769866,
+      "learning_rate": 2.07323462236422e-06,
+      "loss": 0.4727,
+      "step": 4735
+    },
+    {
+      "epoch": 0.7082398684013758,
+      "grad_norm": 1.3694940891653622,
+      "learning_rate": 2.071271401332789e-06,
+      "loss": 0.4066,
+      "step": 4736
+    },
+    {
+      "epoch": 0.7083894122925078,
+      "grad_norm": 3.4800517006814067,
+      "learning_rate": 2.0693088674093146e-06,
+      "loss": 0.3289,
+      "step": 4737
+    },
+    {
+      "epoch": 0.7085389561836399,
+      "grad_norm": 1.7965126916254985,
+      "learning_rate": 2.0673470210542223e-06,
+      "loss": 0.4228,
+      "step": 4738
+    },
+    {
+      "epoch": 0.7086885000747719,
+      "grad_norm": 2.123641224815276,
+      "learning_rate": 2.0653858627277816e-06,
+      "loss": 0.2414,
+      "step": 4739
+    },
+    {
+      "epoch": 0.708838043965904,
+      "grad_norm": 1.7814274017768608,
+      "learning_rate": 2.0634253928900997e-06,
+      "loss": 0.1977,
+      "step": 4740
+    },
+    {
+      "epoch": 0.7089875878570361,
+      "grad_norm": 1.634731670651409,
+      "learning_rate": 2.0614656120011217e-06,
+      "loss": 0.3291,
+      "step": 4741
+    },
+    {
+      "epoch": 0.7091371317481681,
+      "grad_norm": 1.5987346246232053,
+      "learning_rate": 2.059506520520632e-06,
+      "loss": 0.4413,
+      "step": 4742
+    },
+    {
+      "epoch": 0.7092866756393001,
+      "grad_norm": 1.2557267316237954,
+      "learning_rate": 2.05754811890825e-06,
+      "loss": 0.1741,
+      "step": 4743
+    },
+    {
+      "epoch": 0.7094362195304322,
+      "grad_norm": 1.347421034036709,
+      "learning_rate": 2.055590407623437e-06,
+      "loss": 0.163,
+      "step": 4744
+    },
+    {
+      "epoch": 0.7095857634215642,
+      "grad_norm": 1.3672739274091292,
+      "learning_rate": 2.05363338712549e-06,
+      "loss": 0.2058,
+      "step": 4745
+    },
+    {
+      "epoch": 0.7097353073126963,
+      "grad_norm": 1.8020750267925179,
+      "learning_rate": 2.051677057873546e-06,
+      "loss": 0.2011,
+      "step": 4746
+    },
+    {
+      "epoch": 0.7098848512038283,
+      "grad_norm": 1.6621833497920957,
+      "learning_rate": 2.049721420326578e-06,
+      "loss": 0.4209,
+      "step": 4747
+    },
+    {
+      "epoch": 0.7100343950949604,
+      "grad_norm": 1.5679121222959431,
+      "learning_rate": 2.047766474943398e-06,
+      "loss": 0.3769,
+      "step": 4748
+    },
+    {
+      "epoch": 0.7101839389860924,
+      "grad_norm": 1.5967246886278332,
+      "learning_rate": 2.0458122221826542e-06,
+      "loss": 0.235,
+      "step": 4749
+    },
+    {
+      "epoch": 0.7103334828772244,
+      "grad_norm": 1.3136951590796198,
+      "learning_rate": 2.043858662502834e-06,
+      "loss": 0.2018,
+      "step": 4750
+    },
+    {
+      "epoch": 0.7104830267683565,
+      "grad_norm": 1.9399765437775247,
+      "learning_rate": 2.0419057963622613e-06,
+      "loss": 0.3376,
+      "step": 4751
+    },
+    {
+      "epoch": 0.7106325706594886,
+      "grad_norm": 1.3936869392113336,
+      "learning_rate": 2.039953624219098e-06,
+      "loss": 0.2123,
+      "step": 4752
+    },
+    {
+      "epoch": 0.7107821145506206,
+      "grad_norm": 1.486853182232922,
+      "learning_rate": 2.0380021465313406e-06,
+      "loss": 0.3239,
+      "step": 4753
+    },
+    {
+      "epoch": 0.7109316584417527,
+      "grad_norm": 1.7220891594328658,
+      "learning_rate": 2.0360513637568254e-06,
+      "loss": 0.2217,
+      "step": 4754
+    },
+    {
+      "epoch": 0.7110812023328847,
+      "grad_norm": 1.2988328861923717,
+      "learning_rate": 2.0341012763532243e-06,
+      "loss": 0.2319,
+      "step": 4755
+    },
+    {
+      "epoch": 0.7112307462240167,
+      "grad_norm": 1.7196881854807897,
+      "learning_rate": 2.0321518847780474e-06,
+      "loss": 0.3627,
+      "step": 4756
+    },
+    {
+      "epoch": 0.7113802901151488,
+      "grad_norm": 1.679935485808833,
+      "learning_rate": 2.0302031894886424e-06,
+      "loss": 0.1865,
+      "step": 4757
+    },
+    {
+      "epoch": 0.7115298340062809,
+      "grad_norm": 1.0926826923454207,
+      "learning_rate": 2.0282551909421886e-06,
+      "loss": 0.2034,
+      "step": 4758
+    },
+    {
+      "epoch": 0.7116793778974129,
+      "grad_norm": 0.9404517673265307,
+      "learning_rate": 2.0263078895957065e-06,
+      "loss": 0.1358,
+      "step": 4759
+    },
+    {
+      "epoch": 0.711828921788545,
+      "grad_norm": 2.3549408265695786,
+      "learning_rate": 2.0243612859060526e-06,
+      "loss": 0.5566,
+      "step": 4760
+    },
+    {
+      "epoch": 0.7119784656796769,
+      "grad_norm": 1.5894161675694791,
+      "learning_rate": 2.022415380329918e-06,
+      "loss": 0.1957,
+      "step": 4761
+    },
+    {
+      "epoch": 0.712128009570809,
+      "grad_norm": 1.486464360092802,
+      "learning_rate": 2.020470173323831e-06,
+      "loss": 0.3143,
+      "step": 4762
+    },
+    {
+      "epoch": 0.7122775534619411,
+      "grad_norm": 1.4606498955354394,
+      "learning_rate": 2.0185256653441563e-06,
+      "loss": 0.2465,
+      "step": 4763
+    },
+    {
+      "epoch": 0.7124270973530731,
+      "grad_norm": 2.090611910244458,
+      "learning_rate": 2.016581856847094e-06,
+      "loss": 0.2162,
+      "step": 4764
+    },
+    {
+      "epoch": 0.7125766412442052,
+      "grad_norm": 1.7807448909080528,
+      "learning_rate": 2.0146387482886804e-06,
+      "loss": 0.3518,
+      "step": 4765
+    },
+    {
+      "epoch": 0.7127261851353373,
+      "grad_norm": 1.6666354092011324,
+      "learning_rate": 2.012696340124788e-06,
+      "loss": 0.3949,
+      "step": 4766
+    },
+    {
+      "epoch": 0.7128757290264692,
+      "grad_norm": 1.5273494671570687,
+      "learning_rate": 2.0107546328111235e-06,
+      "loss": 0.2107,
+      "step": 4767
+    },
+    {
+      "epoch": 0.7130252729176013,
+      "grad_norm": 1.5632155877868674,
+      "learning_rate": 2.0088136268032323e-06,
+      "loss": 0.2155,
+      "step": 4768
+    },
+    {
+      "epoch": 0.7131748168087334,
+      "grad_norm": 1.4260160070005103,
+      "learning_rate": 2.0068733225564894e-06,
+      "loss": 0.3297,
+      "step": 4769
+    },
+    {
+      "epoch": 0.7133243606998654,
+      "grad_norm": 1.8836676685538953,
+      "learning_rate": 2.004933720526111e-06,
+      "loss": 0.4145,
+      "step": 4770
+    },
+    {
+      "epoch": 0.7134739045909975,
+      "grad_norm": 1.9231850260831183,
+      "learning_rate": 2.002994821167147e-06,
+      "loss": 0.6001,
+      "step": 4771
+    },
+    {
+      "epoch": 0.7136234484821296,
+      "grad_norm": 1.437024286586543,
+      "learning_rate": 2.00105662493448e-06,
+      "loss": 0.1722,
+      "step": 4772
+    },
+    {
+      "epoch": 0.7137729923732615,
+      "grad_norm": 1.6320153713595178,
+      "learning_rate": 1.9991191322828336e-06,
+      "loss": 0.3157,
+      "step": 4773
+    },
+    {
+      "epoch": 0.7139225362643936,
+      "grad_norm": 1.3855744704624602,
+      "learning_rate": 1.997182343666757e-06,
+      "loss": 0.1999,
+      "step": 4774
+    },
+    {
+      "epoch": 0.7140720801555257,
+      "grad_norm": 1.2355394755747025,
+      "learning_rate": 1.995246259540642e-06,
+      "loss": 0.1652,
+      "step": 4775
+    },
+    {
+      "epoch": 0.7142216240466577,
+      "grad_norm": 1.411860252766047,
+      "learning_rate": 1.9933108803587105e-06,
+      "loss": 0.2646,
+      "step": 4776
+    },
+    {
+      "epoch": 0.7143711679377898,
+      "grad_norm": 1.1769599835854019,
+      "learning_rate": 1.991376206575025e-06,
+      "loss": 0.1923,
+      "step": 4777
+    },
+    {
+      "epoch": 0.7145207118289217,
+      "grad_norm": 1.725317793317903,
+      "learning_rate": 1.989442238643478e-06,
+      "loss": 0.3052,
+      "step": 4778
+    },
+    {
+      "epoch": 0.7146702557200538,
+      "grad_norm": 1.2735072479571612,
+      "learning_rate": 1.987508977017794e-06,
+      "loss": 0.2956,
+      "step": 4779
+    },
+    {
+      "epoch": 0.7148197996111859,
+      "grad_norm": 1.3919949320239815,
+      "learning_rate": 1.9855764221515366e-06,
+      "loss": 0.2915,
+      "step": 4780
+    },
+    {
+      "epoch": 0.7149693435023179,
+      "grad_norm": 2.2506955383039444,
+      "learning_rate": 1.983644574498102e-06,
+      "loss": 0.6163,
+      "step": 4781
+    },
+    {
+      "epoch": 0.71511888739345,
+      "grad_norm": 1.6414432755678665,
+      "learning_rate": 1.9817134345107196e-06,
+      "loss": 0.366,
+      "step": 4782
+    },
+    {
+      "epoch": 0.7152684312845821,
+      "grad_norm": 2.14593769293468,
+      "learning_rate": 1.979783002642456e-06,
+      "loss": 0.3334,
+      "step": 4783
+    },
+    {
+      "epoch": 0.715417975175714,
+      "grad_norm": 1.5813323184211674,
+      "learning_rate": 1.9778532793462046e-06,
+      "loss": 0.3541,
+      "step": 4784
+    },
+    {
+      "epoch": 0.7155675190668461,
+      "grad_norm": 2.027730327933588,
+      "learning_rate": 1.9759242650746992e-06,
+      "loss": 0.4575,
+      "step": 4785
+    },
+    {
+      "epoch": 0.7157170629579782,
+      "grad_norm": 1.2625581234592238,
+      "learning_rate": 1.973995960280506e-06,
+      "loss": 0.2204,
+      "step": 4786
+    },
+    {
+      "epoch": 0.7158666068491102,
+      "grad_norm": 1.567565004211618,
+      "learning_rate": 1.972068365416023e-06,
+      "loss": 0.3029,
+      "step": 4787
+    },
+    {
+      "epoch": 0.7160161507402423,
+      "grad_norm": 1.370260561499202,
+      "learning_rate": 1.9701414809334825e-06,
+      "loss": 0.1973,
+      "step": 4788
+    },
+    {
+      "epoch": 0.7161656946313744,
+      "grad_norm": 1.6439443796052537,
+      "learning_rate": 1.96821530728495e-06,
+      "loss": 0.2937,
+      "step": 4789
+    },
+    {
+      "epoch": 0.7163152385225063,
+      "grad_norm": 1.7062807681205492,
+      "learning_rate": 1.9662898449223254e-06,
+      "loss": 0.4497,
+      "step": 4790
+    },
+    {
+      "epoch": 0.7164647824136384,
+      "grad_norm": 1.7415982092251505,
+      "learning_rate": 1.9643650942973403e-06,
+      "loss": 0.359,
+      "step": 4791
+    },
+    {
+      "epoch": 0.7166143263047704,
+      "grad_norm": 1.3681752043113142,
+      "learning_rate": 1.9624410558615596e-06,
+      "loss": 0.1842,
+      "step": 4792
+    },
+    {
+      "epoch": 0.7167638701959025,
+      "grad_norm": 1.2908931102706829,
+      "learning_rate": 1.960517730066383e-06,
+      "loss": 0.2988,
+      "step": 4793
+    },
+    {
+      "epoch": 0.7169134140870346,
+      "grad_norm": 2.4298641226498687,
+      "learning_rate": 1.9585951173630376e-06,
+      "loss": 0.3598,
+      "step": 4794
+    },
+    {
+      "epoch": 0.7170629579781665,
+      "grad_norm": 1.5797060515449204,
+      "learning_rate": 1.956673218202589e-06,
+      "loss": 0.3032,
+      "step": 4795
+    },
+    {
+      "epoch": 0.7172125018692986,
+      "grad_norm": 1.7219811152387416,
+      "learning_rate": 1.9547520330359327e-06,
+      "loss": 0.3501,
+      "step": 4796
+    },
+    {
+      "epoch": 0.7173620457604307,
+      "grad_norm": 1.9552098247974243,
+      "learning_rate": 1.952831562313798e-06,
+      "loss": 0.3774,
+      "step": 4797
+    },
+    {
+      "epoch": 0.7175115896515627,
+      "grad_norm": 1.210272332516529,
+      "learning_rate": 1.9509118064867467e-06,
+      "loss": 0.1831,
+      "step": 4798
+    },
+    {
+      "epoch": 0.7176611335426948,
+      "grad_norm": 1.8071744882668304,
+      "learning_rate": 1.9489927660051696e-06,
+      "loss": 0.4436,
+      "step": 4799
+    },
+    {
+      "epoch": 0.7178106774338269,
+      "grad_norm": 1.2996122015649152,
+      "learning_rate": 1.9470744413192928e-06,
+      "loss": 0.1861,
+      "step": 4800
+    },
+    {
+      "epoch": 0.7179602213249588,
+      "grad_norm": 4.014526497100866,
+      "learning_rate": 1.945156832879174e-06,
+      "loss": 0.2961,
+      "step": 4801
+    },
+    {
+      "epoch": 0.7181097652160909,
+      "grad_norm": 1.6479807487579767,
+      "learning_rate": 1.9432399411347035e-06,
+      "loss": 0.1841,
+      "step": 4802
+    },
+    {
+      "epoch": 0.718259309107223,
+      "grad_norm": 1.0653125198511966,
+      "learning_rate": 1.9413237665356018e-06,
+      "loss": 0.201,
+      "step": 4803
+    },
+    {
+      "epoch": 0.718408852998355,
+      "grad_norm": 1.624087003839024,
+      "learning_rate": 1.939408309531422e-06,
+      "loss": 0.3056,
+      "step": 4804
+    },
+    {
+      "epoch": 0.7185583968894871,
+      "grad_norm": 5.947475483516305,
+      "learning_rate": 1.9374935705715485e-06,
+      "loss": 0.5231,
+      "step": 4805
+    },
+    {
+      "epoch": 0.7187079407806191,
+      "grad_norm": 1.6483056300420804,
+      "learning_rate": 1.9355795501051984e-06,
+      "loss": 0.2196,
+      "step": 4806
+    },
+    {
+      "epoch": 0.7188574846717511,
+      "grad_norm": 1.3894340473073978,
+      "learning_rate": 1.933666248581418e-06,
+      "loss": 0.2128,
+      "step": 4807
+    },
+    {
+      "epoch": 0.7190070285628832,
+      "grad_norm": 1.0253204031014467,
+      "learning_rate": 1.9317536664490894e-06,
+      "loss": 0.1502,
+      "step": 4808
+    },
+    {
+      "epoch": 0.7191565724540152,
+      "grad_norm": 1.6364028006620832,
+      "learning_rate": 1.9298418041569185e-06,
+      "loss": 0.3203,
+      "step": 4809
+    },
+    {
+      "epoch": 0.7193061163451473,
+      "grad_norm": 0.9078633975338446,
+      "learning_rate": 1.9279306621534488e-06,
+      "loss": 0.1478,
+      "step": 4810
+    },
+    {
+      "epoch": 0.7194556602362794,
+      "grad_norm": 1.8363602253036038,
+      "learning_rate": 1.926020240887052e-06,
+      "loss": 0.4678,
+      "step": 4811
+    },
+    {
+      "epoch": 0.7196052041274114,
+      "grad_norm": 1.736846468850028,
+      "learning_rate": 1.924110540805932e-06,
+      "loss": 0.3426,
