diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,14070 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.3001345895020188,
+  "eval_steps": 500,
+  "global_step": 2007,
+  "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
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 6687,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 669,
+  "total_flos": 124940545228800.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}