+      "step": 4812
+    },
+    {
+      "epoch": 0.7197547480185434,
+      "grad_norm": 1.2066282537831992,
+      "learning_rate": 1.922201562358124e-06,
+      "loss": 0.2292,
+      "step": 4813
+    },
+    {
+      "epoch": 0.7199042919096755,
+      "grad_norm": 1.7762661108753415,
+      "learning_rate": 1.9202933059914904e-06,
+      "loss": 0.4426,
+      "step": 4814
+    },
+    {
+      "epoch": 0.7200538358008075,
+      "grad_norm": 1.959880308308892,
+      "learning_rate": 1.918385772153727e-06,
+      "loss": 0.3761,
+      "step": 4815
+    },
+    {
+      "epoch": 0.7202033796919396,
+      "grad_norm": 1.4598193479538135,
+      "learning_rate": 1.91647896129236e-06,
+      "loss": 0.1963,
+      "step": 4816
+    },
+    {
+      "epoch": 0.7203529235830717,
+      "grad_norm": 1.395554848072498,
+      "learning_rate": 1.9145728738547444e-06,
+      "loss": 0.3177,
+      "step": 4817
+    },
+    {
+      "epoch": 0.7205024674742037,
+      "grad_norm": 1.511186762410378,
+      "learning_rate": 1.912667510288071e-06,
+      "loss": 0.3569,
+      "step": 4818
+    },
+    {
+      "epoch": 0.7206520113653357,
+      "grad_norm": 1.3490597700277405,
+      "learning_rate": 1.9107628710393515e-06,
+      "loss": 0.2194,
+      "step": 4819
+    },
+    {
+      "epoch": 0.7208015552564677,
+      "grad_norm": 1.6936008232835473,
+      "learning_rate": 1.908858956555435e-06,
+      "loss": 0.2214,
+      "step": 4820
+    },
+    {
+      "epoch": 0.7209510991475998,
+      "grad_norm": 1.9458510254771038,
+      "learning_rate": 1.9069557672829974e-06,
+      "loss": 0.4656,
+      "step": 4821
+    },
+    {
+      "epoch": 0.7211006430387319,
+      "grad_norm": 1.81801663865579,
+      "learning_rate": 1.9050533036685458e-06,
+      "loss": 0.4798,
+      "step": 4822
+    },
+    {
+      "epoch": 0.7212501869298639,
+      "grad_norm": 1.7906911175356348,
+      "learning_rate": 1.9031515661584181e-06,
+      "loss": 0.406,
+      "step": 4823
+    },
+    {
+      "epoch": 0.721399730820996,
+      "grad_norm": 1.1444791808017973,
+      "learning_rate": 1.9012505551987764e-06,
+      "loss": 0.1843,
+      "step": 4824
+    },
+    {
+      "epoch": 0.721549274712128,
+      "grad_norm": 1.4698689388624135,
+      "learning_rate": 1.8993502712356188e-06,
+      "loss": 0.2031,
+      "step": 4825
+    },
+    {
+      "epoch": 0.72169881860326,
+      "grad_norm": 2.2276819861424375,
+      "learning_rate": 1.8974507147147692e-06,
+      "loss": 0.4352,
+      "step": 4826
+    },
+    {
+      "epoch": 0.7218483624943921,
+      "grad_norm": 1.7631748973597112,
+      "learning_rate": 1.8955518860818823e-06,
+      "loss": 0.453,
+      "step": 4827
+    },
+    {
+      "epoch": 0.7219979063855242,
+      "grad_norm": 1.7994665448141303,
+      "learning_rate": 1.8936537857824422e-06,
+      "loss": 0.3077,
+      "step": 4828
+    },
+    {
+      "epoch": 0.7221474502766562,
+      "grad_norm": 1.8288572589190288,
+      "learning_rate": 1.89175641426176e-06,
+      "loss": 0.3802,
+      "step": 4829
+    },
+    {
+      "epoch": 0.7222969941677883,
+      "grad_norm": 1.7911243468121087,
+      "learning_rate": 1.889859771964979e-06,
+      "loss": 0.3058,
+      "step": 4830
+    },
+    {
+      "epoch": 0.7224465380589203,
+      "grad_norm": 1.8399567741455625,
+      "learning_rate": 1.8879638593370686e-06,
+      "loss": 0.459,
+      "step": 4831
+    },
+    {
+      "epoch": 0.7225960819500523,
+      "grad_norm": 4.102889088039247,
+      "learning_rate": 1.8860686768228293e-06,
+      "loss": 0.3128,
+      "step": 4832
+    },
+    {
+      "epoch": 0.7227456258411844,
+      "grad_norm": 1.865842023204259,
+      "learning_rate": 1.88417422486689e-06,
+      "loss": 0.6379,
+      "step": 4833
+    },
+    {
+      "epoch": 0.7228951697323165,
+      "grad_norm": 1.35548465730419,
+      "learning_rate": 1.8822805039137038e-06,
+      "loss": 0.229,
+      "step": 4834
+    },
+    {
+      "epoch": 0.7230447136234485,
+      "grad_norm": 1.588183686824576,
+      "learning_rate": 1.8803875144075578e-06,
+      "loss": 0.3465,
+      "step": 4835
+    },
+    {
+      "epoch": 0.7231942575145806,
+      "grad_norm": 2.0359646807808205,
+      "learning_rate": 1.8784952567925662e-06,
+      "loss": 0.5241,
+      "step": 4836
+    },
+    {
+      "epoch": 0.7233438014057125,
+      "grad_norm": 2.0625164688782207,
+      "learning_rate": 1.8766037315126705e-06,
+      "loss": 0.2283,
+      "step": 4837
+    },
+    {
+      "epoch": 0.7234933452968446,
+      "grad_norm": 1.4915802204223139,
+      "learning_rate": 1.8747129390116419e-06,
+      "loss": 0.3036,
+      "step": 4838
+    },
+    {
+      "epoch": 0.7236428891879767,
+      "grad_norm": 2.157183804310608,
+      "learning_rate": 1.872822879733076e-06,
+      "loss": 0.2216,
+      "step": 4839
+    },
+    {
+      "epoch": 0.7237924330791087,
+      "grad_norm": 1.4191482750591682,
+      "learning_rate": 1.8709335541204006e-06,
+      "loss": 0.2153,
+      "step": 4840
+    },
+    {
+      "epoch": 0.7239419769702408,
+      "grad_norm": 1.6683750685067826,
+      "learning_rate": 1.8690449626168688e-06,
+      "loss": 0.3868,
+      "step": 4841
+    },
+    {
+      "epoch": 0.7240915208613729,
+      "grad_norm": 1.0627132124708731,
+      "learning_rate": 1.867157105665563e-06,
+      "loss": 0.1548,
+      "step": 4842
+    },
+    {
+      "epoch": 0.7242410647525048,
+      "grad_norm": 1.4481188000664424,
+      "learning_rate": 1.8652699837093929e-06,
+      "loss": 0.3344,
+      "step": 4843
+    },
+    {
+      "epoch": 0.7243906086436369,
+      "grad_norm": 2.015374759547004,
+      "learning_rate": 1.8633835971910952e-06,
+      "loss": 0.1807,
+      "step": 4844
+    },
+    {
+      "epoch": 0.724540152534769,
+      "grad_norm": 1.0388650670740842,
+      "learning_rate": 1.8614979465532341e-06,
+      "loss": 0.1624,
+      "step": 4845
+    },
+    {
+      "epoch": 0.724689696425901,
+      "grad_norm": 1.6403608926999131,
+      "learning_rate": 1.8596130322382011e-06,
+      "loss": 0.2447,
+      "step": 4846
+    },
+    {
+      "epoch": 0.7248392403170331,
+      "grad_norm": 1.6039276522970418,
+      "learning_rate": 1.8577288546882167e-06,
+      "loss": 0.1841,
+      "step": 4847
+    },
+    {
+      "epoch": 0.7249887842081651,
+      "grad_norm": 1.8156004235972893,
+      "learning_rate": 1.8558454143453276e-06,
+      "loss": 0.5061,
+      "step": 4848
+    },
+    {
+      "epoch": 0.7251383280992971,
+      "grad_norm": 1.5083541020940234,
+      "learning_rate": 1.8539627116514036e-06,
+      "loss": 0.2733,
+      "step": 4849
+    },
+    {
+      "epoch": 0.7252878719904292,
+      "grad_norm": 1.245802761649543,
+      "learning_rate": 1.8520807470481472e-06,
+      "loss": 0.1944,
+      "step": 4850
+    },
+    {
+      "epoch": 0.7254374158815612,
+      "grad_norm": 1.8348700402906997,
+      "learning_rate": 1.8501995209770852e-06,
+      "loss": 0.4883,
+      "step": 4851
+    },
+    {
+      "epoch": 0.7255869597726933,
+      "grad_norm": 1.8471514030276397,
+      "learning_rate": 1.848319033879571e-06,
+      "loss": 0.4281,
+      "step": 4852
+    },
+    {
+      "epoch": 0.7257365036638254,
+      "grad_norm": 2.1838526290814144,
+      "learning_rate": 1.8464392861967857e-06,
+      "loss": 0.4389,
+      "step": 4853
+    },
+    {
+      "epoch": 0.7258860475549573,
+      "grad_norm": 1.7698329643054902,
+      "learning_rate": 1.8445602783697375e-06,
+      "loss": 0.3224,
+      "step": 4854
+    },
+    {
+      "epoch": 0.7260355914460894,
+      "grad_norm": 1.926272037635014,
+      "learning_rate": 1.8426820108392569e-06,
+      "loss": 0.483,
+      "step": 4855
+    },
+    {
+      "epoch": 0.7261851353372215,
+      "grad_norm": 1.4306210768559613,
+      "learning_rate": 1.8408044840460042e-06,
+      "loss": 0.1788,
+      "step": 4856
+    },
+    {
+      "epoch": 0.7263346792283535,
+      "grad_norm": 1.391732198372493,
+      "learning_rate": 1.8389276984304645e-06,
+      "loss": 0.3414,
+      "step": 4857
+    },
+    {
+      "epoch": 0.7264842231194856,
+      "grad_norm": 1.5167767238727428,
+      "learning_rate": 1.8370516544329525e-06,
+      "loss": 0.2735,
+      "step": 4858
+    },
+    {
+      "epoch": 0.7266337670106177,
+      "grad_norm": 1.9064808924563106,
+      "learning_rate": 1.8351763524936068e-06,
+      "loss": 0.5048,
+      "step": 4859
+    },
+    {
+      "epoch": 0.7267833109017496,
+      "grad_norm": 1.158958937877205,
+      "learning_rate": 1.8333017930523872e-06,
+      "loss": 0.221,
+      "step": 4860
+    },
+    {
+      "epoch": 0.7269328547928817,
+      "grad_norm": 1.8553248379762768,
+      "learning_rate": 1.8314279765490861e-06,
+      "loss": 0.3385,
+      "step": 4861
+    },
+    {
+      "epoch": 0.7270823986840138,
+      "grad_norm": 1.4634104519416975,
+      "learning_rate": 1.8295549034233185e-06,
+      "loss": 0.172,
+      "step": 4862
+    },
+    {
+      "epoch": 0.7272319425751458,
+      "grad_norm": 1.2767641928380464,
+      "learning_rate": 1.827682574114525e-06,
+      "loss": 0.2092,
+      "step": 4863
+    },
+    {
+      "epoch": 0.7273814864662779,
+      "grad_norm": 1.677784649203382,
+      "learning_rate": 1.8258109890619741e-06,
+      "loss": 0.274,
+      "step": 4864
+    },
+    {
+      "epoch": 0.7275310303574098,
+      "grad_norm": 1.6008734639734468,
+      "learning_rate": 1.8239401487047542e-06,
+      "loss": 0.1879,
+      "step": 4865
+    },
+    {
+      "epoch": 0.7276805742485419,
+      "grad_norm": 2.1299546556280275,
+      "learning_rate": 1.8220700534817843e-06,
+      "loss": 0.5216,
+      "step": 4866
+    },
+    {
+      "epoch": 0.727830118139674,
+      "grad_norm": 1.2805904807692716,
+      "learning_rate": 1.8202007038318065e-06,
+      "loss": 0.2046,
+      "step": 4867
+    },
+    {
+      "epoch": 0.727979662030806,
+      "grad_norm": 1.6699024043147148,
+      "learning_rate": 1.818332100193389e-06,
+      "loss": 0.2377,
+      "step": 4868
+    },
+    {
+      "epoch": 0.7281292059219381,
+      "grad_norm": 2.0378048844203205,
+      "learning_rate": 1.8164642430049235e-06,
+      "loss": 0.371,
+      "step": 4869
+    },
+    {
+      "epoch": 0.7282787498130702,
+      "grad_norm": 1.2611107675280688,
+      "learning_rate": 1.8145971327046274e-06,
+      "loss": 0.1743,
+      "step": 4870
+    },
+    {
+      "epoch": 0.7284282937042021,
+      "grad_norm": 2.8489240171235926,
+      "learning_rate": 1.812730769730543e-06,
+      "loss": 0.1427,
+      "step": 4871
+    },
+    {
+      "epoch": 0.7285778375953342,
+      "grad_norm": 1.8451784710253105,
+      "learning_rate": 1.8108651545205364e-06,
+      "loss": 0.209,
+      "step": 4872
+    },
+    {
+      "epoch": 0.7287273814864663,
+      "grad_norm": 2.050825866940483,
+      "learning_rate": 1.8090002875122998e-06,
+      "loss": 0.4463,
+      "step": 4873
+    },
+    {
+      "epoch": 0.7288769253775983,
+      "grad_norm": 1.525180680093273,
+      "learning_rate": 1.8071361691433504e-06,
+      "loss": 0.2058,
+      "step": 4874
+    },
+    {
+      "epoch": 0.7290264692687304,
+      "grad_norm": 1.8675724634253468,
+      "learning_rate": 1.8052727998510244e-06,
+      "loss": 0.2193,
+      "step": 4875
+    },
+    {
+      "epoch": 0.7291760131598625,
+      "grad_norm": 1.5555303292225289,
+      "learning_rate": 1.8034101800724891e-06,
+      "loss": 0.1661,
+      "step": 4876
+    },
+    {
+      "epoch": 0.7293255570509944,
+      "grad_norm": 2.071699548717637,
+      "learning_rate": 1.8015483102447313e-06,
+      "loss": 0.4313,
+      "step": 4877
+    },
+    {
+      "epoch": 0.7294751009421265,
+      "grad_norm": 2.0316685667135346,
+      "learning_rate": 1.7996871908045648e-06,
+      "loss": 0.1906,
+      "step": 4878
+    },
+    {
+      "epoch": 0.7296246448332585,
+      "grad_norm": 1.5087990868491579,
+      "learning_rate": 1.797826822188628e-06,
+      "loss": 0.2367,
+      "step": 4879
+    },
+    {
+      "epoch": 0.7297741887243906,
+      "grad_norm": 1.3936681026294564,
+      "learning_rate": 1.7959672048333776e-06,
+      "loss": 0.3461,
+      "step": 4880
+    },
+    {
+      "epoch": 0.7299237326155227,
+      "grad_norm": 2.1186030606992894,
+      "learning_rate": 1.794108339175099e-06,
+      "loss": 0.3271,
+      "step": 4881
+    },
+    {
+      "epoch": 0.7300732765066547,
+      "grad_norm": 1.5491715251363514,
+      "learning_rate": 1.7922502256499002e-06,
+      "loss": 0.3444,
+      "step": 4882
+    },
+    {
+      "epoch": 0.7302228203977867,
+      "grad_norm": 1.9455178010759004,
+      "learning_rate": 1.790392864693713e-06,
+      "loss": 0.2189,
+      "step": 4883
+    },
+    {
+      "epoch": 0.7303723642889188,
+      "grad_norm": 2.166131466243641,
+      "learning_rate": 1.788536256742292e-06,
+      "loss": 0.5102,
+      "step": 4884
+    },
+    {
+      "epoch": 0.7305219081800508,
+      "grad_norm": 1.9183208080307343,
+      "learning_rate": 1.786680402231215e-06,
+      "loss": 0.5595,
+      "step": 4885
+    },
+    {
+      "epoch": 0.7306714520711829,
+      "grad_norm": 2.484328983288574,
+      "learning_rate": 1.784825301595884e-06,
+      "loss": 0.716,
+      "step": 4886
+    },
+    {
+      "epoch": 0.730820995962315,
+      "grad_norm": 1.763515816138748,
+      "learning_rate": 1.7829709552715225e-06,
+      "loss": 0.3711,
+      "step": 4887
+    },
+    {
+      "epoch": 0.730970539853447,
+      "grad_norm": 2.1009495277293135,
+      "learning_rate": 1.781117363693179e-06,
+      "loss": 0.3544,
+      "step": 4888
+    },
+    {
+      "epoch": 0.731120083744579,
+      "grad_norm": 1.5866469126094087,
+      "learning_rate": 1.7792645272957248e-06,
+      "loss": 0.3558,
+      "step": 4889
+    },
+    {
+      "epoch": 0.7312696276357111,
+      "grad_norm": 1.7021011459775688,
+      "learning_rate": 1.7774124465138504e-06,
+      "loss": 0.3143,
+      "step": 4890
+    },
+    {
+      "epoch": 0.7314191715268431,
+      "grad_norm": 1.6676742642163374,
+      "learning_rate": 1.7755611217820735e-06,
+      "loss": 0.3288,
+      "step": 4891
+    },
+    {
+      "epoch": 0.7315687154179752,
+      "grad_norm": 1.4418481185682541,
+      "learning_rate": 1.773710553534732e-06,
+      "loss": 0.1749,
+      "step": 4892
+    },
+    {
+      "epoch": 0.7317182593091073,
+      "grad_norm": 1.398991855838437,
+      "learning_rate": 1.771860742205988e-06,
+      "loss": 0.2401,
+      "step": 4893
+    },
+    {
+      "epoch": 0.7318678032002393,
+      "grad_norm": 1.4952873454574098,
+      "learning_rate": 1.7700116882298263e-06,
+      "loss": 0.1886,
+      "step": 4894
+    },
+    {
+      "epoch": 0.7320173470913713,
+      "grad_norm": 1.5264426985774242,
+      "learning_rate": 1.7681633920400492e-06,
+      "loss": 0.2927,
+      "step": 4895
+    },
+    {
+      "epoch": 0.7321668909825033,
+      "grad_norm": 3.238416533180956,
+      "learning_rate": 1.7663158540702875e-06,
+      "loss": 0.3544,
+      "step": 4896
+    },
+    {
+      "epoch": 0.7323164348736354,
+      "grad_norm": 1.8829901876535788,
+      "learning_rate": 1.7644690747539894e-06,
+      "loss": 0.237,
+      "step": 4897
+    },
+    {
+      "epoch": 0.7324659787647675,
+      "grad_norm": 1.8913438943497576,
+      "learning_rate": 1.7626230545244278e-06,
+      "loss": 0.4691,
+      "step": 4898
+    },
+    {
+      "epoch": 0.7326155226558995,
+      "grad_norm": 1.576090948870641,
+      "learning_rate": 1.7607777938146998e-06,
+      "loss": 0.247,
+      "step": 4899
+    },
+    {
+      "epoch": 0.7327650665470316,
+      "grad_norm": 1.6647802985248792,
+      "learning_rate": 1.758933293057718e-06,
+      "loss": 0.1818,
+      "step": 4900
+    },
+    {
+      "epoch": 0.7329146104381636,
+      "grad_norm": 1.6465920925539999,
+      "learning_rate": 1.7570895526862202e-06,
+      "loss": 0.3598,
+      "step": 4901
+    },
+    {
+      "epoch": 0.7330641543292956,
+      "grad_norm": 2.065760448476014,
+      "learning_rate": 1.7552465731327673e-06,
+      "loss": 0.4715,
+      "step": 4902
+    },
+    {
+      "epoch": 0.7332136982204277,
+      "grad_norm": 2.375540972886836,
+      "learning_rate": 1.7534043548297386e-06,
+      "loss": 0.226,
+      "step": 4903
+    },
+    {
+      "epoch": 0.7333632421115598,
+      "grad_norm": 1.2725697263760996,
+      "learning_rate": 1.751562898209339e-06,
+      "loss": 0.1696,
+      "step": 4904
+    },
+    {
+      "epoch": 0.7335127860026918,
+      "grad_norm": 2.233901453855518,
+      "learning_rate": 1.7497222037035876e-06,
+      "loss": 0.6218,
+      "step": 4905
+    },
+    {
+      "epoch": 0.7336623298938239,
+      "grad_norm": 1.9614451537767406,
+      "learning_rate": 1.7478822717443323e-06,
+      "loss": 0.3986,
+      "step": 4906
+    },
+    {
+      "epoch": 0.7338118737849559,
+      "grad_norm": 1.6046003025473952,
+      "learning_rate": 1.7460431027632368e-06,
+      "loss": 0.429,
+      "step": 4907
+    },
+    {
+      "epoch": 0.7339614176760879,
+      "grad_norm": 1.3989763819945016,
+      "learning_rate": 1.7442046971917898e-06,
+      "loss": 0.1538,
+      "step": 4908
+    },
+    {
+      "epoch": 0.73411096156722,
+      "grad_norm": 1.9975116158208546,
+      "learning_rate": 1.7423670554612976e-06,
+      "loss": 0.3242,
+      "step": 4909
+    },
+    {
+      "epoch": 0.734260505458352,
+      "grad_norm": 1.7807636600919978,
+      "learning_rate": 1.7405301780028893e-06,
+      "loss": 0.3002,
+      "step": 4910
+    },
+    {
+      "epoch": 0.7344100493494841,
+      "grad_norm": 1.8439287810205307,
+      "learning_rate": 1.7386940652475142e-06,
+      "loss": 0.425,
+      "step": 4911
+    },
+    {
+      "epoch": 0.7345595932406161,
+      "grad_norm": 1.7255465422592986,
+      "learning_rate": 1.7368587176259416e-06,
+      "loss": 0.2988,
+      "step": 4912
+    },
+    {
+      "epoch": 0.7347091371317481,
+      "grad_norm": 1.1163651100410859,
+      "learning_rate": 1.735024135568762e-06,
+      "loss": 0.2057,
+      "step": 4913
+    },
+    {
+      "epoch": 0.7348586810228802,
+      "grad_norm": 1.9838908798921489,
+      "learning_rate": 1.7331903195063871e-06,
+      "loss": 0.4594,
+      "step": 4914
+    },
+    {
+      "epoch": 0.7350082249140123,
+      "grad_norm": 1.22811541687039,
+      "learning_rate": 1.7313572698690456e-06,
+      "loss": 0.184,
+      "step": 4915
+    },
+    {
+      "epoch": 0.7351577688051443,
+      "grad_norm": 1.130576098248055,
+      "learning_rate": 1.7295249870867898e-06,
+      "loss": 0.1796,
+      "step": 4916
+    },
+    {
+      "epoch": 0.7353073126962764,
+      "grad_norm": 1.3162366435196555,
+      "learning_rate": 1.72769347158949e-06,
+      "loss": 0.2159,
+      "step": 4917
+    },
+    {
+      "epoch": 0.7354568565874084,
+      "grad_norm": 1.3336102127082576,
+      "learning_rate": 1.7258627238068387e-06,
+      "loss": 0.2036,
+      "step": 4918
+    },
+    {
+      "epoch": 0.7356064004785404,
+      "grad_norm": 1.7092416101048638,
+      "learning_rate": 1.7240327441683486e-06,
+      "loss": 0.2639,
+      "step": 4919
+    },
+    {
+      "epoch": 0.7357559443696725,
+      "grad_norm": 1.842857209339753,
+      "learning_rate": 1.722203533103346e-06,
+      "loss": 0.4788,
+      "step": 4920
+    },
+    {
+      "epoch": 0.7359054882608046,
+      "grad_norm": 1.7145975931191868,
+      "learning_rate": 1.720375091040984e-06,
+      "loss": 0.4216,
+      "step": 4921
+    },
+    {
+      "epoch": 0.7360550321519366,
+      "grad_norm": 1.3731269596018154,
+      "learning_rate": 1.7185474184102329e-06,
+      "loss": 0.1715,
+      "step": 4922
+    },
+    {
+      "epoch": 0.7362045760430687,
+      "grad_norm": 1.3050309431583982,
+      "learning_rate": 1.7167205156398813e-06,
+      "loss": 0.2078,
+      "step": 4923
+    },
+    {
+      "epoch": 0.7363541199342006,
+      "grad_norm": 1.25888103507202,
+      "learning_rate": 1.7148943831585395e-06,
+      "loss": 0.2148,
+      "step": 4924
+    },
+    {
+      "epoch": 0.7365036638253327,
+      "grad_norm": 1.4463970574157627,
+      "learning_rate": 1.7130690213946355e-06,
+      "loss": 0.2037,
+      "step": 4925
+    },
+    {
+      "epoch": 0.7366532077164648,
+      "grad_norm": 2.2062906558691804,
+      "learning_rate": 1.7112444307764159e-06,
+      "loss": 0.3348,
+      "step": 4926
+    },
+    {
+      "epoch": 0.7368027516075968,
+      "grad_norm": 1.488056015633072,
+      "learning_rate": 1.7094206117319479e-06,
+      "loss": 0.1917,
+      "step": 4927
+    },
+    {
+      "epoch": 0.7369522954987289,
+      "grad_norm": 1.9098820401759844,
+      "learning_rate": 1.7075975646891164e-06,
+      "loss": 0.4147,
+      "step": 4928
+    },
+    {
+      "epoch": 0.737101839389861,
+      "grad_norm": 1.6911497207877315,
+      "learning_rate": 1.7057752900756286e-06,
+      "loss": 0.2547,
+      "step": 4929
+    },
+    {
+      "epoch": 0.7372513832809929,
+      "grad_norm": 1.74444983156129,
+      "learning_rate": 1.7039537883190038e-06,
+      "loss": 0.2635,
+      "step": 4930
+    },
+    {
+      "epoch": 0.737400927172125,
+      "grad_norm": 1.7734453394376264,
+      "learning_rate": 1.7021330598465852e-06,
+      "loss": 0.4908,
+      "step": 4931
+    },
+    {
+      "epoch": 0.7375504710632571,
+      "grad_norm": 1.6236984629707876,
+      "learning_rate": 1.7003131050855332e-06,
+      "loss": 0.3127,
+      "step": 4932
+    },
+    {
+      "epoch": 0.7377000149543891,
+      "grad_norm": 1.5560907471994876,
+      "learning_rate": 1.6984939244628274e-06,
+      "loss": 0.3418,
+      "step": 4933
+    },
+    {
+      "epoch": 0.7378495588455212,
+      "grad_norm": 1.3794025036066626,
+      "learning_rate": 1.6966755184052664e-06,
+      "loss": 0.1274,
+      "step": 4934
+    },
+    {
+      "epoch": 0.7379991027366533,
+      "grad_norm": 1.9367957142582564,
+      "learning_rate": 1.6948578873394623e-06,
+      "loss": 0.5532,
+      "step": 4935
+    },
+    {
+      "epoch": 0.7381486466277852,
+      "grad_norm": 1.3721298422066388,
+      "learning_rate": 1.693041031691851e-06,
+      "loss": 0.1855,
+      "step": 4936
+    },
+    {
+      "epoch": 0.7382981905189173,
+      "grad_norm": 2.0769708052609626,
+      "learning_rate": 1.6912249518886837e-06,
+      "loss": 0.3111,
+      "step": 4937
+    },
+    {
+      "epoch": 0.7384477344100493,
+      "grad_norm": 2.143921562202975,
+      "learning_rate": 1.689409648356029e-06,
+      "loss": 0.2064,
+      "step": 4938
+    },
+    {
+      "epoch": 0.7385972783011814,
+      "grad_norm": 1.7748062548541788,
+      "learning_rate": 1.6875951215197779e-06,
+      "loss": 0.4532,
+      "step": 4939
+    },
+    {
+      "epoch": 0.7387468221923135,
+      "grad_norm": 1.9049841516003214,
+      "learning_rate": 1.6857813718056354e-06,
+      "loss": 0.4454,
+      "step": 4940
+    },
+    {
+      "epoch": 0.7388963660834454,
+      "grad_norm": 1.3090947430515303,
+      "learning_rate": 1.6839683996391227e-06,
+      "loss": 0.2271,
+      "step": 4941
+    },
+    {
+      "epoch": 0.7390459099745775,
+      "grad_norm": 2.237405229404865,
+      "learning_rate": 1.6821562054455804e-06,
+      "loss": 0.4707,
+      "step": 4942
+    },
+    {
+      "epoch": 0.7391954538657096,
+      "grad_norm": 1.6574169357553463,
+      "learning_rate": 1.6803447896501679e-06,
+      "loss": 0.2455,
+      "step": 4943
+    },
+    {
+      "epoch": 0.7393449977568416,
+      "grad_norm": 1.5698625859485724,
+      "learning_rate": 1.6785341526778603e-06,
+      "loss": 0.2139,
+      "step": 4944
+    },
+    {
+      "epoch": 0.7394945416479737,
+      "grad_norm": 1.1445448389646231,
+      "learning_rate": 1.6767242949534523e-06,
+      "loss": 0.1877,
+      "step": 4945
+    },
+    {
+      "epoch": 0.7396440855391058,
+      "grad_norm": 1.8341193550658388,
+      "learning_rate": 1.674915216901551e-06,
+      "loss": 0.3422,
+      "step": 4946
+    },
+    {
+      "epoch": 0.7397936294302377,
+      "grad_norm": 1.2986018162948336,
+      "learning_rate": 1.6731069189465843e-06,
+      "loss": 0.1937,
+      "step": 4947
+    },
+    {
+      "epoch": 0.7399431733213698,
+      "grad_norm": 3.4458268834752923,
+      "learning_rate": 1.6712994015127976e-06,
+      "loss": 0.1899,
+      "step": 4948
+    },
+    {
+      "epoch": 0.7400927172125019,
+      "grad_norm": 1.6288023419251345,
+      "learning_rate": 1.6694926650242504e-06,
+      "loss": 0.2016,
+      "step": 4949
+    },
+    {
+      "epoch": 0.7402422611036339,
+      "grad_norm": 1.2690199042261172,
+      "learning_rate": 1.667686709904822e-06,
+      "loss": 0.3038,
+      "step": 4950
+    },
+    {
+      "epoch": 0.740391804994766,
+      "grad_norm": 1.8684985693384317,
+      "learning_rate": 1.665881536578206e-06,
+      "loss": 0.514,
+      "step": 4951
+    },
+    {
+      "epoch": 0.7405413488858981,
+      "grad_norm": 1.9533150504153602,
+      "learning_rate": 1.6640771454679134e-06,
+      "loss": 0.3309,
+      "step": 4952
+    },
+    {
+      "epoch": 0.74069089277703,
+      "grad_norm": 1.8484944061522781,
+      "learning_rate": 1.662273536997272e-06,
+      "loss": 0.1938,
+      "step": 4953
+    },
+    {
+      "epoch": 0.7408404366681621,
+      "grad_norm": 2.028078126292207,
+      "learning_rate": 1.6604707115894259e-06,
+      "loss": 0.4777,
+      "step": 4954
+    },
+    {
+      "epoch": 0.7409899805592941,
+      "grad_norm": 1.2667704096878298,
+      "learning_rate": 1.6586686696673364e-06,
+      "loss": 0.2216,
+      "step": 4955
+    },
+    {
+      "epoch": 0.7411395244504262,
+      "grad_norm": 1.4777059517531128,
+      "learning_rate": 1.6568674116537775e-06,
+      "loss": 0.2011,
+      "step": 4956
+    },
+    {
+      "epoch": 0.7412890683415583,
+      "grad_norm": 1.6294110291319974,
+      "learning_rate": 1.6550669379713425e-06,
+      "loss": 0.1741,
+      "step": 4957
+    },
+    {
+      "epoch": 0.7414386122326903,
+      "grad_norm": 1.622994606838209,
+      "learning_rate": 1.6532672490424406e-06,
+      "loss": 0.3036,
+      "step": 4958
+    },
+    {
+      "epoch": 0.7415881561238223,
+      "grad_norm": 2.046538656570282,
+      "learning_rate": 1.6514683452892955e-06,
+      "loss": 0.2847,
+      "step": 4959
+    },
+    {
+      "epoch": 0.7417377000149544,
+      "grad_norm": 1.7786604143906892,
+      "learning_rate": 1.6496702271339487e-06,
+      "loss": 0.1677,
+      "step": 4960
+    },
+    {
+      "epoch": 0.7418872439060864,
+      "grad_norm": 1.2090273342604059,
+      "learning_rate": 1.6478728949982542e-06,
+      "loss": 0.1785,
+      "step": 4961
+    },
+    {
+      "epoch": 0.7420367877972185,
+      "grad_norm": 1.5854642471996911,
+      "learning_rate": 1.646076349303884e-06,
+      "loss": 0.2253,
+      "step": 4962
+    },
+    {
+      "epoch": 0.7421863316883506,
+      "grad_norm": 1.242112373383462,
+      "learning_rate": 1.6442805904723246e-06,
+      "loss": 0.3119,
+      "step": 4963
+    },
+    {
+      "epoch": 0.7423358755794826,
+      "grad_norm": 1.5378831663137362,
+      "learning_rate": 1.6424856189248794e-06,
+      "loss": 0.3666,
+      "step": 4964
+    },
+    {
+      "epoch": 0.7424854194706146,
+      "grad_norm": 1.5142227056599435,
+      "learning_rate": 1.6406914350826657e-06,
+      "loss": 0.3037,
+      "step": 4965
+    },
+    {
+      "epoch": 0.7426349633617467,
+      "grad_norm": 1.8172948735036927,
+      "learning_rate": 1.638898039366616e-06,
+      "loss": 0.1961,
+      "step": 4966
+    },
+    {
+      "epoch": 0.7427845072528787,
+      "grad_norm": 1.2749929231559476,
+      "learning_rate": 1.637105432197479e-06,
+      "loss": 0.166,
+      "step": 4967
+    },
+    {
+      "epoch": 0.7429340511440108,
+      "grad_norm": 2.324535822678269,
+      "learning_rate": 1.6353136139958164e-06,
+      "loss": 0.6694,
+      "step": 4968
+    },
+    {
+      "epoch": 0.7430835950351428,
+      "grad_norm": 1.480140025499841,
+      "learning_rate": 1.6335225851820068e-06,
+      "loss": 0.2349,
+      "step": 4969
+    },
+    {
+      "epoch": 0.7432331389262749,
+      "grad_norm": 4.800174733796218,
+      "learning_rate": 1.6317323461762447e-06,
+      "loss": 0.21,
+      "step": 4970
+    },
+    {
+      "epoch": 0.7433826828174069,
+      "grad_norm": 1.2593497070252637,
+      "learning_rate": 1.6299428973985332e-06,
+      "loss": 0.1609,
+      "step": 4971
+    },
+    {
+      "epoch": 0.7435322267085389,
+      "grad_norm": 1.255074221488897,
+      "learning_rate": 1.6281542392686967e-06,
+      "loss": 0.1761,
+      "step": 4972
+    },
+    {
+      "epoch": 0.743681770599671,
+      "grad_norm": 1.5729841639918394,
+      "learning_rate": 1.626366372206371e-06,
+      "loss": 0.3125,
+      "step": 4973
+    },
+    {
+      "epoch": 0.7438313144908031,
+      "grad_norm": 1.3804863151982847,
+      "learning_rate": 1.6245792966310081e-06,
+      "loss": 0.2061,
+      "step": 4974
+    },
+    {
+      "epoch": 0.7439808583819351,
+      "grad_norm": 1.4988752161362586,
+      "learning_rate": 1.6227930129618735e-06,
+      "loss": 0.2357,
+      "step": 4975
+    },
+    {
+      "epoch": 0.7441304022730671,
+      "grad_norm": 1.1559635335873584,
+      "learning_rate": 1.6210075216180438e-06,
+      "loss": 0.2257,
+      "step": 4976
+    },
+    {
+      "epoch": 0.7442799461641992,
+      "grad_norm": 1.1787528604986488,
+      "learning_rate": 1.6192228230184143e-06,
+      "loss": 0.2064,
+      "step": 4977
+    },
+    {
+      "epoch": 0.7444294900553312,
+      "grad_norm": 1.8395020477405877,
+      "learning_rate": 1.6174389175816925e-06,
+      "loss": 0.2721,
+      "step": 4978
+    },
+    {
+      "epoch": 0.7445790339464633,
+      "grad_norm": 2.2983433448079214,
+      "learning_rate": 1.6156558057263972e-06,
+      "loss": 0.3296,
+      "step": 4979
+    },
+    {
+      "epoch": 0.7447285778375954,
+      "grad_norm": 1.7476169505025692,
+      "learning_rate": 1.6138734878708701e-06,
+      "loss": 0.1581,
+      "step": 4980
+    },
+    {
+      "epoch": 0.7448781217287274,
+      "grad_norm": 1.6880815219258662,
+      "learning_rate": 1.6120919644332538e-06,
+      "loss": 0.3554,
+      "step": 4981
+    },
+    {
+      "epoch": 0.7450276656198594,
+      "grad_norm": 1.5553252000157987,
+      "learning_rate": 1.6103112358315137e-06,
+      "loss": 0.3049,
+      "step": 4982
+    },
+    {
+      "epoch": 0.7451772095109914,
+      "grad_norm": 1.508629076231468,
+      "learning_rate": 1.6085313024834248e-06,
+      "loss": 0.1949,
+      "step": 4983
+    },
+    {
+      "epoch": 0.7453267534021235,
+      "grad_norm": 1.6111729402662986,
+      "learning_rate": 1.6067521648065775e-06,
+      "loss": 0.2028,
+      "step": 4984
+    },
+    {
+      "epoch": 0.7454762972932556,
+      "grad_norm": 1.6882761786162728,
+      "learning_rate": 1.604973823218376e-06,
+      "loss": 0.4145,
+      "step": 4985
+    },
+    {
+      "epoch": 0.7456258411843876,
+      "grad_norm": 1.9886796023882398,
+      "learning_rate": 1.6031962781360327e-06,
+      "loss": 0.3631,
+      "step": 4986
+    },
+    {
+      "epoch": 0.7457753850755197,
+      "grad_norm": 1.0743322277819911,
+      "learning_rate": 1.6014195299765795e-06,
+      "loss": 0.2002,
+      "step": 4987
+    },
+    {
+      "epoch": 0.7459249289666517,
+      "grad_norm": 2.959920703560377,
+      "learning_rate": 1.599643579156857e-06,
+      "loss": 0.6486,
+      "step": 4988
+    },
+    {
+      "epoch": 0.7460744728577837,
+      "grad_norm": 1.7124238982311148,
+      "learning_rate": 1.5978684260935218e-06,
+      "loss": 0.3407,
+      "step": 4989
+    },
+    {
+      "epoch": 0.7462240167489158,
+      "grad_norm": 1.5419789507577413,
+      "learning_rate": 1.5960940712030415e-06,
+      "loss": 0.3379,
+      "step": 4990
+    },
+    {
+      "epoch": 0.7463735606400479,
+      "grad_norm": 2.0568302715565063,
+      "learning_rate": 1.5943205149016967e-06,
+      "loss": 0.1542,
+      "step": 4991
+    },
+    {
+      "epoch": 0.7465231045311799,
+      "grad_norm": 1.1442810084475554,
+      "learning_rate": 1.5925477576055808e-06,
+      "loss": 0.2005,
+      "step": 4992
+    },
+    {
+      "epoch": 0.746672648422312,
+      "grad_norm": 1.0323642072329389,
+      "learning_rate": 1.5907757997306e-06,
+      "loss": 0.1558,
+      "step": 4993
+    },
+    {
+      "epoch": 0.746822192313444,
+      "grad_norm": 1.7151094243880038,
+      "learning_rate": 1.5890046416924727e-06,
+      "loss": 0.201,
+      "step": 4994
+    },
+    {
+      "epoch": 0.746971736204576,
+      "grad_norm": 1.9038837865131188,
+      "learning_rate": 1.5872342839067305e-06,
+      "loss": 0.2425,
+      "step": 4995
+    },
+    {
+      "epoch": 0.7471212800957081,
+      "grad_norm": 1.4737968126049896,
+      "learning_rate": 1.5854647267887141e-06,
+      "loss": 0.1929,
+      "step": 4996
+    },
+    {
+      "epoch": 0.7472708239868401,
+      "grad_norm": 1.9338720487436842,
+      "learning_rate": 1.5836959707535798e-06,
+      "loss": 0.4908,
+      "step": 4997
+    },
+    {
+      "epoch": 0.7474203678779722,
+      "grad_norm": 1.7687296746572252,
+      "learning_rate": 1.5819280162162948e-06,
+      "loss": 0.4599,
+      "step": 4998
+    },
+    {
+      "epoch": 0.7475699117691043,
+      "grad_norm": 1.7216072867885186,
+      "learning_rate": 1.5801608635916387e-06,
+      "loss": 0.3506,
+      "step": 4999
+    },
+    {
+      "epoch": 0.7477194556602362,
+      "grad_norm": 1.105197711513627,
+      "learning_rate": 1.5783945132942037e-06,
+      "loss": 0.1791,
+      "step": 5000
+    },
+    {
+      "epoch": 0.7478689995513683,
+      "grad_norm": 1.8433637537725378,
+      "learning_rate": 1.5766289657383893e-06,
+      "loss": 0.2155,
+      "step": 5001
+    },
+    {
+      "epoch": 0.7480185434425004,
+      "grad_norm": 1.851082703438374,
+      "learning_rate": 1.574864221338412e-06,
+      "loss": 0.2261,
+      "step": 5002
+    },
+    {
+      "epoch": 0.7481680873336324,
+      "grad_norm": 1.1235403154462775,
+      "learning_rate": 1.5731002805082979e-06,
+      "loss": 0.1976,
+      "step": 5003
+    },
+    {
+      "epoch": 0.7483176312247645,
+      "grad_norm": 1.4005311844858053,
+      "learning_rate": 1.571337143661884e-06,
+      "loss": 0.1214,
+      "step": 5004
+    },
+    {
+      "epoch": 0.7484671751158966,
+      "grad_norm": 1.3991927499661123,
+      "learning_rate": 1.5695748112128195e-06,
+      "loss": 0.2138,
+      "step": 5005
+    },
+    {
+      "epoch": 0.7486167190070285,
+      "grad_norm": 1.5742832057053338,
+      "learning_rate": 1.5678132835745647e-06,
+      "loss": 0.3268,
+      "step": 5006
+    },
+    {
+      "epoch": 0.7487662628981606,
+      "grad_norm": 1.7017372496995755,
+      "learning_rate": 1.5660525611603904e-06,
+      "loss": 0.3067,
+      "step": 5007
+    },
+    {
+      "epoch": 0.7489158067892927,
+      "grad_norm": 2.2239030850506607,
+      "learning_rate": 1.56429264438338e-06,
+      "loss": 0.4399,
+      "step": 5008
+    },
+    {
+      "epoch": 0.7490653506804247,
+      "grad_norm": 1.4825358039189538,
+      "learning_rate": 1.5625335336564262e-06,
+      "loss": 0.1838,
+      "step": 5009
+    },
+    {
+      "epoch": 0.7492148945715568,
+      "grad_norm": 1.843373781549283,
+      "learning_rate": 1.560775229392235e-06,
+      "loss": 0.5073,
+      "step": 5010
+    },
+    {
+      "epoch": 0.7493644384626889,
+      "grad_norm": 1.6081766972792195,
+      "learning_rate": 1.5590177320033184e-06,
+      "loss": 0.2901,
+      "step": 5011
+    },
+    {
+      "epoch": 0.7495139823538208,
+      "grad_norm": 1.6372337350507788,
+      "learning_rate": 1.557261041902004e-06,
+      "loss": 0.2235,
+      "step": 5012
+    },
+    {
+      "epoch": 0.7496635262449529,
+      "grad_norm": 1.889672133817125,
+      "learning_rate": 1.5555051595004279e-06,
+      "loss": 0.3568,
+      "step": 5013
+    },
+    {
+      "epoch": 0.7498130701360849,
+      "grad_norm": 1.8353270753658117,
+      "learning_rate": 1.5537500852105376e-06,
+      "loss": 0.469,
+      "step": 5014
+    },
+    {
+      "epoch": 0.749962614027217,
+      "grad_norm": 1.2515052360240535,
+      "learning_rate": 1.5519958194440914e-06,
+      "loss": 0.1958,
+      "step": 5015
+    },
+    {
+      "epoch": 0.7501121579183491,
+      "grad_norm": 1.411832489473013,
+      "learning_rate": 1.550242362612654e-06,
+      "loss": 0.1959,
+      "step": 5016
+    },
+    {
+      "epoch": 0.750261701809481,
+      "grad_norm": 2.304156057863376,
+      "learning_rate": 1.5484897151276052e-06,
+      "loss": 0.3776,
+      "step": 5017
+    },
+    {
+      "epoch": 0.7504112457006131,
+      "grad_norm": 1.2686373160265543,
+      "learning_rate": 1.5467378774001325e-06,
+      "loss": 0.1746,
+      "step": 5018
+    },
+    {
+      "epoch": 0.7505607895917452,
+      "grad_norm": 1.4290150379380038,
+      "learning_rate": 1.544986849841234e-06,
+      "loss": 0.2043,
+      "step": 5019
+    },
+    {
+      "epoch": 0.7507103334828772,
+      "grad_norm": 1.674021514973367,
+      "learning_rate": 1.5432366328617182e-06,
+      "loss": 0.3528,
+      "step": 5020
+    },
+    {
+      "epoch": 0.7508598773740093,
+      "grad_norm": 1.8225088458049317,
+      "learning_rate": 1.541487226872202e-06,
+      "loss": 0.3571,
+      "step": 5021
+    },
+    {
+      "epoch": 0.7510094212651414,
+      "grad_norm": 1.231715373134182,
+      "learning_rate": 1.5397386322831131e-06,
+      "loss": 0.196,
+      "step": 5022
+    },
+    {
+      "epoch": 0.7511589651562733,
+      "grad_norm": 1.2398172172587025,
+      "learning_rate": 1.5379908495046892e-06,
+      "loss": 0.2013,
+      "step": 5023
+    },
+    {
+      "epoch": 0.7513085090474054,
+      "grad_norm": 1.3768389715397804,
+      "learning_rate": 1.5362438789469768e-06,
+      "loss": 0.1743,
+      "step": 5024
+    },
+    {
+      "epoch": 0.7514580529385375,
+      "grad_norm": 1.97450836406249,
+      "learning_rate": 1.5344977210198313e-06,
+      "loss": 0.2024,
+      "step": 5025
+    },
+    {
+      "epoch": 0.7516075968296695,
+      "grad_norm": 1.7489178198059616,
+      "learning_rate": 1.532752376132921e-06,
+      "loss": 0.2381,
+      "step": 5026
+    },
+    {
+      "epoch": 0.7517571407208016,
+      "grad_norm": 1.9006601987299636,
+      "learning_rate": 1.5310078446957166e-06,
+      "loss": 0.5114,
+      "step": 5027
+    },
+    {
+      "epoch": 0.7519066846119336,
+      "grad_norm": 1.9487728533894322,
+      "learning_rate": 1.5292641271175035e-06,
+      "loss": 0.3343,
+      "step": 5028
+    },
+    {
+      "epoch": 0.7520562285030656,
+      "grad_norm": 1.4232534967996326,
+      "learning_rate": 1.5275212238073755e-06,
+      "loss": 0.2949,
+      "step": 5029
+    },
+    {
+      "epoch": 0.7522057723941977,
+      "grad_norm": 2.466386702045916,
+      "learning_rate": 1.5257791351742335e-06,
+      "loss": 0.4721,
+      "step": 5030
+    },
+    {
+      "epoch": 0.7523553162853297,
+      "grad_norm": 1.116081887809245,
+      "learning_rate": 1.5240378616267887e-06,
+      "loss": 0.1345,
+      "step": 5031
+    },
+    {
+      "epoch": 0.7525048601764618,
+      "grad_norm": 1.2001444074048684,
+      "learning_rate": 1.522297403573561e-06,
+      "loss": 0.2245,
+      "step": 5032
+    },
+    {
+      "epoch": 0.7526544040675939,
+      "grad_norm": 1.401524907716876,
+      "learning_rate": 1.520557761422878e-06,
+      "loss": 0.326,
+      "step": 5033
+    },
+    {
+      "epoch": 0.7528039479587258,
+      "grad_norm": 1.6483945434180083,
+      "learning_rate": 1.518818935582877e-06,
+      "loss": 0.2961,
+      "step": 5034
+    },
+    {
+      "epoch": 0.7529534918498579,
+      "grad_norm": 1.7718583609120988,
+      "learning_rate": 1.5170809264615028e-06,
+      "loss": 0.3553,
+      "step": 5035
+    },
+    {
+      "epoch": 0.75310303574099,
+      "grad_norm": 1.5109861506112048,
+      "learning_rate": 1.5153437344665112e-06,
+      "loss": 0.2958,
+      "step": 5036
+    },
+    {
+      "epoch": 0.753252579632122,
+      "grad_norm": 1.8242522161056527,
+      "learning_rate": 1.5136073600054606e-06,
+      "loss": 0.3677,
+      "step": 5037
+    },
+    {
+      "epoch": 0.7534021235232541,
+      "grad_norm": 1.5061857510551848,
+      "learning_rate": 1.5118718034857238e-06,
+      "loss": 0.3322,
+      "step": 5038
+    },
+    {
+      "epoch": 0.7535516674143862,
+      "grad_norm": 1.7516086286447914,
+      "learning_rate": 1.5101370653144775e-06,
+      "loss": 0.3317,
+      "step": 5039
+    },
+    {
+      "epoch": 0.7537012113055181,
+      "grad_norm": 1.9410830253618627,
+      "learning_rate": 1.5084031458987086e-06,
+      "loss": 0.477,
+      "step": 5040
+    },
+    {
+      "epoch": 0.7538507551966502,
+      "grad_norm": 1.624047738522535,
+      "learning_rate": 1.5066700456452138e-06,
+      "loss": 0.2883,
+      "step": 5041
+    },
+    {
+      "epoch": 0.7540002990877822,
+      "grad_norm": 2.032351730576193,
+      "learning_rate": 1.5049377649605906e-06,
+      "loss": 0.2289,
+      "step": 5042
+    },
+    {
+      "epoch": 0.7541498429789143,
+      "grad_norm": 1.9571845552317846,
+      "learning_rate": 1.5032063042512512e-06,
+      "loss": 0.4659,
+      "step": 5043
+    },
+    {
+      "epoch": 0.7542993868700464,
+      "grad_norm": 1.4169043100292917,
+      "learning_rate": 1.5014756639234123e-06,
+      "loss": 0.2016,
+      "step": 5044
+    },
+    {
+      "epoch": 0.7544489307611784,
+      "grad_norm": 1.6271245758464608,
+      "learning_rate": 1.4997458443830997e-06,
+      "loss": 0.1912,
+      "step": 5045
+    },
+    {
+      "epoch": 0.7545984746523104,
+      "grad_norm": 1.730182102995152,
+      "learning_rate": 1.498016846036145e-06,
+      "loss": 0.1766,
+      "step": 5046
+    },
+    {
+      "epoch": 0.7547480185434425,
+      "grad_norm": 1.844207557379883,
+      "learning_rate": 1.4962886692881884e-06,
+      "loss": 0.3448,
+      "step": 5047
+    },
+    {
+      "epoch": 0.7548975624345745,
+      "grad_norm": 1.686654479519285,
+      "learning_rate": 1.4945613145446764e-06,
+      "loss": 0.3413,
+      "step": 5048
+    },
+    {
+      "epoch": 0.7550471063257066,
+      "grad_norm": 1.5689209518146487,
+      "learning_rate": 1.492834782210863e-06,
+      "loss": 0.2017,
+      "step": 5049
+    },
+    {
+      "epoch": 0.7551966502168387,
+      "grad_norm": 1.2979889962686177,
+      "learning_rate": 1.4911090726918092e-06,
+      "loss": 0.1642,
+      "step": 5050
+    },
+    {
+      "epoch": 0.7553461941079707,
+      "grad_norm": 1.509481815444356,
+      "learning_rate": 1.4893841863923858e-06,
+      "loss": 0.2808,
+      "step": 5051
+    },
+    {
+      "epoch": 0.7554957379991027,
+      "grad_norm": 1.2945562913921012,
+      "learning_rate": 1.4876601237172633e-06,
+      "loss": 0.1775,
+      "step": 5052
+    },
+    {
+      "epoch": 0.7556452818902348,
+      "grad_norm": 1.3900598384820892,
+      "learning_rate": 1.4859368850709256e-06,
+      "loss": 0.1958,
+      "step": 5053
+    },
+    {
+      "epoch": 0.7557948257813668,
+      "grad_norm": 1.599403060631151,
+      "learning_rate": 1.4842144708576606e-06,
+      "loss": 0.2444,
+      "step": 5054
+    },
+    {
+      "epoch": 0.7559443696724989,
+      "grad_norm": 1.3957515741240658,
+      "learning_rate": 1.4824928814815642e-06,
+      "loss": 0.1523,
+      "step": 5055
+    },
+    {
+      "epoch": 0.7560939135636309,
+      "grad_norm": 1.8649507157100929,
+      "learning_rate": 1.4807721173465384e-06,
+      "loss": 0.6052,
+      "step": 5056
+    },
+    {
+      "epoch": 0.756243457454763,
+      "grad_norm": 1.3054300810441961,
+      "learning_rate": 1.479052178856289e-06,
+      "loss": 0.1867,
+      "step": 5057
+    },
+    {
+      "epoch": 0.756393001345895,
+      "grad_norm": 1.2409359547056185,
+      "learning_rate": 1.4773330664143303e-06,
+      "loss": 0.2161,
+      "step": 5058
+    },
+    {
+      "epoch": 0.756542545237027,
+      "grad_norm": 1.517794796120005,
+      "learning_rate": 1.475614780423984e-06,
+      "loss": 0.3075,
+      "step": 5059
+    },
+    {
+      "epoch": 0.7566920891281591,
+      "grad_norm": 2.2683584626782114,
+      "learning_rate": 1.4738973212883744e-06,
+      "loss": 0.1662,
+      "step": 5060
+    },
+    {
+      "epoch": 0.7568416330192912,
+      "grad_norm": 1.9968898477136643,
+      "learning_rate": 1.4721806894104385e-06,
+      "loss": 0.491,
+      "step": 5061
+    },
+    {
+      "epoch": 0.7569911769104232,
+      "grad_norm": 1.646432552223488,
+      "learning_rate": 1.4704648851929104e-06,
+      "loss": 0.3103,
+      "step": 5062
+    },
+    {
+      "epoch": 0.7571407208015553,
+      "grad_norm": 1.7629010278592214,
+      "learning_rate": 1.4687499090383351e-06,
+      "loss": 0.2925,
+      "step": 5063
+    },
+    {
+      "epoch": 0.7572902646926873,
+      "grad_norm": 1.9232933811639126,
+      "learning_rate": 1.4670357613490639e-06,
+      "loss": 0.1801,
+      "step": 5064
+    },
+    {
+      "epoch": 0.7574398085838193,
+      "grad_norm": 1.686411431441648,
+      "learning_rate": 1.4653224425272512e-06,
+      "loss": 0.2741,
+      "step": 5065
+    },
+    {
+      "epoch": 0.7575893524749514,
+      "grad_norm": 1.6200170825201825,
+      "learning_rate": 1.46360995297486e-06,
+      "loss": 0.4698,
+      "step": 5066
+    },
+    {
+      "epoch": 0.7577388963660835,
+      "grad_norm": 1.6203826723565258,
+      "learning_rate": 1.4618982930936532e-06,
+      "loss": 0.2704,
+      "step": 5067
+    },
+    {
+      "epoch": 0.7578884402572155,
+      "grad_norm": 2.026413566334365,
+      "learning_rate": 1.4601874632852053e-06,
+      "loss": 0.448,
+      "step": 5068
+    },
+    {
+      "epoch": 0.7580379841483476,
+      "grad_norm": 1.879075603823234,
+      "learning_rate": 1.4584774639508931e-06,
+      "loss": 0.5288,
+      "step": 5069
+    },
+    {
+      "epoch": 0.7581875280394796,
+      "grad_norm": 1.5003273725894806,
+      "learning_rate": 1.4567682954918988e-06,
+      "loss": 0.2539,
+      "step": 5070
+    },
+    {
+      "epoch": 0.7583370719306116,
+      "grad_norm": 1.2553245034283764,
+      "learning_rate": 1.4550599583092111e-06,
+      "loss": 0.1335,
+      "step": 5071
+    },
+    {
+      "epoch": 0.7584866158217437,
+      "grad_norm": 1.5877306530313855,
+      "learning_rate": 1.453352452803618e-06,
+      "loss": 0.2332,
+      "step": 5072
+    },
+    {
+      "epoch": 0.7586361597128757,
+      "grad_norm": 1.7701093390470832,
+      "learning_rate": 1.4516457793757215e-06,
+      "loss": 0.3121,
+      "step": 5073
+    },
+    {
+      "epoch": 0.7587857036040078,
+      "grad_norm": 0.8674066155063845,
+      "learning_rate": 1.4499399384259222e-06,
+      "loss": 0.1345,
+      "step": 5074
+    },
+    {
+      "epoch": 0.7589352474951399,
+      "grad_norm": 1.2117217139497436,
+      "learning_rate": 1.4482349303544264e-06,
+      "loss": 0.195,
+      "step": 5075
+    },
+    {
+      "epoch": 0.7590847913862718,
+      "grad_norm": 2.2219961154832544,
+      "learning_rate": 1.4465307555612467e-06,
+      "loss": 0.3763,
+      "step": 5076
+    },
+    {
+      "epoch": 0.7592343352774039,
+      "grad_norm": 1.5061167634309305,
+      "learning_rate": 1.4448274144461965e-06,
+      "loss": 0.1569,
+      "step": 5077
+    },
+    {
+      "epoch": 0.759383879168536,
+      "grad_norm": 1.8680090274082126,
+      "learning_rate": 1.4431249074088976e-06,
+      "loss": 0.2104,
+      "step": 5078
+    },
+    {
+      "epoch": 0.759533423059668,
+      "grad_norm": 1.9039417503030744,
+      "learning_rate": 1.441423234848774e-06,
+      "loss": 0.273,
+      "step": 5079
+    },
+    {
+      "epoch": 0.7596829669508001,
+      "grad_norm": 1.4514849698536556,
+      "learning_rate": 1.4397223971650554e-06,
+      "loss": 0.1839,
+      "step": 5080
+    },
+    {
+      "epoch": 0.7598325108419322,
+      "grad_norm": 1.5069640247940201,
+      "learning_rate": 1.438022394756775e-06,
+      "loss": 0.3132,
+      "step": 5081
+    },
+    {
+      "epoch": 0.7599820547330641,
+      "grad_norm": 2.243758810711406,
+      "learning_rate": 1.4363232280227673e-06,
+      "loss": 0.3157,
+      "step": 5082
+    },
+    {
+      "epoch": 0.7601315986241962,
+      "grad_norm": 1.3666381604237785,
+      "learning_rate": 1.434624897361675e-06,
+      "loss": 0.1897,
+      "step": 5083
+    },
+    {
+      "epoch": 0.7602811425153283,
+      "grad_norm": 1.5194858603084764,
+      "learning_rate": 1.4329274031719427e-06,
+      "loss": 0.3733,
+      "step": 5084
+    },
+    {
+      "epoch": 0.7604306864064603,
+      "grad_norm": 1.7839247130392213,
+      "learning_rate": 1.4312307458518187e-06,
+      "loss": 0.3784,
+      "step": 5085
+    },
+    {
+      "epoch": 0.7605802302975924,
+      "grad_norm": 1.7923584025943586,
+      "learning_rate": 1.4295349257993551e-06,
+      "loss": 0.3015,
+      "step": 5086
+    },
+    {
+      "epoch": 0.7607297741887243,
+      "grad_norm": 1.6712705441976172,
+      "learning_rate": 1.427839943412408e-06,
+      "loss": 0.1954,
+      "step": 5087
+    },
+    {
+      "epoch": 0.7608793180798564,
+      "grad_norm": 1.2072895353933133,
+      "learning_rate": 1.4261457990886363e-06,
+      "loss": 0.1781,
+      "step": 5088
+    },
+    {
+      "epoch": 0.7610288619709885,
+      "grad_norm": 1.5042497663667336,
+      "learning_rate": 1.4244524932255026e-06,
+      "loss": 0.1984,
+      "step": 5089
+    },
+    {
+      "epoch": 0.7611784058621205,
+      "grad_norm": 2.107355295283857,
+      "learning_rate": 1.4227600262202735e-06,
+      "loss": 0.4434,
+      "step": 5090
+    },
+    {
+      "epoch": 0.7613279497532526,
+      "grad_norm": 1.8766492583764405,
+      "learning_rate": 1.421068398470018e-06,
+      "loss": 0.2612,
+      "step": 5091
+    },
+    {
+      "epoch": 0.7614774936443847,
+      "grad_norm": 1.796138305576019,
+      "learning_rate": 1.4193776103716067e-06,
+      "loss": 0.2362,
+      "step": 5092
+    },
+    {
+      "epoch": 0.7616270375355166,
+      "grad_norm": 2.181925533887432,
+      "learning_rate": 1.4176876623217155e-06,
+      "loss": 0.2451,
+      "step": 5093
+    },
+    {
+      "epoch": 0.7617765814266487,
+      "grad_norm": 1.856890477039263,
+      "learning_rate": 1.4159985547168231e-06,
+      "loss": 0.4182,
+      "step": 5094
+    },
+    {
+      "epoch": 0.7619261253177808,
+      "grad_norm": 2.1785753345566157,
+      "learning_rate": 1.4143102879532095e-06,
+      "loss": 0.3683,
+      "step": 5095
+    },
+    {
+      "epoch": 0.7620756692089128,
+      "grad_norm": 2.39515228059131,
+      "learning_rate": 1.412622862426961e-06,
+      "loss": 0.4871,
+      "step": 5096
+    },
+    {
+      "epoch": 0.7622252131000449,
+      "grad_norm": 1.8122946113103227,
+      "learning_rate": 1.4109362785339603e-06,
+      "loss": 0.3102,
+      "step": 5097
+    },
+    {
+      "epoch": 0.762374756991177,
+      "grad_norm": 2.025031228030744,
+      "learning_rate": 1.4092505366698977e-06,
+      "loss": 0.3625,
+      "step": 5098
+    },
+    {
+      "epoch": 0.7625243008823089,
+      "grad_norm": 1.914219199792331,
+      "learning_rate": 1.4075656372302643e-06,
+      "loss": 0.197,
+      "step": 5099
+    },
+    {
+      "epoch": 0.762673844773441,
+      "grad_norm": 1.9228829981383313,
+      "learning_rate": 1.4058815806103542e-06,
+      "loss": 0.3423,
+      "step": 5100
+    },
+    {
+      "epoch": 0.762823388664573,
+      "grad_norm": 1.4348421326181278,
+      "learning_rate": 1.404198367205263e-06,
+      "loss": 0.203,
+      "step": 5101
+    },
+    {
+      "epoch": 0.7629729325557051,
+      "grad_norm": 2.226462398755096,
+      "learning_rate": 1.402515997409889e-06,
+      "loss": 0.5939,
+      "step": 5102
+    },
+    {
+      "epoch": 0.7631224764468372,
+      "grad_norm": 2.4144455997712044,
+      "learning_rate": 1.400834471618932e-06,
+      "loss": 0.5583,
+      "step": 5103
+    },
+    {
+      "epoch": 0.7632720203379691,
+      "grad_norm": 1.8002819684827804,
+      "learning_rate": 1.3991537902268949e-06,
+      "loss": 0.3969,
+      "step": 5104
+    },
+    {
+      "epoch": 0.7634215642291012,
+      "grad_norm": 1.131912222291808,
+      "learning_rate": 1.3974739536280813e-06,
+      "loss": 0.1708,
+      "step": 5105
+    },
+    {
+      "epoch": 0.7635711081202333,
+      "grad_norm": 1.1518573963105492,
+      "learning_rate": 1.3957949622165984e-06,
+      "loss": 0.2141,
+      "step": 5106
+    },
+    {
+      "epoch": 0.7637206520113653,
+      "grad_norm": 1.261206361936778,
+      "learning_rate": 1.3941168163863505e-06,
+      "loss": 0.1754,
+      "step": 5107
+    },
+    {
+      "epoch": 0.7638701959024974,
+      "grad_norm": 1.483154100959671,
+      "learning_rate": 1.392439516531049e-06,
+      "loss": 0.2928,
+      "step": 5108
+    },
+    {
+      "epoch": 0.7640197397936295,
+      "grad_norm": 2.0259968651342946,
+      "learning_rate": 1.3907630630442044e-06,
+      "loss": 0.3666,
+      "step": 5109
+    },
+    {
+      "epoch": 0.7641692836847614,
+      "grad_norm": 1.643634583406319,
+      "learning_rate": 1.3890874563191288e-06,
+      "loss": 0.164,
+      "step": 5110
+    },
+    {
+      "epoch": 0.7643188275758935,
+      "grad_norm": 2.3198327361640825,
+      "learning_rate": 1.3874126967489355e-06,
+      "loss": 0.6347,
+      "step": 5111
+    },
+    {
+      "epoch": 0.7644683714670256,
+      "grad_norm": 2.0489051011628545,
+      "learning_rate": 1.3857387847265391e-06,
+      "loss": 0.1899,
+      "step": 5112
+    },
+    {
+      "epoch": 0.7646179153581576,
+      "grad_norm": 1.451286296004528,
+      "learning_rate": 1.3840657206446562e-06,
+      "loss": 0.2919,
+      "step": 5113
+    },
+    {
+      "epoch": 0.7647674592492897,
+      "grad_norm": 1.3652283718866698,
+      "learning_rate": 1.3823935048958026e-06,
+      "loss": 0.2496,
+      "step": 5114
+    },
+    {
+      "epoch": 0.7649170031404217,
+      "grad_norm": 1.2708107882464574,
+      "learning_rate": 1.380722137872298e-06,
+      "loss": 0.2123,
+      "step": 5115
+    },
+    {
+      "epoch": 0.7650665470315537,
+      "grad_norm": 1.7482422831337887,
+      "learning_rate": 1.379051619966259e-06,
+      "loss": 0.3106,
+      "step": 5116
+    },
+    {
+      "epoch": 0.7652160909226858,
+      "grad_norm": 2.012034563626334,
+      "learning_rate": 1.3773819515696085e-06,
+      "loss": 0.4865,
+      "step": 5117
+    },
+    {
+      "epoch": 0.7653656348138178,
+      "grad_norm": 1.4340694999731798,
+      "learning_rate": 1.3757131330740636e-06,
+      "loss": 0.3106,
+      "step": 5118
+    },
+    {
+      "epoch": 0.7655151787049499,
+      "grad_norm": 1.8139431075466297,
+      "learning_rate": 1.3740451648711457e-06,
+      "loss": 0.2158,
+      "step": 5119
+    },
+    {
+      "epoch": 0.765664722596082,
+      "grad_norm": 1.6642052674277892,
+      "learning_rate": 1.3723780473521765e-06,
+      "loss": 0.3219,
+      "step": 5120
+    },
+    {
+      "epoch": 0.765814266487214,
+      "grad_norm": 1.7580275785472688,
+      "learning_rate": 1.3707117809082787e-06,
+      "loss": 0.2053,
+      "step": 5121
+    },
+    {
+      "epoch": 0.765963810378346,
+      "grad_norm": 1.6002774083782452,
+      "learning_rate": 1.369046365930375e-06,
+      "loss": 0.1749,
+      "step": 5122
+    },
+    {
+      "epoch": 0.7661133542694781,
+      "grad_norm": 1.4260068897783758,
+      "learning_rate": 1.367381802809185e-06,
+      "loss": 0.1994,
+      "step": 5123
+    },
+    {
+      "epoch": 0.7662628981606101,
+      "grad_norm": 1.2705607099025549,
+      "learning_rate": 1.3657180919352336e-06,
+      "loss": 0.1984,
+      "step": 5124
+    },
+    {
+      "epoch": 0.7664124420517422,
+      "grad_norm": 1.2946783271168116,
+      "learning_rate": 1.3640552336988421e-06,
+      "loss": 0.3071,
+      "step": 5125
+    },
+    {
+      "epoch": 0.7665619859428743,
+      "grad_norm": 2.030415403862267,
+      "learning_rate": 1.3623932284901342e-06,
+      "loss": 0.3528,
+      "step": 5126
+    },
+    {
+      "epoch": 0.7667115298340063,
+      "grad_norm": 2.24138226820361,
+      "learning_rate": 1.3607320766990311e-06,
+      "loss": 0.5005,
+      "step": 5127
+    },
+    {
+      "epoch": 0.7668610737251383,
+      "grad_norm": 1.8442276481420865,
+      "learning_rate": 1.3590717787152564e-06,
+      "loss": 0.3878,
+      "step": 5128
+    },
+    {
+      "epoch": 0.7670106176162704,
+      "grad_norm": 1.1641423978843881,
+      "learning_rate": 1.3574123349283314e-06,
+      "loss": 0.1934,
+      "step": 5129
+    },
+    {
+      "epoch": 0.7671601615074024,
+      "grad_norm": 2.0359668691512005,
+      "learning_rate": 1.3557537457275772e-06,
+      "loss": 0.5111,
+      "step": 5130
+    },
+    {
+      "epoch": 0.7673097053985345,
+      "grad_norm": 1.2486381085511884,
+      "learning_rate": 1.3540960115021151e-06,
+      "loss": 0.1556,
+      "step": 5131
+    },
+    {
+      "epoch": 0.7674592492896665,
+      "grad_norm": 1.5112396289258176,
+      "learning_rate": 1.3524391326408675e-06,
+      "loss": 0.1686,
+      "step": 5132
+    },
+    {
+      "epoch": 0.7676087931807986,
+      "grad_norm": 1.7511319178239615,
+      "learning_rate": 1.3507831095325507e-06,
+      "loss": 0.3124,
+      "step": 5133
+    },
+    {
+      "epoch": 0.7677583370719306,
+      "grad_norm": 1.3527522831391534,
+      "learning_rate": 1.349127942565685e-06,
+      "loss": 0.1773,
+      "step": 5134
+    },
+    {
+      "epoch": 0.7679078809630626,
+      "grad_norm": 1.9345906394726324,
+      "learning_rate": 1.3474736321285892e-06,
+      "loss": 0.4342,
+      "step": 5135
+    },
+    {
+      "epoch": 0.7680574248541947,
+      "grad_norm": 0.9833467274833904,
+      "learning_rate": 1.3458201786093795e-06,
+      "loss": 0.1713,
+      "step": 5136
+    },
+    {
+      "epoch": 0.7682069687453268,
+      "grad_norm": 1.988559163874012,
+      "learning_rate": 1.3441675823959743e-06,
+      "loss": 0.2121,
+      "step": 5137
+    },
+    {
+      "epoch": 0.7683565126364588,
+      "grad_norm": 1.6989904140653285,
+      "learning_rate": 1.3425158438760854e-06,
+      "loss": 0.3356,
+      "step": 5138
+    },
+    {
+      "epoch": 0.7685060565275909,
+      "grad_norm": 1.7597113504624267,
+      "learning_rate": 1.3408649634372278e-06,
+      "loss": 0.2969,
+      "step": 5139
+    },
+    {
+      "epoch": 0.7686556004187229,
+      "grad_norm": 1.5808745547264054,
+      "learning_rate": 1.3392149414667144e-06,
+      "loss": 0.3108,
+      "step": 5140
+    },
+    {
+      "epoch": 0.7688051443098549,
+      "grad_norm": 1.183072326233728,
+      "learning_rate": 1.337565778351656e-06,
+      "loss": 0.1785,
+      "step": 5141
+    },
+    {
+      "epoch": 0.768954688200987,
+      "grad_norm": 1.377181885025977,
+      "learning_rate": 1.3359174744789627e-06,
+      "loss": 0.1877,
+      "step": 5142
+    },
+    {
+      "epoch": 0.7691042320921191,
+      "grad_norm": 1.6390355630703077,
+      "learning_rate": 1.3342700302353411e-06,
+      "loss": 0.1863,
+      "step": 5143
+    },
+    {
+      "epoch": 0.7692537759832511,
+      "grad_norm": 1.4233754017806632,
+      "learning_rate": 1.332623446007299e-06,
+      "loss": 0.2292,
+      "step": 5144
+    },
+    {
+      "epoch": 0.7694033198743832,
+      "grad_norm": 1.5453883598923903,
+      "learning_rate": 1.3309777221811398e-06,
+      "loss": 0.1724,
+      "step": 5145
+    },
+    {
+      "epoch": 0.7695528637655151,
+      "grad_norm": 1.3471592166252704,
+      "learning_rate": 1.329332859142967e-06,
+      "loss": 0.1794,
+      "step": 5146
+    },
+    {
+      "epoch": 0.7697024076566472,
+      "grad_norm": 1.866000528416247,
+      "learning_rate": 1.3276888572786829e-06,
+      "loss": 0.3123,
+      "step": 5147
+    },
+    {
+      "epoch": 0.7698519515477793,
+      "grad_norm": 1.9653238426653945,
+      "learning_rate": 1.3260457169739821e-06,
+      "loss": 0.5209,
+      "step": 5148
+    },
+    {
+      "epoch": 0.7700014954389113,
+      "grad_norm": 1.744041123921355,
+      "learning_rate": 1.324403438614364e-06,
+      "loss": 0.2774,
+      "step": 5149
+    },
+    {
+      "epoch": 0.7701510393300434,
+      "grad_norm": 1.7153767081245714,
+      "learning_rate": 1.3227620225851218e-06,
+      "loss": 0.3048,
+      "step": 5150
+    },
+    {
+      "epoch": 0.7703005832211754,
+      "grad_norm": 2.0578185400732982,
+      "learning_rate": 1.321121469271348e-06,
+      "loss": 0.4632,
+      "step": 5151
+    },
+    {
+      "epoch": 0.7704501271123074,
+      "grad_norm": 1.943828621957741,
+      "learning_rate": 1.319481779057934e-06,
+      "loss": 0.3479,
+      "step": 5152
+    },
+    {
+      "epoch": 0.7705996710034395,
+      "grad_norm": 1.9588943567254984,
+      "learning_rate": 1.3178429523295615e-06,
+      "loss": 0.3768,
+      "step": 5153
+    },
+    {
+      "epoch": 0.7707492148945716,
+      "grad_norm": 1.7862138182310494,
+      "learning_rate": 1.3162049894707208e-06,
+      "loss": 0.3491,
+      "step": 5154
+    },
+    {
+      "epoch": 0.7708987587857036,
+      "grad_norm": 1.9559502499943882,
+      "learning_rate": 1.314567890865691e-06,
+      "loss": 0.4638,
+      "step": 5155
+    },
+    {
+      "epoch": 0.7710483026768357,
+      "grad_norm": 1.6711506968064866,
+      "learning_rate": 1.3129316568985523e-06,
+      "loss": 0.2376,
+      "step": 5156
+    },
+    {
+      "epoch": 0.7711978465679677,
+      "grad_norm": 1.4349317828854198,
+      "learning_rate": 1.311296287953182e-06,
+      "loss": 0.2305,
+      "step": 5157
+    },
+    {
+      "epoch": 0.7713473904590997,
+      "grad_norm": 3.467349895371973,
+      "learning_rate": 1.3096617844132509e-06,
+      "loss": 0.2566,
+      "step": 5158
+    },
+    {
+      "epoch": 0.7714969343502318,
+      "grad_norm": 2.094062732001438,
+      "learning_rate": 1.3080281466622297e-06,
+      "loss": 0.3332,
+      "step": 5159
+    },
+    {
+      "epoch": 0.7716464782413638,
+      "grad_norm": 2.2304024427615894,
+      "learning_rate": 1.3063953750833863e-06,
+      "loss": 0.6173,
+      "step": 5160
+    },
+    {
+      "epoch": 0.7717960221324959,
+      "grad_norm": 1.5226679105199643,
+      "learning_rate": 1.3047634700597845e-06,
+      "loss": 0.205,
+      "step": 5161
+    },
+    {
+      "epoch": 0.771945566023628,
+      "grad_norm": 1.6996586942607637,
+      "learning_rate": 1.3031324319742862e-06,
+      "loss": 0.3515,
+      "step": 5162
+    },
+    {
+      "epoch": 0.7720951099147599,
+      "grad_norm": 1.4850067551054624,
+      "learning_rate": 1.3015022612095462e-06,
+      "loss": 0.1903,
+      "step": 5163
+    },
+    {
+      "epoch": 0.772244653805892,
+      "grad_norm": 1.1231636882236478,
+      "learning_rate": 1.2998729581480196e-06,
+      "loss": 0.1425,
+      "step": 5164
+    },
+    {
+      "epoch": 0.7723941976970241,
+      "grad_norm": 1.7939064424526463,
+      "learning_rate": 1.2982445231719565e-06,
+      "loss": 0.3739,
+      "step": 5165
+    },
+    {
+      "epoch": 0.7725437415881561,
+      "grad_norm": 1.6489311898199002,
+      "learning_rate": 1.2966169566634029e-06,
+      "loss": 0.3142,
+      "step": 5166
+    },
+    {
+      "epoch": 0.7726932854792882,
+      "grad_norm": 1.785459582394644,
+      "learning_rate": 1.294990259004203e-06,
+      "loss": 0.2079,
+      "step": 5167
+    },
+    {
+      "epoch": 0.7728428293704203,
+      "grad_norm": 2.0711860253417345,
+      "learning_rate": 1.293364430575994e-06,
+      "loss": 0.6204,
+      "step": 5168
+    },
+    {
+      "epoch": 0.7729923732615522,
+      "grad_norm": 1.957511424168931,
+      "learning_rate": 1.2917394717602123e-06,
+      "loss": 0.3876,
+      "step": 5169
+    },
+    {
+      "epoch": 0.7731419171526843,
+      "grad_norm": 1.521492202699478,
+      "learning_rate": 1.290115382938088e-06,
+      "loss": 0.3246,
+      "step": 5170
+    },
+    {
+      "epoch": 0.7732914610438164,
+      "grad_norm": 1.3139500725240865,
+      "learning_rate": 1.288492164490649e-06,
+      "loss": 0.1721,
+      "step": 5171
+    },
+    {
+      "epoch": 0.7734410049349484,
+      "grad_norm": 1.5448621697348717,
+      "learning_rate": 1.286869816798718e-06,
+      "loss": 0.1732,
+      "step": 5172
+    },
+    {
+      "epoch": 0.7735905488260805,
+      "grad_norm": 1.3058555765008295,
+      "learning_rate": 1.2852483402429123e-06,
+      "loss": 0.1706,
+      "step": 5173
+    },
+    {
+      "epoch": 0.7737400927172124,
+      "grad_norm": 1.4533480811391561,
+      "learning_rate": 1.2836277352036458e-06,
+      "loss": 0.3222,
+      "step": 5174
+    },
+    {
+      "epoch": 0.7738896366083445,
+      "grad_norm": 1.784889376535602,
+      "learning_rate": 1.2820080020611287e-06,
+      "loss": 0.4126,
+      "step": 5175
+    },
+    {
+      "epoch": 0.7740391804994766,
+      "grad_norm": 1.6998787486087408,
+      "learning_rate": 1.2803891411953656e-06,
+      "loss": 0.2804,
+      "step": 5176
+    },
+    {
+      "epoch": 0.7741887243906086,
+      "grad_norm": 1.593636695723748,
+      "learning_rate": 1.278771152986159e-06,
+      "loss": 0.2147,
+      "step": 5177
+    },
+    {
+      "epoch": 0.7743382682817407,
+      "grad_norm": 1.978272034241196,
+      "learning_rate": 1.2771540378131015e-06,
+      "loss": 0.3492,
+      "step": 5178
+    },
+    {
+      "epoch": 0.7744878121728728,
+      "grad_norm": 1.9360723446212331,
+      "learning_rate": 1.2755377960555848e-06,
+      "loss": 0.1768,
+      "step": 5179
+    },
+    {
+      "epoch": 0.7746373560640047,
+      "grad_norm": 1.7891672511855676,
+      "learning_rate": 1.2739224280927959e-06,
+      "loss": 0.2124,
+      "step": 5180
+    },
+    {
+      "epoch": 0.7747868999551368,
+      "grad_norm": 2.1658464164443925,
+      "learning_rate": 1.2723079343037143e-06,
+      "loss": 0.6568,
+      "step": 5181
+    },
+    {
+      "epoch": 0.7749364438462689,
+      "grad_norm": 1.47056363116747,
+      "learning_rate": 1.2706943150671163e-06,
+      "loss": 0.1576,
+      "step": 5182
+    },
+    {
+      "epoch": 0.7750859877374009,
+      "grad_norm": 2.0532403498373704,
+      "learning_rate": 1.2690815707615727e-06,
+      "loss": 0.7706,
+      "step": 5183
+    },
+    {
+      "epoch": 0.775235531628533,
+      "grad_norm": 1.5654608710902007,
+      "learning_rate": 1.267469701765449e-06,
+      "loss": 0.2848,
+      "step": 5184
+    },
+    {
+      "epoch": 0.7753850755196651,
+      "grad_norm": 1.9066760024915341,
+      "learning_rate": 1.2658587084569052e-06,
+      "loss": 0.3584,
+      "step": 5185
+    },
+    {
+      "epoch": 0.775534619410797,
+      "grad_norm": 1.8489299239183394,
+      "learning_rate": 1.2642485912138952e-06,
+      "loss": 0.5046,
+      "step": 5186
+    },
+    {
+      "epoch": 0.7756841633019291,
+      "grad_norm": 1.853116095909097,
+      "learning_rate": 1.2626393504141703e-06,
+      "loss": 0.2301,
+      "step": 5187
+    },
+    {
+      "epoch": 0.7758337071930612,
+      "grad_norm": 1.6492497580832963,
+      "learning_rate": 1.261030986435271e-06,
+      "loss": 0.3376,
+      "step": 5188
+    },
+    {
+      "epoch": 0.7759832510841932,
+      "grad_norm": 2.0608610085499293,
+      "learning_rate": 1.2594234996545358e-06,
+      "loss": 0.4805,
+      "step": 5189
+    },
+    {
+      "epoch": 0.7761327949753253,
+      "grad_norm": 1.5281935374174347,
+      "learning_rate": 1.2578168904490961e-06,
+      "loss": 0.2001,
+      "step": 5190
+    },
+    {
+      "epoch": 0.7762823388664573,
+      "grad_norm": 1.4752945527057928,
+      "learning_rate": 1.2562111591958798e-06,
+      "loss": 0.1915,
+      "step": 5191
+    },
+    {
+      "epoch": 0.7764318827575893,
+      "grad_norm": 1.2257318903732652,
+      "learning_rate": 1.2546063062716069e-06,
+      "loss": 0.1994,
+      "step": 5192
+    },
+    {
+      "epoch": 0.7765814266487214,
+      "grad_norm": 1.3328539273933164,
+      "learning_rate": 1.2530023320527885e-06,
+      "loss": 0.1759,
+      "step": 5193
+    },
+    {
+      "epoch": 0.7767309705398534,
+      "grad_norm": 2.3149216704783524,
+      "learning_rate": 1.251399236915733e-06,
+      "loss": 0.3307,
+      "step": 5194
+    },
+    {
+      "epoch": 0.7768805144309855,
+      "grad_norm": 2.207985609196849,
+      "learning_rate": 1.2497970212365445e-06,
+      "loss": 0.3746,
+      "step": 5195
+    },
+    {
+      "epoch": 0.7770300583221176,
+      "grad_norm": 1.5457106552731106,
+      "learning_rate": 1.248195685391117e-06,
+      "loss": 0.2302,
+      "step": 5196
+    },
+    {
+      "epoch": 0.7771796022132496,
+      "grad_norm": 1.8337112808430374,
+      "learning_rate": 1.2465952297551392e-06,
+      "loss": 0.3605,
+      "step": 5197
+    },
+    {
+      "epoch": 0.7773291461043816,
+      "grad_norm": 1.1180432616700489,
+      "learning_rate": 1.2449956547040947e-06,
+      "loss": 0.1833,
+      "step": 5198
+    },
+    {
+      "epoch": 0.7774786899955137,
+      "grad_norm": 2.502925484334131,
+      "learning_rate": 1.2433969606132567e-06,
+      "loss": 0.2465,
+      "step": 5199
+    },
+    {
+      "epoch": 0.7776282338866457,
+      "grad_norm": 2.2065069740569,
+      "learning_rate": 1.241799147857695e-06,
+      "loss": 0.1764,
+      "step": 5200
+    },
+    {
+      "epoch": 0.7777777777777778,
+      "grad_norm": 1.4519395658790257,
+      "learning_rate": 1.2402022168122728e-06,
+      "loss": 0.1469,
+      "step": 5201
+    },
+    {
+      "epoch": 0.7779273216689099,
+      "grad_norm": 1.861219549064482,
+      "learning_rate": 1.2386061678516454e-06,
+      "loss": 0.3527,
+      "step": 5202
+    },
+    {
+      "epoch": 0.7780768655600419,
+      "grad_norm": 1.9902252706182828,
+      "learning_rate": 1.2370110013502619e-06,
+      "loss": 0.2107,
+      "step": 5203
+    },
+    {
+      "epoch": 0.7782264094511739,
+      "grad_norm": 2.769366718134359,
+      "learning_rate": 1.2354167176823617e-06,
+      "loss": 0.3197,
+      "step": 5204
+    },
+    {
+      "epoch": 0.7783759533423059,
+      "grad_norm": 1.866646748213707,
+      "learning_rate": 1.23382331722198e-06,
+      "loss": 0.1758,
+      "step": 5205
+    },
+    {
+      "epoch": 0.778525497233438,
+      "grad_norm": 1.7292566295186207,
+      "learning_rate": 1.232230800342944e-06,
+      "loss": 0.3464,
+      "step": 5206
+    },
+    {
+      "epoch": 0.7786750411245701,
+      "grad_norm": 1.9534158032411064,
+      "learning_rate": 1.2306391674188733e-06,
+      "loss": 0.4686,
+      "step": 5207
+    },
+    {
+      "epoch": 0.7788245850157021,
+      "grad_norm": 2.059181429892903,
+      "learning_rate": 1.2290484188231817e-06,
+      "loss": 0.4902,
+      "step": 5208
+    },
+    {
+      "epoch": 0.7789741289068342,
+      "grad_norm": 1.641058928007562,
+      "learning_rate": 1.2274585549290718e-06,
+      "loss": 0.2876,
+      "step": 5209
+    },
+    {
+      "epoch": 0.7791236727979662,
+      "grad_norm": 1.6929071960032525,
+      "learning_rate": 1.225869576109543e-06,
+      "loss": 0.1776,
+      "step": 5210
+    },
+    {
+      "epoch": 0.7792732166890982,
+      "grad_norm": 1.6253245698964378,
+      "learning_rate": 1.2242814827373844e-06,
+      "loss": 0.3526,
+      "step": 5211
+    },
+    {
+      "epoch": 0.7794227605802303,
+      "grad_norm": 2.04295837808168,
+      "learning_rate": 1.2226942751851773e-06,
+      "loss": 0.5467,
+      "step": 5212
+    },
+    {
+      "epoch": 0.7795723044713624,
+      "grad_norm": 1.743780870061826,
+      "learning_rate": 1.2211079538252985e-06,
+      "loss": 0.334,
+      "step": 5213
+    },
+    {
+      "epoch": 0.7797218483624944,
+      "grad_norm": 1.9591199498980094,
+      "learning_rate": 1.2195225190299108e-06,
+      "loss": 0.4594,
+      "step": 5214
+    },
+    {
+      "epoch": 0.7798713922536264,
+      "grad_norm": 1.7219054473843514,
+      "learning_rate": 1.2179379711709738e-06,
+      "loss": 0.3348,
+      "step": 5215
+    },
+    {
+      "epoch": 0.7800209361447585,
+      "grad_norm": 2.093883168554664,
+      "learning_rate": 1.216354310620238e-06,
+      "loss": 0.2057,
+      "step": 5216
+    },
+    {
+      "epoch": 0.7801704800358905,
+      "grad_norm": 1.7834362366509595,
+      "learning_rate": 1.2147715377492453e-06,
+      "loss": 0.4279,
+      "step": 5217
+    },
+    {
+      "epoch": 0.7803200239270226,
+      "grad_norm": 1.926663024705307,
+      "learning_rate": 1.2131896529293307e-06,
+      "loss": 0.2761,
+      "step": 5218
+    },
+    {
+      "epoch": 0.7804695678181546,
+      "grad_norm": 1.96588487893358,
+      "learning_rate": 1.2116086565316172e-06,
+      "loss": 0.4397,
+      "step": 5219
+    },
+    {
+      "epoch": 0.7806191117092867,
+      "grad_norm": 1.5321799903925233,
+      "learning_rate": 1.210028548927023e-06,
+      "loss": 0.2289,
+      "step": 5220
+    },
+    {
+      "epoch": 0.7807686556004187,
+      "grad_norm": 1.8998642571797673,
+      "learning_rate": 1.2084493304862566e-06,
+      "loss": 0.3277,
+      "step": 5221
+    },
+    {
+      "epoch": 0.7809181994915507,
+      "grad_norm": 1.9600185581588903,
+      "learning_rate": 1.2068710015798173e-06,
+      "loss": 0.2714,
+      "step": 5222
+    },
+    {
+      "epoch": 0.7810677433826828,
+      "grad_norm": 1.7864009599583153,
+      "learning_rate": 1.2052935625779971e-06,
+      "loss": 0.1933,
+      "step": 5223
+    },
+    {
+      "epoch": 0.7812172872738149,
+      "grad_norm": 1.869397734653961,
+      "learning_rate": 1.2037170138508785e-06,
+      "loss": 0.4623,
+      "step": 5224
+    },
+    {
+      "epoch": 0.7813668311649469,
+      "grad_norm": 1.4205112519649394,
+      "learning_rate": 1.2021413557683341e-06,
+      "loss": 0.1947,
+      "step": 5225
+    },
+    {
+      "epoch": 0.781516375056079,
+      "grad_norm": 2.047271238805902,
+      "learning_rate": 1.200566588700029e-06,
+      "loss": 0.4764,
+      "step": 5226
+    },
+    {
+      "epoch": 0.781665918947211,
+      "grad_norm": 1.7536964843303968,
+      "learning_rate": 1.1989927130154188e-06,
+      "loss": 0.1888,
+      "step": 5227
+    },
+    {
+      "epoch": 0.781815462838343,
+      "grad_norm": 1.3080390665559112,
+      "learning_rate": 1.1974197290837513e-06,
+      "loss": 0.214,
+      "step": 5228
+    },
+    {
+      "epoch": 0.7819650067294751,
+      "grad_norm": 1.8553890211141864,
+      "learning_rate": 1.1958476372740613e-06,
+      "loss": 0.4046,
+      "step": 5229
+    },
+    {
+      "epoch": 0.7821145506206072,
+      "grad_norm": 1.2388166720119576,
+      "learning_rate": 1.194276437955177e-06,
+      "loss": 0.2053,
+      "step": 5230
+    },
+    {
+      "epoch": 0.7822640945117392,
+      "grad_norm": 1.8911900345926427,
+      "learning_rate": 1.1927061314957173e-06,
+      "loss": 0.1435,
+      "step": 5231
+    },
+    {
+      "epoch": 0.7824136384028713,
+      "grad_norm": 1.8591047459231786,
+      "learning_rate": 1.191136718264092e-06,
+      "loss": 0.5541,
+      "step": 5232
+    },
+    {
+      "epoch": 0.7825631822940032,
+      "grad_norm": 2.1194471568246764,
+      "learning_rate": 1.1895681986285013e-06,
+      "loss": 0.1896,
+      "step": 5233
+    },
+    {
+      "epoch": 0.7827127261851353,
+      "grad_norm": 1.5309028231915267,
+      "learning_rate": 1.1880005729569305e-06,
+      "loss": 0.2254,
+      "step": 5234
+    },
+    {
+      "epoch": 0.7828622700762674,
+      "grad_norm": 1.5598996459031256,
+      "learning_rate": 1.1864338416171645e-06,
+      "loss": 0.3328,
+      "step": 5235
+    },
+    {
+      "epoch": 0.7830118139673994,
+      "grad_norm": 1.474528440565962,
+      "learning_rate": 1.184868004976772e-06,
+      "loss": 0.2349,
+      "step": 5236
+    },
+    {
+      "epoch": 0.7831613578585315,
+      "grad_norm": 1.6887170014803405,
+      "learning_rate": 1.1833030634031133e-06,
+      "loss": 0.1614,
+      "step": 5237
+    },
+    {
+      "epoch": 0.7833109017496636,
+      "grad_norm": 1.979797844043359,
+      "learning_rate": 1.1817390172633402e-06,
+      "loss": 0.1771,
+      "step": 5238
+    },
+    {
+      "epoch": 0.7834604456407955,
+      "grad_norm": 2.022927768675501,
+      "learning_rate": 1.1801758669243906e-06,
+      "loss": 0.1835,
+      "step": 5239
+    },
+    {
+      "epoch": 0.7836099895319276,
+      "grad_norm": 1.6051070044772568,
+      "learning_rate": 1.1786136127529956e-06,
+      "loss": 0.1793,
+      "step": 5240
+    },
+    {
+      "epoch": 0.7837595334230597,
+      "grad_norm": 1.5363371968334916,
+      "learning_rate": 1.1770522551156749e-06,
+      "loss": 0.179,
+      "step": 5241
+    },
+    {
+      "epoch": 0.7839090773141917,
+      "grad_norm": 1.3112160726563684,
+      "learning_rate": 1.175491794378738e-06,
+      "loss": 0.1864,
+      "step": 5242
+    },
+    {
+      "epoch": 0.7840586212053238,
+      "grad_norm": 2.0595831619664957,
+      "learning_rate": 1.1739322309082863e-06,
+      "loss": 0.329,
+      "step": 5243
+    },
+    {
+      "epoch": 0.7842081650964559,
+      "grad_norm": 1.7661938552680008,
+      "learning_rate": 1.172373565070205e-06,
+      "loss": 0.3183,
+      "step": 5244
+    },
+    {
+      "epoch": 0.7843577089875878,
+      "grad_norm": 1.789087420294166,
+      "learning_rate": 1.170815797230173e-06,
+      "loss": 0.4763,
+      "step": 5245
+    },
+    {
+      "epoch": 0.7845072528787199,
+      "grad_norm": 2.0256151536550973,
+      "learning_rate": 1.1692589277536587e-06,
+      "loss": 0.3072,
+      "step": 5246
+    },
+    {
+      "epoch": 0.784656796769852,
+      "grad_norm": 1.1749305277181807,
+      "learning_rate": 1.1677029570059179e-06,
+      "loss": 0.2655,
+      "step": 5247
+    },
+    {
+      "epoch": 0.784806340660984,
+      "grad_norm": 1.679136798094466,
+      "learning_rate": 1.166147885351997e-06,
+      "loss": 0.1859,
+      "step": 5248
+    },
+    {
+      "epoch": 0.7849558845521161,
+      "grad_norm": 1.5237961664786974,
+      "learning_rate": 1.1645937131567303e-06,
+      "loss": 0.3246,
+      "step": 5249
+    },
+    {
+      "epoch": 0.785105428443248,
+      "grad_norm": 2.0072470584961386,
+      "learning_rate": 1.1630404407847412e-06,
+      "loss": 0.368,
+      "step": 5250
+    },
+    {
+      "epoch": 0.7852549723343801,
+      "grad_norm": 1.885681017017375,
+      "learning_rate": 1.1614880686004438e-06,
+      "loss": 0.5029,
+      "step": 5251
+    },
+    {
+      "epoch": 0.7854045162255122,
+      "grad_norm": 1.491129645009784,
+      "learning_rate": 1.1599365969680376e-06,
+      "loss": 0.2128,
+      "step": 5252
+    },
+    {
+      "epoch": 0.7855540601166442,
+      "grad_norm": 1.5958841470973304,
+      "learning_rate": 1.1583860262515162e-06,
+      "loss": 0.2492,
+      "step": 5253
+    },
+    {
+      "epoch": 0.7857036040077763,
+      "grad_norm": 1.2571700036739464,
+      "learning_rate": 1.156836356814654e-06,
+      "loss": 0.1872,
+      "step": 5254
+    },
+    {
+      "epoch": 0.7858531478989084,
+      "grad_norm": 1.1067274429193164,
+      "learning_rate": 1.1552875890210208e-06,
+      "loss": 0.1585,
+      "step": 5255
+    },
+    {
+      "epoch": 0.7860026917900403,
+      "grad_norm": 1.925194296846744,
+      "learning_rate": 1.153739723233972e-06,
+      "loss": 0.3717,
+      "step": 5256
+    },
+    {
+      "epoch": 0.7861522356811724,
+      "grad_norm": 1.2686743088868078,
+      "learning_rate": 1.152192759816652e-06,
+      "loss": 0.1384,
+      "step": 5257
+    },
+    {
+      "epoch": 0.7863017795723045,
+      "grad_norm": 1.7201319860534536,
+      "learning_rate": 1.1506466991319948e-06,
+      "loss": 0.318,
+      "step": 5258
+    },
+    {
+      "epoch": 0.7864513234634365,
+      "grad_norm": 2.092071828571842,
+      "learning_rate": 1.1491015415427176e-06,
+      "loss": 0.466,
+      "step": 5259
+    },
+    {
+      "epoch": 0.7866008673545686,
+      "grad_norm": 1.536769538548223,
+      "learning_rate": 1.1475572874113317e-06,
+      "loss": 0.3516,
+      "step": 5260
+    },
+    {
+      "epoch": 0.7867504112457007,
+      "grad_norm": 2.0644681744130726,
+      "learning_rate": 1.1460139371001339e-06,
+      "loss": 0.2093,
+      "step": 5261
+    },
+    {
+      "epoch": 0.7868999551368326,
+      "grad_norm": 1.3289832091207237,
+      "learning_rate": 1.1444714909712085e-06,
+      "loss": 0.179,
+      "step": 5262
+    },
+    {
+      "epoch": 0.7870494990279647,
+      "grad_norm": 2.2028406852534927,
+      "learning_rate": 1.1429299493864283e-06,
+      "loss": 0.359,
+      "step": 5263
+    },
+    {
+      "epoch": 0.7871990429190967,
+      "grad_norm": 1.9810878662502185,
+      "learning_rate": 1.1413893127074537e-06,
+      "loss": 0.2074,
+      "step": 5264
+    },
+    {
+      "epoch": 0.7873485868102288,
+      "grad_norm": 1.6394785644797076,
+      "learning_rate": 1.1398495812957333e-06,
+      "loss": 0.326,
+      "step": 5265
+    },
+    {
+      "epoch": 0.7874981307013609,
+      "grad_norm": 1.8364777318031276,
+      "learning_rate": 1.1383107555125033e-06,
+      "loss": 0.2928,
+      "step": 5266
+    },
+    {
+      "epoch": 0.7876476745924929,
+      "grad_norm": 1.5493249757699845,
+      "learning_rate": 1.1367728357187863e-06,
+      "loss": 0.2279,
+      "step": 5267
+    },
+    {
+      "epoch": 0.7877972184836249,
+      "grad_norm": 1.641815970339515,
+      "learning_rate": 1.1352358222753944e-06,
+      "loss": 0.1938,
+      "step": 5268
+    },
+    {
+      "epoch": 0.787946762374757,
+      "grad_norm": 2.37354184235881,
+      "learning_rate": 1.1336997155429235e-06,
+      "loss": 0.1637,
+      "step": 5269
+    },
+    {
+      "epoch": 0.788096306265889,
+      "grad_norm": 1.7033674653722883,
+      "learning_rate": 1.1321645158817607e-06,
+      "loss": 0.1882,
+      "step": 5270
+    },
+    {
+      "epoch": 0.7882458501570211,
+      "grad_norm": 1.2846955702747882,
+      "learning_rate": 1.1306302236520778e-06,
+      "loss": 0.1998,
+      "step": 5271
+    },
+    {
+      "epoch": 0.7883953940481532,
+      "grad_norm": 1.3745056043797175,
+      "learning_rate": 1.1290968392138351e-06,
+      "loss": 0.1856,
+      "step": 5272
+    },
+    {
+      "epoch": 0.7885449379392852,
+      "grad_norm": 1.8410278810400311,
+      "learning_rate": 1.1275643629267808e-06,
+      "loss": 0.354,
+      "step": 5273
+    },
+    {
+      "epoch": 0.7886944818304172,
+      "grad_norm": 1.5622941167611866,
+      "learning_rate": 1.126032795150445e-06,
+      "loss": 0.179,
+      "step": 5274
+    },
+    {
+      "epoch": 0.7888440257215493,
+      "grad_norm": 1.1947229651920328,
+      "learning_rate": 1.12450213624415e-06,
+      "loss": 0.2075,
+      "step": 5275
+    },
+    {
+      "epoch": 0.7889935696126813,
+      "grad_norm": 1.7705591933910914,
+      "learning_rate": 1.122972386567004e-06,
+      "loss": 0.3807,
+      "step": 5276
+    },
+    {
+      "epoch": 0.7891431135038134,
+      "grad_norm": 1.294351141284705,
+      "learning_rate": 1.1214435464779006e-06,
+      "loss": 0.1878,
+      "step": 5277
+    },
+    {
+      "epoch": 0.7892926573949454,
+      "grad_norm": 1.4514453727762606,
+      "learning_rate": 1.1199156163355214e-06,
+      "loss": 0.1872,
+      "step": 5278
+    },
+    {
+      "epoch": 0.7894422012860774,
+      "grad_norm": 1.6565436809106264,
+      "learning_rate": 1.1183885964983316e-06,
+      "loss": 0.3287,
+      "step": 5279
+    },
+    {
+      "epoch": 0.7895917451772095,
+      "grad_norm": 2.0008845608339176,
+      "learning_rate": 1.1168624873245848e-06,
+      "loss": 0.1799,
+      "step": 5280
+    },
+    {
+      "epoch": 0.7897412890683415,
+      "grad_norm": 2.182714868872967,
+      "learning_rate": 1.1153372891723225e-06,
+      "loss": 0.2159,
+      "step": 5281
+    },
+    {
+      "epoch": 0.7898908329594736,
+      "grad_norm": 1.1527198307124225,
+      "learning_rate": 1.1138130023993692e-06,
+      "loss": 0.1631,
+      "step": 5282
+    },
+    {
+      "epoch": 0.7900403768506057,
+      "grad_norm": 1.7822532798123156,
+      "learning_rate": 1.1122896273633387e-06,
+      "loss": 0.299,
+      "step": 5283
+    },
+    {
+      "epoch": 0.7901899207417377,
+      "grad_norm": 1.1673369724310274,
+      "learning_rate": 1.1107671644216305e-06,
+      "loss": 0.1843,
+      "step": 5284
+    },
+    {
+      "epoch": 0.7903394646328697,
+      "grad_norm": 1.8090909105667299,
+      "learning_rate": 1.1092456139314257e-06,
+      "loss": 0.2422,
+      "step": 5285
+    },
+    {
+      "epoch": 0.7904890085240018,
+      "grad_norm": 2.3212239437253417,
+      "learning_rate": 1.1077249762496966e-06,
+      "loss": 0.2279,
+      "step": 5286
+    },
+    {
+      "epoch": 0.7906385524151338,
+      "grad_norm": 1.7897473069927503,
+      "learning_rate": 1.1062052517331995e-06,
+      "loss": 0.484,
+      "step": 5287
+    },
+    {
+      "epoch": 0.7907880963062659,
+      "grad_norm": 2.015197139247361,
+      "learning_rate": 1.1046864407384754e-06,
+      "loss": 0.184,
+      "step": 5288
+    },
+    {
+      "epoch": 0.790937640197398,
+      "grad_norm": 1.688163150185381,
+      "learning_rate": 1.1031685436218541e-06,
+      "loss": 0.3039,
+      "step": 5289
+    },
+    {
+      "epoch": 0.79108718408853,
+      "grad_norm": 1.9066573434763772,
+      "learning_rate": 1.1016515607394469e-06,
+      "loss": 0.3647,
+      "step": 5290
+    },
+    {
+      "epoch": 0.791236727979662,
+      "grad_norm": 1.8738226015137964,
+      "learning_rate": 1.1001354924471536e-06,
+      "loss": 0.4081,
+      "step": 5291
+    },
+    {
+      "epoch": 0.7913862718707941,
+      "grad_norm": 1.7435080907369735,
+      "learning_rate": 1.0986203391006584e-06,
+      "loss": 0.1527,
+      "step": 5292
+    },
+    {
+      "epoch": 0.7915358157619261,
+      "grad_norm": 1.3645650617332317,
+      "learning_rate": 1.09710610105543e-06,
+      "loss": 0.1843,
+      "step": 5293
+    },
+    {
+      "epoch": 0.7916853596530582,
+      "grad_norm": 2.0670218389079014,
+      "learning_rate": 1.0955927786667259e-06,
+      "loss": 0.3774,
+      "step": 5294
+    },
+    {
+      "epoch": 0.7918349035441902,
+      "grad_norm": 1.8569683078727917,
+      "learning_rate": 1.0940803722895826e-06,
+      "loss": 0.2881,
+      "step": 5295
+    },
+    {
+      "epoch": 0.7919844474353223,
+      "grad_norm": 1.9175298153574774,
+      "learning_rate": 1.0925688822788266e-06,
+      "loss": 0.5207,
+      "step": 5296
+    },
+    {
+      "epoch": 0.7921339913264543,
+      "grad_norm": 1.7749496684527544,
+      "learning_rate": 1.091058308989068e-06,
+      "loss": 0.2527,
+      "step": 5297
+    },
+    {
+      "epoch": 0.7922835352175863,
+      "grad_norm": 1.2217285731938283,
+      "learning_rate": 1.0895486527747023e-06,
+      "loss": 0.1754,
+      "step": 5298
+    },
+    {
+      "epoch": 0.7924330791087184,
+      "grad_norm": 1.6995366117086494,
+      "learning_rate": 1.0880399139899105e-06,
+      "loss": 0.1996,
+      "step": 5299
+    },
+    {
+      "epoch": 0.7925826229998505,
+      "grad_norm": 1.50247326241567,
+      "learning_rate": 1.0865320929886542e-06,
+      "loss": 0.135,
+      "step": 5300
+    },
+    {
+      "epoch": 0.7927321668909825,
+      "grad_norm": 1.635063768588377,
+      "learning_rate": 1.0850251901246844e-06,
+      "loss": 0.3197,
+      "step": 5301
+    },
+    {
+      "epoch": 0.7928817107821146,
+      "grad_norm": 1.8838567884738247,
+      "learning_rate": 1.0835192057515353e-06,
+      "loss": 0.2191,
+      "step": 5302
+    },
+    {
+      "epoch": 0.7930312546732466,
+      "grad_norm": 1.3004964224536268,
+      "learning_rate": 1.0820141402225253e-06,
+      "loss": 0.2527,
+      "step": 5303
+    },
+    {
+      "epoch": 0.7931807985643786,
+      "grad_norm": 1.6665155660163822,
+      "learning_rate": 1.0805099938907566e-06,
+      "loss": 0.2872,
+      "step": 5304
+    },
+    {
+      "epoch": 0.7933303424555107,
+      "grad_norm": 1.903361555409123,
+      "learning_rate": 1.079006767109117e-06,
+      "loss": 0.4715,
+      "step": 5305
+    },
+    {
+      "epoch": 0.7934798863466428,
+      "grad_norm": 1.406430656696842,
+      "learning_rate": 1.0775044602302781e-06,
+      "loss": 0.2823,
+      "step": 5306
+    },
+    {
+      "epoch": 0.7936294302377748,
+      "grad_norm": 1.7277035890924473,
+      "learning_rate": 1.0760030736066952e-06,
+      "loss": 0.3272,
+      "step": 5307
+    },
+    {
+      "epoch": 0.7937789741289069,
+      "grad_norm": 1.1554649336199252,
+      "learning_rate": 1.0745026075906078e-06,
+      "loss": 0.1612,
+      "step": 5308
+    },
+    {
+      "epoch": 0.7939285180200388,
+      "grad_norm": 1.7791123530058242,
+      "learning_rate": 1.073003062534042e-06,
+      "loss": 0.308,
+      "step": 5309
+    },
+    {
+      "epoch": 0.7940780619111709,
+      "grad_norm": 1.9658245876851332,
+      "learning_rate": 1.0715044387888024e-06,
+      "loss": 0.2799,
+      "step": 5310
+    },
+    {
+      "epoch": 0.794227605802303,
+      "grad_norm": 1.6959397258782598,
+      "learning_rate": 1.0700067367064814e-06,
+      "loss": 0.2732,
+      "step": 5311
+    },
+    {
+      "epoch": 0.794377149693435,
+      "grad_norm": 1.81531320314136,
+      "learning_rate": 1.0685099566384548e-06,
+      "loss": 0.3395,
+      "step": 5312
+    },
+    {
+      "epoch": 0.7945266935845671,
+      "grad_norm": 1.847318033631291,
+      "learning_rate": 1.0670140989358818e-06,
+      "loss": 0.1702,
+      "step": 5313
+    },
+    {
+      "epoch": 0.7946762374756992,
+      "grad_norm": 2.0257824488414506,
+      "learning_rate": 1.0655191639497058e-06,
+      "loss": 0.3722,
+      "step": 5314
+    },
+    {
+      "epoch": 0.7948257813668311,
+      "grad_norm": 1.4192797285786958,
+      "learning_rate": 1.0640251520306493e-06,
+      "loss": 0.2118,
+      "step": 5315
+    },
+    {
+      "epoch": 0.7949753252579632,
+      "grad_norm": 1.8984963639477945,
+      "learning_rate": 1.0625320635292263e-06,
+      "loss": 0.1579,
+      "step": 5316
+    },
+    {
+      "epoch": 0.7951248691490953,
+      "grad_norm": 1.5382028977262527,
+      "learning_rate": 1.0610398987957277e-06,
+      "loss": 0.2171,
+      "step": 5317
+    },
+    {
+      "epoch": 0.7952744130402273,
+      "grad_norm": 1.7690670255014513,
+      "learning_rate": 1.0595486581802305e-06,
+      "loss": 0.2159,
+      "step": 5318
+    },
+    {
+      "epoch": 0.7954239569313594,
+      "grad_norm": 1.432027975142127,
+      "learning_rate": 1.0580583420325946e-06,
+      "loss": 0.1665,
+      "step": 5319
+    },
+    {
+      "epoch": 0.7955735008224915,
+      "grad_norm": 1.648537122057397,
+      "learning_rate": 1.05656895070246e-06,
+      "loss": 0.2823,
+      "step": 5320
+    },
+    {
+      "epoch": 0.7957230447136234,
+      "grad_norm": 2.1181607138649907,
+      "learning_rate": 1.0550804845392542e-06,
+      "loss": 0.3154,
+      "step": 5321
+    },
+    {
+      "epoch": 0.7958725886047555,
+      "grad_norm": 1.823107673084047,
+      "learning_rate": 1.0535929438921854e-06,
+      "loss": 0.4188,
+      "step": 5322
+    },
+    {
+      "epoch": 0.7960221324958875,
+      "grad_norm": 1.3206600685401113,
+      "learning_rate": 1.0521063291102446e-06,
+      "loss": 0.1963,
+      "step": 5323
+    },
+    {
+      "epoch": 0.7961716763870196,
+      "grad_norm": 1.6798018571826112,
+      "learning_rate": 1.050620640542208e-06,
+      "loss": 0.4739,
+      "step": 5324
+    },
+    {
+      "epoch": 0.7963212202781517,
+      "grad_norm": 1.9935111342215572,
+      "learning_rate": 1.049135878536629e-06,
+      "loss": 0.5629,
+      "step": 5325
+    },
+    {
+      "epoch": 0.7964707641692836,
+      "grad_norm": 2.406131698127812,
+      "learning_rate": 1.0476520434418492e-06,
+      "loss": 0.5238,
+      "step": 5326
+    },
+    {
+      "epoch": 0.7966203080604157,
+      "grad_norm": 1.413432708140045,
+      "learning_rate": 1.04616913560599e-06,
+      "loss": 0.3231,
+      "step": 5327
+    },
+    {
+      "epoch": 0.7967698519515478,
+      "grad_norm": 1.66376053841891,
+      "learning_rate": 1.0446871553769562e-06,
+      "loss": 0.3455,
+      "step": 5328
+    },
+    {
+      "epoch": 0.7969193958426798,
+      "grad_norm": 1.5649481834229166,
+      "learning_rate": 1.0432061031024343e-06,
+      "loss": 0.2248,
+      "step": 5329
+    },
+    {
+      "epoch": 0.7970689397338119,
+      "grad_norm": 1.7736064981609523,
+      "learning_rate": 1.041725979129894e-06,
+      "loss": 0.3487,
+      "step": 5330
+    },
+    {
+      "epoch": 0.797218483624944,
+      "grad_norm": 1.9953614177398118,
+      "learning_rate": 1.040246783806586e-06,
+      "loss": 0.4229,
+      "step": 5331
+    },
+    {
+      "epoch": 0.7973680275160759,
+      "grad_norm": 1.58227968973177,
+      "learning_rate": 1.0387685174795443e-06,
+      "loss": 0.1912,
+      "step": 5332
+    },
+    {
+      "epoch": 0.797517571407208,
+      "grad_norm": 1.4695702098932486,
+      "learning_rate": 1.0372911804955844e-06,
+      "loss": 0.1987,
+      "step": 5333
+    },
+    {
+      "epoch": 0.7976671152983401,
+      "grad_norm": 1.640238457511369,
+      "learning_rate": 1.0358147732013046e-06,
+      "loss": 0.1873,
+      "step": 5334
+    },
+    {
+      "epoch": 0.7978166591894721,
+      "grad_norm": 1.4936634174296828,
+      "learning_rate": 1.0343392959430826e-06,
+      "loss": 0.1916,
+      "step": 5335
+    },
+    {
+      "epoch": 0.7979662030806042,
+      "grad_norm": 1.2395120514748919,
+      "learning_rate": 1.0328647490670795e-06,
+      "loss": 0.1748,
+      "step": 5336
+    },
+    {
+      "epoch": 0.7981157469717362,
+      "grad_norm": 1.8154926750825382,
+      "learning_rate": 1.0313911329192393e-06,
+      "loss": 0.3312,
+      "step": 5337
+    },
+    {
+      "epoch": 0.7982652908628682,
+      "grad_norm": 2.1376464545398655,
+      "learning_rate": 1.0299184478452861e-06,
+      "loss": 0.3098,
+      "step": 5338
+    },
+    {
+      "epoch": 0.7984148347540003,
+      "grad_norm": 2.4417223184078236,
+      "learning_rate": 1.0284466941907278e-06,
+      "loss": 0.4359,
+      "step": 5339
+    },
+    {
+      "epoch": 0.7985643786451323,
+      "grad_norm": 1.6721365002201523,
+      "learning_rate": 1.0269758723008487e-06,
+      "loss": 0.3503,
+      "step": 5340
+    },
+    {
+      "epoch": 0.7987139225362644,
+      "grad_norm": 1.8985931112931775,
+      "learning_rate": 1.0255059825207197e-06,
+      "loss": 0.2176,
+      "step": 5341
+    },
+    {
+      "epoch": 0.7988634664273965,
+      "grad_norm": 1.8160831656299763,
+      "learning_rate": 1.024037025195191e-06,
+      "loss": 0.2219,
+      "step": 5342
+    },
+    {
+      "epoch": 0.7990130103185284,
+      "grad_norm": 1.910633721773146,
+      "learning_rate": 1.0225690006688938e-06,
+      "loss": 0.3561,
+      "step": 5343
+    },
+    {
+      "epoch": 0.7991625542096605,
+      "grad_norm": 1.5888253964377586,
+      "learning_rate": 1.0211019092862411e-06,
+      "loss": 0.1843,
+      "step": 5344
+    },
+    {
+      "epoch": 0.7993120981007926,
+      "grad_norm": 1.5931030344056767,
+      "learning_rate": 1.0196357513914261e-06,
+      "loss": 0.2058,
+      "step": 5345
+    },
+    {
+      "epoch": 0.7994616419919246,
+      "grad_norm": 2.1545258276730967,
+      "learning_rate": 1.0181705273284247e-06,
+      "loss": 0.6205,
+      "step": 5346
+    },
+    {
+      "epoch": 0.7996111858830567,
+      "grad_norm": 1.5763172805820727,
+      "learning_rate": 1.0167062374409914e-06,
+      "loss": 0.2119,
+      "step": 5347
+    },
+    {
+      "epoch": 0.7997607297741888,
+      "grad_norm": 1.9360623024652444,
+      "learning_rate": 1.0152428820726635e-06,
+      "loss": 0.2238,
+      "step": 5348
+    },
+    {
+      "epoch": 0.7999102736653207,
+      "grad_norm": 1.2466634924719853,
+      "learning_rate": 1.0137804615667596e-06,
+      "loss": 0.1847,
+      "step": 5349
+    },
+    {
+      "epoch": 0.8000598175564528,
+      "grad_norm": 2.260903832607403,
+      "learning_rate": 1.0123189762663743e-06,
+      "loss": 0.6352,
+      "step": 5350
+    },
+    {
+      "epoch": 0.8002093614475849,
+      "grad_norm": 1.6975391900562669,
+      "learning_rate": 1.0108584265143878e-06,
+      "loss": 0.2045,
+      "step": 5351
+    },
+    {
+      "epoch": 0.8003589053387169,
+      "grad_norm": 2.012799448635203,
+      "learning_rate": 1.0093988126534598e-06,
+      "loss": 0.3329,
+      "step": 5352
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 6687,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 669,
+  "total_flos": 333058895462400.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}